Resonance avoidance control method for variable frequency compressor based on acoustic sensing

Abstract

The application provides a variable-frequency compressor resonance avoidance control method based on acoustic sensing, which comprises the following steps: collecting an operating acoustic signal through an acoustic sensor of a compressor cavity; applying a small disturbance of ±1 Hz above and below the current compressor operating frequency to calculate a disturbance frequency spectrum, combining the current compressor operating frequency to generate an energy change amount and a resonance sensitivity index, and generating a frequency modulation direction suggestion; determining an initial frequency modulation amplitude based on the resonance sensitivity index to generate a frequency modulation suggestion amplitude, a frequency modulation target frequency, and a linkage control flag; executing the frequency modulation target frequency and monitoring the resonance sensitivity index in real time, and triggering a reverse fine adjustment if a new resonance is detected to generate a new frequency modulation amplitude. The application replaces the traditional amplitude sudden increase method with a disturbance-driven frequency spectrum identification, and replaces the fixed frequency modulation rule with a strategy compression and real-time feedback, so that the control system has higher identification accuracy, better resonance avoidance strategy, and stronger robustness under complex working conditions.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of acoustic sensing, and particularly relates to a variable frequency compressor resonance avoidance control method based on acoustic sensing. BACKGROUND

[0002] Variable frequency compressors are widely used in heat pump, air conditioning and refrigeration scenes, which realize efficient operation and load matching by adjusting the motor operating frequency. However, at certain frequencies and their multiple frequency points, the internal structure of the compressor is prone to excite resonance phenomenon, causing problems such as increased noise, severe vibration and even mechanical fatigue. The traditional control strategy generally adopts a static method of preset resonance frequency avoidance, that is, the controller skips a specific frequency band to avoid resonance risk. However, due to the dynamic drift of actual resonance points affected by factors such as equipment aging, load changes and refrigerant state, the static rule is easy to fail in long-term operation. At the same time, some existing technologies use acoustic sensors combined with FFT to extract frequency spectrum energy for resonance identification, but mainly rely on threshold judgment of amplitude mutation, which is easily disturbed by background noise fluctuations and other disturbance sources, leading to misjudgment, missed judgment or false triggering of frequency modulation, and cannot judge whether it is in a resonance critical state, nor can it deduce the frequency modulation direction based on disturbance response. Even if the resonance state is identified, the controller often uses a fixed proportion or one-way frequency modulation strategy, which lacks frequency modulation rate adjustment and working condition feedback mechanism, and is difficult to cope with changes in multiple device models, complex installation structures and dynamic use scenarios. The frequency modulation action may deviate from the optimal working zone or repeatedly jump frequency, affecting the stability, comfort and energy efficiency of the system. SUMMARY

[0003] The purpose of the present application is to propose a variable frequency compressor resonance avoidance control method based on acoustic sensing, which adopts a phased execution and dynamic feedback mechanism to construct a complete closed-loop control logic.

[0004] In order to achieve the above purpose, the present application provides a variable frequency compressor resonance avoidance control method based on acoustic sensing, which comprises:

[0005] S1: Acquire running acoustic signals through acoustic sensors in the compressor cavity, generate frequency spectrum vectors through framing and Fourier transform, extract 2 times frequency and 3 times frequency characteristic frequency bands of the current frequency spectrum vector, and synchronously acquire the current compressor operating frequency;

[0006] S2: Apply ±1Hz small perturbation to the 2 times frequency and 3 times frequency characteristic frequency bands of the current frequency spectrum vector above and below the current compressor operating frequency to calculate the perturbation frequency spectrum, generate energy change amount in combination with the current compressor operating frequency, then the controller statistically analyzes the high frequency band sound spectrum fluctuation degree of the current frame of the system to generate an anti-interference term, obtains a resonance sensitivity index in combination with the energy change amount, and generates a frequency modulation direction suggestion; wherein the frequency modulation direction includes frequency increase and frequency decrease;

[0007] S3: determining an initial frequency modulation amplitude based on the resonance sensitivity index, suppressing the amplitude in combination with the deviation of the current compressor operating frequency from the rated frequency, generating a frequency modulation suggestion amplitude, a frequency modulation target frequency, and a linkage control flag;

