Active noise reduction system, instability detection method and control method thereof, and household appliance

Abstract

The invention provides an active noise reduction system, an instability detection method thereof, a control method and a household appliance. The instability detection method comprises the steps that sound signals are collected in the operation process of the active noise reduction system, and sound features are extracted from the sound signals; according to the sound characteristics, determining a suspected instability frequency of stable single-frequency abnormal sound in the active noise reduction system; and determining the suspected instability frequency meeting the instability condition as the final instability frequency of the stable single-frequency abnormal sound. According to the invention, two-step screening is carried out on the frequencies of the sound signals according to the sound features of the sound signals, so that the accuracy of instability detection is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of noise reduction, and in particular to an active noise reduction system and an instability detection method, a control method, and a household appliance thereof. Background Art

[0002] There are three main ways to detect instability in active noise reduction algorithms: 1. Based on the power amplitude of time domain signals, monitor power fluctuations or power size. If it suddenly increases or fluctuates violently, it is considered that instability has occurred; 2. Based on frequency domain signals, monitor the difference between the power of each frequency and the surrounding power. If the power at a certain frequency is abnormally prominent compared to the adjacent frequency band, it is considered that instability has occurred; 3. The artificial neural network method commonly used in the field of headphones, collects a large amount of voice data (abnormal sound group and normal group) in the early stage, builds a discriminant neural network, and then extracts features from the audio data to be discriminated according to a certain method, and imports it into the artificial neural network to determine whether instability has occurred. In practice, the above schemes cannot detect the instability of the active noise reduction system very well, and the accuracy of the active noise reduction algorithm needs to be improved. Summary of the invention

[0003] The technical problem to be solved by the present disclosure is to overcome the defect of inaccurate instability detection of active noise reduction systems in the prior art, and to provide an active noise reduction system and an instability detection method, a control method, and a home appliance thereof.

[0004] The present invention solves the above technical problems through the following technical solutions:

[0005] In a first aspect, a method for detecting instability of an active noise reduction system is provided, the method comprising:

[0006] Collecting sound signals during the operation of the active noise reduction system and extracting sound features from the sound signals; the sound signals include noise signals of noise sources and / or control signals of speakers of the active noise reduction system;

[0007] Determine the suspected instability of the stable single-frequency abnormal sound in the active noise reduction system according to the sound characteristics frequency;

[0008] The suspected unstable frequency that meets the instability conditions is determined as the final unstable frequency of the stable single-frequency abnormal sound.

[0009] Optionally, the sound feature includes a peak adjacent power ratio; determining the suspected instability frequency of the stable single-frequency abnormal sound in the active noise reduction system according to the sound feature extracted from the sound signal includes:

[0010] Determine the peak frequency of each frame of the sound signal in the multiple frames of sound signals respectively; the peak frequency is the frequency corresponding to the maximum value of at least three adjacent peaks in the sound signal;

[0011] Calculating a peak adjacent power ratio corresponding to the peak frequency, and determining a peak frequency corresponding to the peak adjacent power ratio falling within a preset abnormal sound range as an initial instability frequency;

[0012] The initial unstable frequencies that meet the preset conditions are determined as suspected unstable frequencies of stable single-frequency abnormal sounds.

[0013] Optionally, the preset abnormal sound range is determined according to the peak adjacent power ratio of the frequency points of the spectrum of the analog signal; wherein, the analog signal is a sound signal collected when the active noise reduction system simulates a stable single-frequency abnormal sound.

[0014] Optionally, the preset condition includes at least one of the following:

[0015] The number of occurrences of the initial instability frequency in a plurality of adjacent frames of sound signals is greater than a first instability threshold;

[0016] The probability of occurrence of the initial instability frequency in a plurality of adjacent frames of sound signals is greater than a second instability threshold;

[0017] The sum of the peak adjacent power ratios of the initial instability frequency in a plurality of adjacent frames of sound signals is greater than a third instability threshold;

[0018] An average value of peak adjacent power ratios of the initial instability frequency in a plurality of adjacent frames of sound signals is greater than a fourth instability threshold.

[0019] Optionally, determining the suspected instability frequency that meets the instability condition as the final instability frequency includes:

[0020] Partitioning the sound signal according to a preset partitioning strategy; the preset partitioning strategy is determined according to the frequency band difference of the sound signal;

[0021] For each partition, when the sound feature in the partition meets the instability condition, the suspected instability frequency in the partition is determined as the final instability frequency of the stable single-frequency abnormal sound.

[0022] Optionally, the preset partitioning strategy includes partitioning intervals; the partitioning intervals are determined based on the differences in peak values ​​of the sound signal in different frequency bands without instability; and partitioning the sound signal includes: partitioning the sound signal according to the partitioning intervals.

[0023] Optionally, for each partition, when the sound feature in the partition meets the instability condition, determining the suspected instability frequency in the partition as the final instability frequency includes:

[0024] Calculate the standard deviation of the peak adjacent power ratio corresponding to the suspected unstable frequency in the same partition;

[0025] When the standard deviation is greater than the standard deviation threshold of the corresponding partition, determining the corresponding suspected instability frequency in the partition as the final instability frequency;

[0026] The standard deviation threshold is determined according to the sound characteristics when no instability occurs;

[0027] Optionally, calculating the variance of the peak adjacent power ratio corresponding to the suspected instability frequency in the same partition;

[0028] When the variance is greater than the variance threshold corresponding to the partition, determining the suspected instability frequency corresponding to the partition as the final instability frequency;

[0029] The variance threshold is determined according to the sound characteristics when no instability occurs.

[0030] Optionally, extracting sound features from the sound signal includes: performing time domain and / or frequency domain processing on the sound signal after low-pass filtering the sound signal, and then extracting sound features from the processed sound signal.

[0031] In a second aspect, a control method for an active noise reduction system is provided, wherein the active noise reduction system includes a filter; the control method comprises: detecting a final instability frequency according to the instability detection method for the active noise reduction system described in any one of the above items; determining a filtering parameter of the filter according to the final instability frequency, and suppressing an instability signal in a sound signal through the filter.

[0032] In a third aspect, an active noise reduction system is provided, comprising: a speaker, a microphone, a processor, and a filter; the processor is respectively connected to the speaker, the microphone, and the filter, and when the processor executes a computer program, the control method or the instability detection method of the active noise reduction system described in any one of the above items is implemented.

[0033] In a fourth aspect, a household appliance is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the instability detection method or control method described in any one of the above items is implemented.

[0034] In a fifth aspect, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the instability detection method or control method described in any one of the above is implemented.

[0035] In a sixth aspect, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above-mentioned instability detection methods or control methods.

[0036] In a seventh aspect, a range hood is provided, wherein the range hood comprises the active noise reduction system described in the third aspect.

[0037] On the basis of being in accordance with the common sense in the art, the above-mentioned preferred conditions can be arbitrarily combined to obtain the preferred embodiments of the present disclosure.

[0038] The positive progressive effect of the present disclosure lies in that the present disclosure analyzes the sound features used for instability detection in the sound signal, first determines the suspected instability frequency of the stable single-frequency abnormal sound, and then determines the final instability frequency of the stable single-frequency abnormal sound, and through two-step screening, thereby improving the accuracy of instability detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A flowchart of an instability detection method for an active noise reduction system provided by an exemplary embodiment of the present disclosure;

[0040] Figure 2a A schematic diagram of a frequency domain signal curve of a stable single-frequency abnormal sound, howling, and normal noise provided by an exemplary embodiment of the present disclosure;

[0041] Figure 2b A schematic diagram of a time domain signal curve of a stable single-frequency abnormal sound, howling, and normal noise provided by an exemplary embodiment of the present disclosure;

[0042] Figure 3a This is a spectrum diagram of the active noise reduction system when howling occurs;

[0043] Figure 3b This is a spectrum diagram of the active noise reduction system when a stable single-frequency abnormal sound occurs;

[0044] Figure 4a A flow chart for determining a suspected instability frequency in an instability detection method provided by an exemplary embodiment of the present disclosure;

[0045] Figure 4b A specific flow chart of determining a suspected instability frequency and determining a final instability frequency in an instability detection method provided by an exemplary embodiment of the present disclosure;

[0046] Figure 5 The present invention is a schematic structural diagram of a household appliance according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0047] The present disclosure is further described below by way of examples, but the present disclosure is not limited to the scope of the examples.

[0048] Prefixes such as "first" and "second" are used in the embodiments of the present disclosure only to distinguish different description objects, and have no limiting effect on the position, order, priority, quantity or content of the described objects. The use of prefixes such as ordinal numbers to distinguish description objects in the embodiments of the present disclosure does not constitute a limitation on the described objects. For the statement of the described objects, please refer to the description in the context of the claims or embodiments, and no unnecessary limitation should be constituted due to the use of such prefixes. In addition, in the description of the present embodiment, unless otherwise specified, the meaning of "plurality" is two or more.

[0049] The principle of active noise reduction is to achieve noise reduction effect by canceling out sound signals. The active noise reduction system includes microphones, speakers, feedforward amplifiers, controllers, etc. Microphones are arranged around the noise source to pick up noise signals, and the picked up noise signals are transmitted to the controller through the feedforward amplifier. After analog-to-digital conversion, the controller is equipped with an operation unit, which calculates the control signal (anti-phase noise) of the speaker, and then converts the digital-to-analog signal into an analog signal and outputs it through the speaker, finally achieving noise reduction effect in the predetermined area.

[0050] The instability detection of the active noise reduction system is when the active noise reduction system is working, due to some faults or interference, the speaker of the active noise reduction system emits a strange sound, which affects the user experience. The strange sound includes howling and stable single-frequency strange sound (or stable single-frequency sound).

[0051] The characteristics of howling caused by active noise reduction are: Figure 2a and Figure 2b There is a prominent peak in the noise spectrum. The amplitude of the highest point of the peak is significantly higher than the adjacent frequency, and the peak is generally much larger than the peak of the stable single-frequency abnormal sound, accounting for a large proportion of the overall noise power. From the beginning of the howling, its power amplitude in the time domain gradually increases from small to large. The ratio of the maximum power to the minimum power is generally more than 5 times, and the amplitude is generally close to or even exceeds the maximum sound power of the speaker itself. Figure 2a The second to fifth columns of the table refer to the amplitudes at 94.28 Hz, 108.75 Hz, 0 Hz and 6000 Hz respectively, and the sixth column RMS data refers to the root mean square value of the amplitude from 0 to 6000 Hz. The amplitude can be characterized by, but not limited to, power, energy, or decibels.

[0052] Different from howling, the characteristics of stable single-frequency noise / stable single-frequency sound are: there is a prominent peak in the noise spectrum, the amplitude of the highest point of the peak (noise peak) is higher than the amplitude corresponding to the adjacent frequency, and the power of the peak accounts for a small proportion of the total noise power, that is, the power of the peak accounts for a proportion of the total noise power that is less than or equal to the power threshold, where the power threshold is calibrated according to experimental data. Figure 2aAs shown by the blue curve, the peak bandwidth does not exceed 50Hz. Figure 2b When a stable single-frequency abnormal sound is generated, its power amplitude in the time domain is relatively stable, the power changes little, the ratio of maximum power to minimum power will not exceed 3 times, and the amplitude is generally much smaller than the maximum sound power of the speaker itself. Figure 2b The left ordinate represents the power amplitude, and the right ordinate represents the sound pressure amplitude, and the two can be converted into each other.

[0053] Figure 1 A flowchart of an instability detection method for an active noise reduction system provided by an exemplary embodiment of the present disclosure. The instability detection method includes:

[0054] Step 101: Collect sound signals during the operation of the active noise reduction system, and extract sound features from the sound signals.

[0055] The sound signal can be a noise signal emitted by a noise source collected by the first microphone included in the active noise reduction system. The noise signal emitted by the noise source is subsequently used as the basic data for instability detection. The first microphone is deployed near the noise source. It should be noted that the sound signal can be the collection result of one first microphone or a combination of the collection results of multiple first microphones. At this time, the sound signal collected by the first microphone is the superposition result of the original noise signal emitted by the noise source and the feedback signal.

[0056] The sound signal can be a control signal of the speaker, which is then used as the basic data for instability detection. It should be noted that the control signal of the speaker can be obtained directly by obtaining the output result of the controller, or it can be collected by a second microphone deployed near the speaker included in the active noise reduction system. The sound signal can be the control signal of one speaker, or it can be a combination of control signals of multiple speakers.

[0057] The sound signal may also be a combination of a noise signal of a noise source and a control signal of a speaker. The specific combination is not particularly limited in the embodiment of the present disclosure.

[0058] In a stable noise environment, the control signal of the speaker is stable, so the power of the control signal remains in a stable range for a period of time. The power will increase significantly. And the power increase is reflected at a certain frequency. Based on the above introduction,There are two types of instability: the first is howling, which is a sharp sound that increases from low to high in a very short time, and finally the speaker is working at full capacity, and the sound is unacceptable to users; the second is a stable single-frequency abnormal sound, which does not gradually increase to a very sharp state, but it can be clearly heard that in addition to the noise emitted by the noise source, there is a stable or intermittent single-frequency sound, which will make people uncomfortable for a long time. The above two instability situations are a sudden increase in the power of a single frequency, but the degree of increase is different.

[0059] Figure 3a This is a spectrum diagram of the active noise reduction system when howling occurs. The ordinate is the amplitude of the sound signal and the abscissa is the frequency of the sound signal. In the case of howling instability, the power of the frequency where the abnormality occurs will become particularly strong, and even the main peak of the power of the entire frequency band. It can be seen from the figure that the abnormal frequency is 109Hz, and the power of this frequency (300000) is much greater than the power of other frequency bands (the highest is between 40000 and 50000), which is an order of magnitude different. Figure 3b This is a spectrum diagram of the active noise reduction system when a stable single-frequency abnormal sound occurs. The ordinate is the amplitude of the sound signal, and the abscissa is the frequency of the sound signal. The arrow points to the situation where the stable single-frequency abnormal sound is unstable. The power of the abnormal frequency is not like howling. Its power is not particularly strong and is generally lower than the main peak power of the low-frequency band (100-200Hz) wind noise and fan main noise. It can be seen from the figure that the abnormal frequency is around 430Hz, and the power of this frequency is lower than the main peak of the frequency around 100-200Hz. However, since this frequency (around 430Hz) is an isolated peak in its nearby frequency band, it is easy to be heard.

[0060] In step 101, it is necessary to collect sound signals during the operation of the active noise reduction system, and extract sound features for instability detection from the sound signals. It should be noted that in step 101, multiple frames of sound signals can be collected, and instability detection is subsequently performed based on the multiple frames of sound signals; or a single frame of sound signals with a longer duration can be collected, and instability detection is subsequently performed based on the single frame of sound signals.

[0061] In one implementation, the sound signal is collected in real time during the operation of the active noise reduction system, and the sound signal is Fourier transformed to extract the sound features from the Fourier transform results. The Fourier transform can be, but is not limited to, the fast Fourier transform FFT. Considering the system computing power performance, the number of FFT points N does not need to be particularly large. It is sufficient to ensure that the FFT frequency resolution F = fs / N is between 20-50Hz. fs is the system signal sampling rate.

[0062] In one embodiment, the sound signal is Fourier transformed at a target frequency resolution. Frequency resolution The rate is obtained by optimization, and the optimization goal is to accurately detect instability and minimize the computational complexity of instability detection. Small.

[0063] The so-called accurate instability detection means that the instability detection accuracy is greater than the accuracy threshold and / or the false detection rate is less than or equal to the false detection threshold. Among them, the accuracy threshold and the false detection threshold can be set according to actual conditions.

[0064] The frequency resolution determines the amount of calculation in the instability detection process. When the frequency resolution is 1Hz, it means that a Fourier transform is performed every 1s. Assuming that the amount of calculation each time is K1, when the frequency resolution is increased to 4Hz, it means that a Fourier transform is performed every 1 / 4s. At this time, the amount of calculation each time is K2, the amount of data within 1 / 4s, K1=4·K2. The frequency resolution of 4Hz reduces the amount of data by 4 times compared to the frequency resolution of 1Hz, which greatly speeds up the calculation speed of instability detection. As the frequency resolution increases, the amount of data for each Fourier transform decreases.

[0065] In one embodiment, the step of determining the target frequency resolution includes: obtaining abnormal samples, iteratively updating the frequency resolution, and determining the frequency resolution corresponding to accurate identification of abnormal sounds and the minimum calculation amount of instability detection as the target frequency resolution.

[0066] The abnormal sample is a noise signal containing an abnormal sound. The abnormal sample can be collected through experiments, or it can be the result of superimposing a normal sample (a noise signal without an abnormal sound) and an abnormal sound.

[0067] In one embodiment, the optimization algorithm may include, but is not limited to, at least one of the following algorithms: a genetic algorithm, a simulated annealing algorithm, a particle swarm algorithm, a Bayesian optimization algorithm, and a hill climbing algorithm.

[0068] The following example further illustrates the process of determining the target frequency resolution:

[0069] Based on the abnormal samples, the frequency band where the instability occurs is known, such as the stable single-frequency instability at frequencies fa, fb, and fc in the abnormal samples.

[0070] S1. Assuming the initial frequency resolution is 1 Hz, after increasing the frequency resolution to 32 Hz, perform FFT on the abnormal samples and perform instability detection based on FFT.

[0071] S2. If the frequency resolution of 32Hz can accurately identify the existence of stable single-frequency instability in abnormal samples at fa, fb, and fc, it means that increasing the frequency resolution to 32Hz is sufficient to determine the instability. If the frequency resolution of 32Hz cannot accurately identify the instability in all frequencies fa, fb, and fc in the abnormal samples and / or there is a misjudgment (for example, other frequencies are identified as instability), it means that although increasing the frequency resolution to 32Hz reduces the amount of calculation, it is not enough to determine all instability points at this time, and the frequency resolution needs to be lowered.

[0072] S3. If the frequency resolution is sufficient to determine instability, select 32 Hz as the frequency resolution and determine the corresponding amount of calculation. At this time, the amount of calculation has been reduced to 1 / 32 compared to the initial amount of calculation when the frequency resolution is 1 Hz. If the frequency resolution is not sufficient to determine all active noise reduction instability points, continue to repeat step 2 until the frequency resolution that accurately identifies instability and minimizes the amount of calculation for instability detection is found.

[0073] Step 102: Determine the suspected instability frequency of the stable single-frequency abnormal sound in the active noise reduction system according to the sound characteristics.

[0074] According to the above description, a major feature of unstable sound is that the peak of the unstable frequency is generally an isolated / prominent spike, so the purpose of this step is to filter out the frequency with a peak that is more prominent than the surrounding in each frame of the sound signal as an abnormal point. Feature extraction is performed based on the sound features obtained in the previous step. The extracted features can be obtained through spectrum amplitude features, spectrum phase features, spectrum power features, spectrum shape features, etc.

[0075] In one embodiment, in order to determine that the cause of the instability of the active noise reduction is a stable single-frequency abnormal sound rather than a howling, in step 102, the suspected instability frequency of the stable single-frequency abnormal sound in the active noise reduction system is determined according to the sound characteristics. Subsequently, the final instability frequency is determined from the suspected instability frequencies of the stable single-frequency abnormal sound.

[0076] Step 103: determine the suspected unstable frequency that meets the instability condition as the final unstable frequency of the stable single-frequency abnormal sound.

[0077] Among them, the instability condition is determined according to the characteristics of the stable single-frequency abnormal sound.

[0078] In this embodiment, by analyzing the sound features used for instability detection in the sound signal, the suspected instability frequency of the stable single-frequency abnormal sound is first determined, and then the final instability frequency of the stable single-frequency abnormal sound is determined. Through two-step screening, the accuracy of instability detection is improved.

[0079] In one embodiment, in step 103, the sound signal is partitioned according to a preset partitioning strategy; for each partition, when the sound characteristics in the partition meet the instability condition, the suspected instability frequency in the partition is determined as the final instability frequency.

[0080] The preset partitioning strategy is determined according to the frequency band difference of the sound signal, and the instability condition is determined according to the characteristics of the sound signal when the active noise reduction system is unstable.

[0081] In this embodiment, the sound signal is processed in different zones and the instability state is detected in different zones to eliminate mutual interference between the zones, thereby improving the accuracy of instability detection.

[0082] Preferably, the preset partitioning strategy includes partitioning intervals; the partitioning intervals are determined based on the differences in peak values ​​of the sound signal in different frequency bands without instability; the sound signal is partitioned, that is, the frequency of the sound signal is segmented according to the partitioning intervals.

[0083] Taking the active noise reduction system installed in the range hood as an example, in addition to the two characteristics of unstable sound frequencies, which are spikes and continuity, there are also some normal sound frequencies in the low-frequency band (between 100-250Hz) that have the above characteristics. Therefore, in order to distinguish the normal sound in the low-frequency band, the sound signal needs to be partitioned. The partitioning strategy partitions the sound signal mainly based on the differences between frequency bands. First, it is necessary to obtain the range of amplitudes corresponding to different frequencies of the active noise reduction system when it is not unstable. This data can be obtained through user data or laboratory data. Then, the partition interval is determined based on the differences in the above data values, and then the sound signal is partitioned according to the partition interval.

[0084] In one embodiment, the partition interval can be determined in the effective frequency band of active noise reduction / the frequency band where instability is prone to occur. If the effective frequency band of active noise reduction / the frequency band where howling is prone to occur is between 100-700Hz, and the spectrum characteristics (peak values) of every 100Hz are basically the same, then the preset partition intervals can be 100-200Hz, 200-300Hz, 300-400Hz, 400-500Hz, 500-600Hz, 600-700Hz, a total of 6 intervals, that is, to ensure that the amplitude corresponding to each interval is roughly the same. The sound signals of 0-100Hz and 700Hz will not cause instability, and there is no need to perform instability analysis on them. Excluding them can reduce the amount of calculation.

[0085] It should be noted that the intervals of the partition intervals can also be determined by observing the peak and PNPR in the interval. For example, if the peak and / or PNPR in the 300-700Hz interval are relatively close, that is, the rate of change is less than the rate of change threshold, 300-700Hz can also be merged into one partition for processing.

[0086] In this embodiment, in order to distinguish the instability frequencies in different partitions, the sound signals are partitioned according to the frequencies based on the corresponding peak values ​​at different frequencies in the partitions. In subsequent processing, different processing strategies can be adopted for the sound signals in different frequency ranges, which is more conducive to screening out the instability signals and provides a benchmark for instability judgment.

[0087] In one embodiment, a partition strategy is determined according to an actual application scenario of the active noise reduction system. Specifically, the actual application scenario of the active noise reduction system is determined, and a partition strategy matching the actual application scenario is determined according to the correspondence between the application scenario and the partition strategy, and the sound signal is partitioned using the partition strategy. The actual application scenario can be characterized by, but not limited to, the amplitude of the ambient noise, the frequency of the ambient noise, and the like.

[0088] After the sound signal is partitioned as described above, the suspected instability frequencies in each partition are further judged to determine whether they meet the instability conditions. Taking the active noise reduction system installed in the range hood as an example, the peak value of the low-frequency band (the partition between 100-300Hz) will be significantly higher than the peak value of the high-frequency band (the interval above 300Hz). Therefore, the amplitude corresponding to the suspected instability frequency in the low-frequency band partition and the amplitude corresponding to the suspected instability frequency in the high-frequency band will be very different. Therefore, the suspected instability frequencies in each partition need to be distinguished and compared, and further screened to determine the final instability frequency. If the peak value of the suspected instability frequency is significantly higher than the peak values ​​corresponding to other frequencies in this interval, it can be determined as the final instability frequency.

[0089] In this embodiment, by analyzing the sound features used for instability detection in the sound signal, the sound signal is partitioned and processed, and the instability state is detected by partition to eliminate the mutual interference between the partitions, thereby improving the accuracy of instability detection. In particular, for stable single-frequency abnormal sounds where the power of the peak accounts for a small proportion of the total noise power, the influence of the frequency band with higher power in the ambient noise on instability detection can be eliminated through partitioning, thereby improving the accuracy of instability detection when stable single-frequency abnormal sounds occur in the active noise reduction system.

[0090] In one embodiment, the acoustic characteristics include a peak adjacent power ratio, see Figure 4a and Figure 4b , according to the sound characteristics, determine the suspected unstable frequency of the stable single-frequency abnormal sound in the active noise reduction system, including:

[0091] Step 201 : determine the peak frequency of each frame of the sound signal in the multiple frames of the sound signal respectively; the peak frequency is the frequency corresponding to the maximum value of at least three adjacent peaks in the sound signal.

[0092] Specifically, a major feature of unstable sound is that the peak of the unstable frequency is generally an isolated / prominent spike, so this step is used to filter out the frequencies in each frame of sound signal whose peaks are more prominent than the surrounding ones as abnormal points. For example, if the power of frequency A is the largest compared to the two adjacent (A-1 and A+1) or N (AN / 2 to A+N / 2, N is an even number) frequencies, then frequency A is considered to be a peak frequency. And so on, all peak frequency queues Peak are filtered out.

[0093] It can be understood that multiple frames of sound signals can be regarded as a sound signal with a longer time frame. If step 101 collects a sound signal with a longer time frame, step 201 determines the peak frequency of the sound signal with a longer time frame.

[0094] Step 202: Calculate the peak adjacent power ratio corresponding to the peak frequency, and The peak frequency corresponding to the peak adjacent power ratio of the surrounding area being greater than the peak adjacent power ratio threshold is determined as the initial unstable frequency of the stable single-frequency abnormal sound.

[0095] The preset abnormal sound range is determined according to the peak adjacent power ratio of the frequency points of the spectrum of the analog signal. The analog signal is a sound signal collected when the active noise reduction system simulates a stable single-frequency abnormal sound.

[0096] Specifically, for all frequencies in the Peak, calculate the peak neighbor power ratio PNPR (the ratio of the power of the target frequency to the average power of the adjacent frequencies, which can be expressed logarithmically). PNPR is used to characterize the sharpness of the frequency peak. If the peak is sharper and the surrounding frequencies are flatter, the PNPR value is larger. That is, if frequency B is a candidate frequency in the Peak, the power E of frequency B is extracted. B , and the power average Eavg of the N frequencies (BN / 2 to B+N / 2, N is an even number) adjacent to frequency B, then the PNPR value of frequency B is Bpnpr=10*log10(E B / Eavg). And so on, calculate the PNPR values ​​of all frequencies in the peak frequency queue Peak. According to the preset abnormal sound range, filter out the initial unstable frequency of the stable single-frequency abnormal sound. And put its PNPR value into the queue Peak. PNPR middle.

[0097] PNPR has little to do with the absolute power of the peak. Figure 3b Taking the 430Hz peak as an example, it is a sharp isolated peak, and its amplitude is much larger than that of its adjacent frequency bands. By presetting the abnormal sound range and combining PNPR, the initial instability frequency of the stable single-frequency abnormal sound can be accurately identified.

[0098] Step 203: Determine the suspected unstable frequency of the stable single-frequency abnormal sound from the initial unstable frequency that meets the preset condition.

[0099] Among them, the preset conditions are determined according to the characteristics of the stable single-frequency abnormal sound.

[0100] One characteristic of unstable sound is that the sound of unstable frequency generally does not appear and disappear at a very fast speed, but lasts for a period of time. According to the above characteristics, multiple frames of sound signals or a longer frame of sound signal are selected to determine the initial unstable frequency.

[0101] Taking selecting multiple frames of sound signals to identify the initial instability frequency and determine the suspected instability frequency as an example, in one embodiment, the preset condition includes that the number of occurrences of the initial instability frequency in several adjacent frames of sound signals is greater than a first instability threshold.

[0102] Specifically, the initial instability frequency of the N frames (N is an integer) before the current sound signal is obtained. If the current frame is the i-th frame, it is necessary to collect all the initial instability frequency queues from the i-N+1th frame to the i-th frame. Count the frequency queues of N frames, count all the frequencies that appear in the queue, and obtain the number of times each initial instability frequency appears. For example, to obtain the initial instability frequency of the 10 frames of sound signals before the current sound signal, it is necessary to collect all the initial instability frequency queues in the first 10 frames. If frequency A appears 9 times in the initial instability frequency queue of the first 10 frames, which is greater than the first instability threshold (7 times), frequency A is determined as a suspected instability frequency. It should be noted that N is 10 and the first instability threshold is 7 times for illustrative purposes only. The specific values ​​of N and the first instability threshold can be set according to actual conditions.

[0103] In one embodiment, the preset condition includes that the probability of occurrence of the initial instability frequency in a number of adjacent frames of sound signals is greater than the second instability threshold.

[0104] Specifically, the initial instability frequency of the N frames before the current sound signal is obtained. If the current frame is the i-th frame, it is necessary to collect all the initial instability frequency queues from the i-N+1 frame to the i-th frame. Count the frequency queues of N frames, count all the frequencies that appear in the queue, and obtain the probability of each initial instability frequency. For example, to obtain the initial instability frequency of the 10 frames before the current sound signal, it is necessary to collect all the initial instability frequency queues in the first 10 frames. If the probability of frequency A appearing in the initial instability frequency queue of the first 10 frames is 75%, which is greater than the second instability threshold (50%), then frequency A is determined as a suspected instability frequency. It should be noted that the second instability threshold of 50% is only an example, and the specific value of the second instability threshold can be set according to actual conditions.

[0105] In one embodiment, the preset condition includes that the sum of the peak adjacent power ratios of the initial instability frequencies in a plurality of adjacent frames of sound signals is greater than a third instability threshold.

[0106] Specifically, the initial instability frequency of the N frames before the current sound signal is obtained. If the current frame is the i-th frame, it is necessary to collect all the initial instability frequency queues from the i-N+1th frame to the i-th frame. Count the frequency queues of N frames, sum all the PNPR values ​​of the same frequency in the queue, and obtain the total PNPR value of the frequency. In the case of multiple frames, if a certain frequency is an instability point, then this frequency will have a higher PNPR value in each frame, so the PNPR value of each frame is added and judged by the sum. For example, to obtain the initial instability frequency of the 10 frames before the current sound signal, it is necessary to collect all the initial instability frequency queues in the first 10 frames. If the initial instability frequency contains frequency A, calculate the sum of the PNPR values ​​of the first 10 frames corresponding to frequency A. When the sum of the PNPR values ​​of frequency A is greater than the third instability threshold, frequency A is determined as a suspected instability frequency.

[0107] In one embodiment, the preset condition includes that an average value of peak adjacent power ratios of initial instability frequencies in a plurality of adjacent frames of sound signals is greater than a fourth instability threshold.

[0108] Specifically, the initial instability frequency of the N frames before the current sound signal is obtained. If the current frame is the i-th frame, it is necessary to collect all the initial instability frequency queues from the i-N+1th frame to the i-th frame. Count the frequency queues of N frames, calculate the average value of all PNPR values ​​of the same frequency in the queue, and obtain the PNPR average value of the frequency. In the case of multiple frames, if a certain frequency is an instability point, then this frequency will have a higher PNPR value in each frame, so the average value of the PNPR value of each frame is judged by the average value. For example, to obtain the initial instability frequency of the 10 frames before the current sound signal, it is necessary to collect all the initial instability frequency queues in the first 10 frames. If the initial instability frequency contains frequency A, the average value of the PNPR values ​​of the first 10 frames corresponding to frequency A is calculated. When the average value of the PNPR value of frequency A is greater than the fourth instability threshold, frequency A is determined as a suspected instability frequency.

[0109] The above preset conditions can be combined with each other to screen suspected unstable frequencies.

[0110] In the above embodiment, the instability frequency is judged by single-frame and multi-frame sound signals, and the instability frequency is screened multiple times, which lays a foundation for subsequent instability frequency detection and further improves the accuracy of instability frequency detection.

[0111] In one embodiment, for each partition, when the sound feature in the partition meets the instability condition, determining the suspected instability frequency in the partition as the final instability frequency includes:

[0112] Calculate the standard deviation of the peak adjacent power ratio corresponding to the suspected instability frequency in the same partition; when the standard deviation is greater than the standard deviation threshold of the corresponding partition, determine the corresponding suspected instability frequency in the partition as the final instability frequency; the standard deviation threshold is determined based on the sound characteristics when no instability occurs.

[0113] Specifically, the standard deviation of the PNPR corresponding to the suspected instability frequency of the same partition is calculated. The standard deviation can be used to determine the peak fluctuation of the interval. Once any partition exceeds the standard deviation threshold of the corresponding partition, it is considered that the fluctuation of the partition is greater than expected, and instability is considered to have occurred. And the corresponding suspected instability frequency in the partition is determined as the final instability frequency. The setting of the standard deviation threshold will be determined according to the sound characteristics when instability occurs. Taking the range hood as an example, the peak fluctuation in the low-frequency band is relatively high, so the calculated PNPR will also fluctuate more, so the calculation of std (standard deviation) will also be larger. Compared with the high-frequency band, the high-frequency band is more stable and has small fluctuations. The calculated PNPR value also has small fluctuations, so the calculation of std will also be small. Therefore, for the active noise reduction system of the range hood, the standard deviation threshold of the low-frequency band (partition between 100-300Hz) will be significantly higher than the standard deviation threshold of the high-frequency band (interval above 300Hz).

[0114] Frequency band Group 1 Group 2 Group 3 Group 4 Group 5 Group 6 Group 7 Group 8 st threshold 100-200 17.32045 7.491047 17.91207 8.060772 13.42452 23.09753 11.97594 17.57675 25 200-300 15.91188 20.20623 13.47737 12.64061 14.02145 10.45675 14.07649 10.60075 25 300-400 5.474182 7.54878 9.198485 13.50125 6.573411 9.120099 9.136161 9.361063 15 400-500 10.17906 8.993037 5.184132 9.905431 8.069404 6.552608 10.36976 9.828358 15 500-600 10.47169 8.84014 8.522361 11.03411 6.002584 8.113459 8.358088 7.308186 15 600-700 10.15184 13.34889 14.73921 5.985169 7.139955 12.09516 8.692635 13.76853 15

[0115] The above table is the standard deviation data of each partition calculated by the partition under normal operation of the eight sets of noise reduction systems. The first column corresponds to the frequency band of the partition, and the other columns are the standard deviation values ​​in different frequency band partitions corresponding to each set of data. Specifically, first filter out the maximum value of the standard deviation of each frequency band, and round it up as the standard deviation threshold. For example, for 100-200Hz, the maximum value is 23.097. Then you can round up and set the standard deviation threshold to 24, or round it up to a multiple of 5, and the threshold can be set to 25.

[0116]

[0117]

[0118] The figure above shows the different suspected unstable frequencies and corresponding PNPR values ​​of multiple frames of sound signals collected in the unstable state, and the standard deviation values ​​obtained by calculation. According to the above standard deviation values ​​and the corresponding standard deviation thresholds of each partition, the suspected unstable frequencies in the partitions exceeding the standard deviation thresholds are determined as the final unstable frequencies.

[0119] In one embodiment, the variance of the peak adjacent power ratio corresponding to the suspected instability frequency in the same partition is calculated. When the variance is greater than the variance threshold corresponding to the partition, the corresponding suspected instability frequency in the partition is determined as the final instability frequency; the variance threshold is determined according to the sound characteristics when no instability occurs.

[0120] Similarly, the variance method can also be used to determine the fluctuation of the interval. First, the variance of the peak adjacent power ratio of the suspected unstable frequency in the same partition is calculated. When the variance is greater than the variance threshold corresponding to the partition, the corresponding suspected unstable frequency in the partition is determined as the final unstable frequency.

[0121] In this embodiment, the peak fluctuation in the same partition can be used to determine whether instability occurs in each interval, thereby determining the corresponding suspected instability frequency in the partition as the final instability frequency. By screening the instability frequency multiple times, the accuracy of instability frequency detection is greatly improved, especially when the abnormal sound occurs in the low frequency band.

[0122] In one embodiment, extracting sound features from a sound signal includes: performing low-pass filtering on the sound signal, performing time domain and / or frequency domain processing on the sound signal, and then extracting sound features from the processed sound signal.

[0123] Specifically, the collected sound signal can be analyzed to determine the range of the noise frequency band, and the frequency band that does not need to be processed or is not prone to instability can be filtered out through a low-pass filter, and then the sound features of the processed sound signal can be extracted. For example, the main noise frequency band of the range hood is relatively low, and the active noise reduction control frequency band will not be fully controlled, so the sound signal can be low-pass filtered first, and the cutoff frequency can be selected as the upper limit frequency of the active noise reduction control, such as 1KHz.

[0124] In this embodiment, the sound signal is processed after passing through a low-pass filter, which greatly reduces the system computing power and reduces the complexity of the overall calculation.

[0125] In one embodiment, extracting sound features from a sound signal includes: performing bandpass filtering on the sound signal, performing time domain and / or frequency domain processing on the sound signal, and then extracting sound features from the processed sound signal. The specific implementation process is similar to that of a low-pass filter and will not be described in detail here.

[0126] The disclosed embodiments also provide a control method for an active noise reduction system, wherein the active noise reduction system includes a filter; the control method comprises: detecting a final instability frequency according to the instability detection method for the active noise reduction system of any of the above embodiments; determining filtering parameters of the filter according to the final instability frequency, and suppressing the instability signal in the sound signal through the filter.

[0127] For the specific implementation process of determining the filtering parameters of the filter according to the final instability frequency, please refer to the relevant technical description, which will not be repeated here.

[0128] Specifically, since the signal amplitude and signal power at the corresponding instability frequency increase after instability occurs, the instability signal in the sound signal can be suppressed by setting a filter. The parameters of the filter can be automatically adjusted according to the final instability frequency and the corresponding amplitude in the above embodiment.

[0129] In this embodiment, when the active noise reduction system becomes unstable, the filter parameters can be automatically adjusted to suppress the unstable signal in the sound signal, thereby greatly improving the user experience.

[0130] An embodiment of the present disclosure also provides an active noise reduction system, which includes: a speaker, a microphone, a processor, and a filter; the processor is connected to the speaker, the microphone, and the filter, respectively, and when the processor executes a computer program, it implements the instability detection method or the control method of the active noise reduction system provided in any of the above embodiments.

[0131] The disclosed embodiment also provides a range hood, which includes the above active noise reduction system. When the range hood is working, the active noise reduction system is turned on. When the active noise reduction system detects instability, the active noise reduction system uses the above control method of the active noise reduction system to suppress the instability frequency.

[0132] Figure 5 This is a structural schematic diagram of a household appliance shown in an example embodiment of the present disclosure, wherein the household appliance includes a memory, a processor, and a computer program stored in the memory and used to run on the processor, and when the processor executes the computer program, the instability detection method or control method described in any of the above embodiments is implemented. Figure 5 The household appliance 50 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0133] Household electrical appliances may include but are not limited to range hoods, washing machines, dishwashers, etc.

[0134] like Figure 5 As shown, the home appliance 50 may be in the form of a general computing device, for example, it may be a server device. The components of the home appliance 50 may include but are not limited to: at least one processor 51, at least one memory 52, and a bus 53 connecting different system components (including the memory 52 and the processor 51).

[0135] The bus 53 includes a data bus, an address bus, and a control bus.

[0136] The memory 52 may include a volatile memory, such as a random access memory (RAM) 521 and / or a cache memory 522 , and may further include a read-only memory (ROM) 523 .

[0137] The memory 52 may also include a program tool 525 (or utility) having a set (at least one) of program modules 524, such program modules 524 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0138] The processor 51 executes various functional applications and data processing by running the computer program stored in the memory 52, such as the instability detection method or control method provided in any of the above embodiments.

[0139] The home appliance 50 may also communicate with one or more external devices 54 (e.g., keyboards, pointing devices, etc.). Such communication may be performed via an input / output (I / O) interface 55. In addition, the home appliance 50 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 56. As shown, the network adapter 56 communicates with other modules of the home appliance 50 via a bus 53. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the home appliance 50, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems.

[0140] It should be noted that although several units / modules or sub-units / modules of the household appliance are mentioned in the above detailed description, such division is merely exemplary and not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more units / modules described above may be embodied in one unit / module. Conversely, the features and functions of one unit / module described above may be further divided to be embodied by multiple units / modules.

[0141] The embodiments of the present disclosure further provide a computer-readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the instability detection method or control method provided by any of the above embodiments is implemented.

[0142] The readable storage medium may include but is not limited to: a portable disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical storage device, a magnetic storage device or any suitable combination of the above.

[0143] The embodiments of the present disclosure further provide a computer program product, including a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above-mentioned instability detection methods or control methods.

[0144] Among them, the program code for executing the computer program product of the present disclosure can be written in any combination of one or more programming languages, and the program code can be executed completely on the user device, partially on the user device, as an independent software package, partially on the user device and partially on a remote device, or completely on the remote device.

[0145] Although the specific embodiments of the present disclosure are described above, those skilled in the art should understand that this is only an example, and the protection scope of the present disclosure is defined by the appended claims. Those skilled in the art may make various changes or modifications to these embodiments without departing from the principles and essence of the present disclosure, but these changes and modifications all fall within the protection scope of the present disclosure.

Claims

1. A method for detecting instability of an active noise reduction system, characterized in that: The instability detection method comprises: Acquiring a sound signal during the operation of the active noise reduction system, and extracting a sound feature for instability detection from the sound signal; the sound signal includes a noise signal of a noise source and / or a control signal of a speaker of the active noise reduction system; Determining a suspected instability frequency of a stable single-frequency abnormal sound in the active noise reduction system according to the sound characteristics; The suspected unstable frequency that meets the instability conditions is determined as the final unstable frequency of the stable single-frequency abnormal sound.

2. The instability detection method according to claim 1, characterized in that: The acoustic characteristics include a peak-to-adjacent power ratio; Determining the suspected instability frequency of the stable single-frequency abnormal sound in the active noise reduction system according to the sound characteristics includes: Determine the peak frequency of each frame of the sound signal in the multiple frames of sound signals respectively; the peak frequency is the frequency corresponding to the maximum value of at least three adjacent peaks in the sound signal; Calculating a peak adjacent power ratio corresponding to the peak frequency, and determining a peak frequency corresponding to the peak adjacent power ratio falling within a preset abnormal sound range as an initial instability frequency; The initial instability frequency that meets the preset conditions is determined as the suspected instability frequency of the stable single-frequency abnormal sound.

3. The instability detection method according to claim 2, characterized in that: The preset abnormal sound range is determined according to the peak adjacent power ratio of the frequency points of the spectrum of the analog signal; wherein, the analog signal is a sound signal collected when the active noise reduction system simulates a stable single-frequency abnormal sound.

4. The instability detection method according to claim 2, characterized in that: The preset condition includes at least one of the following: The number of occurrences of the initial instability frequency in a plurality of adjacent frames of sound signals is greater than a first instability threshold; The probability of occurrence of the initial instability frequency in a plurality of adjacent frames of sound signals is greater than a second instability threshold; The sum of the peak adjacent power ratios of the initial instability frequency in a plurality of adjacent frames of sound signals is greater than a third instability threshold; An average value of peak adjacent power ratios of the initial instability frequency in a plurality of adjacent frames of sound signals is greater than a fourth instability threshold.

5. The instability detection method according to claim 1, characterized in that: The suspected unstable frequency that meets the instability conditions is determined as the final unstable frequency, including: Partitioning the sound signal according to a preset partitioning strategy; the preset partitioning strategy is determined according to the frequency band difference of the sound signal; For each partition, when the sound feature in the partition meets the instability condition, the suspected instability frequency in the partition is determined as the final instability frequency of the stable single-frequency abnormal sound.

6. The instability detection method according to claim 5, characterized in that: The preset partition strategy includes partition intervals; The partition interval is determined according to the difference in peak values ​​of the sound signal in different frequency bands in a non-instability state; partitioning the sound signal includes: partitioning the sound signal according to the partition interval.

7. The instability detection method according to claim 5, characterized in that: For each partition, when the sound feature in the partition meets the instability condition, determining the suspected instability frequency in the partition as the final instability frequency includes: Calculate the standard deviation of the peak adjacent power ratio corresponding to the suspected unstable frequency in the same partition; When the standard deviation is greater than the standard deviation threshold of the corresponding partition, determining the corresponding suspected instability frequency in the partition as the final instability frequency; The standard deviation threshold is determined according to the sound characteristics when no instability occurs; and / or; Calculate the variance of the peak adjacent power ratio corresponding to the suspected unstable frequency in the same partition; When the variance is greater than the variance threshold corresponding to the partition, determining the suspected instability frequency corresponding to the partition as the final instability frequency; The variance threshold is determined according to the sound characteristics when no instability occurs.

8. The instability detection method according to claim 1, characterized in that: Extracting sound features from the sound signal includes: performing time domain and / or frequency domain processing on the sound signal after low-pass filtering the sound signal, and then extracting sound features from the processed sound signal.

9. A control method for an active noise reduction system, characterized in that: The active noise reduction system includes a filter; the control method includes: According to the instability detection method of the active noise reduction system according to any one of claims 1 to 8, the final instability frequency is detected; The filter parameters of the filter are determined according to the final instability frequency, and the instability signal in the sound signal is suppressed by the filter.

10. An active noise reduction system, characterized in that: The active noise reduction system includes: a speaker, a microphone, a processor, and a filter; the processor is connected to the speaker, the microphone, and the filter respectively, and when the processor executes a computer program, it implements the instability detection method according to any one of claims 1 to 8 or the control method of the active noise reduction system according to claim 9.

11. A household appliance 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, the instability detection method according to any one of claims 1 to 8 or the control method of the active noise reduction system according to claim 9 is implemented.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the instability detection method according to any one of claims 1 to 8 or the control method of the active noise reduction system according to claim 9 is implemented.

13. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the instability detection method according to any one of claims 1 to 8 or the control method of the active noise reduction system according to claim 9 is implemented.

Images