[0008] S4: performing the frequency modulation target frequency in stages with the minimum step size, and monitoring the resonance sensitivity index in real time, triggering a reverse fine adjustment if a new resonance is detected, and generating a new frequency modulation amplitude;

[0009] Wherein, when the linkage control flag = 1, the frequency modulation action is performed according to the frequency modulation suggestion amplitude, the fan PWM signal and the expansion valve step opening are adjusted according to the direction and size of the frequency modulation suggestion amplitude, and after the frequency modulation is executed, the updated operating frequency is generated, and the state flag is marked for the execution result:

[0010] If the updated operating frequency reaches the frequency modulation target frequency, and the resonance sensitivity index is lower than the preset threshold value in two consecutive periods, it is considered that the frequency modulation anti-vibration is successful, and the frequency modulation state flag = 1 is set.

[0011] Further, the acoustic sensor is a capacitive electret microphone module, which is selected as a standard model, the frequency response range of which is 20Hz to 10kHz, and the sensitivity thereof is about -42dB; the capacitive electret microphone module is connected to the ADC channel of the main controller through an analog front end, and the ADC sampling rate is fixedly set as 8000Hz.

[0012] Further, the S1 specifically includes:

[0013] Collecting original acoustic data, denoted as ;

[0014] In each frame processing process, the controller first multiplies the current frame data by a standard Hamming window function to reduce the influence of discontinuity at the edge of the window function on the spectral analysis;

[0015] After windowing, the controller performs a fast Fourier transform operation on the frame data to convert the time domain signal into a frequency domain representation;

[0016] The frequency spectrum vector obtained after conversion is , which represents the acoustic energy amplitude of each frequency component inside the compressor cavity at the current frame time;

[0017] The maximum value of the frequency spectrum vector is normalized to ensure that the system can compare the energy distribution between different frequency bands under different hardware configurations;

[0018] Synchronously acquire the current compressor operating frequency , based on the current compressor operating frequency Two key frequency bands are located in the spectrum vector : the first band is centered at 2 times the current compressor operating frequency, and is extended by 5 Hz upward and downward; the second band is centered at 3 times the current compressor operating frequency, and is also extended by ±5 Hz; the first band and the second band together form a variable , which represents the characteristic frequency region in the sound spectrum that is most relevant to the frequency multiplication resonance.

[0019] Further, the S2 specifically includes:

[0020] Based on the current compressor operating frequency , two perturbation frequency points are generated: an upper perturbation and a lower perturbation ; during the perturbation process, new sound spectrum data is collected in real time to generate perturbation spectrum vectors, including an upper perturbation spectrum vector and a lower perturbation spectrum vector ;

[0021] ±5 Hz range centered at and , two key frequency multiplication feature segments are extracted, denoted as and , which are combined into a total frequency band , and within the total frequency band , the energy difference between the perturbation spectrum vector and the current compressor operating frequency is compared to construct the energy change amount after positive and negative perturbations;

[0022] The mean square fluctuation in the 0-500 Hz high frequency band of the spectrum is counted to construct an anti-interference term;

[0023] According to the anti-interference term and the energy change amount, a resonance sensitivity index is calculated, which is used to measure the asymmetric amplitude of the perturbation response. The larger the value, the more sensitive the current frequency is to the perturbation, and the closer it is to the resonance critical region.

[0024] Further, the frequency modulation direction suggestion is generated as follows:

[0025] An empirical threshold is set;

[0026] When the resonance sensitivity index is greater than the empirical threshold , it is determined that the current frequency has a resonance risk, and the frequency modulation avoidance process is entered, including:

[0027] The frequency modulation direction is determined by the comparison relationship between the upper and lower perturbation response amplitudes of the energy change amount: if the upper perturbation response amplitude is greater than the lower perturbation response amplitude, the frequency modulation direction is less than the disturbance response amplitude , indicating that the response amplitude in the frequency-up direction is smaller, i.e. it is more likely to move out of the resonance region after frequency-up, the system sets the frequency-up direction to +1; otherwise, it is set to -1.

[0028] Further, the S3 specifically comprises:

[0029] accepting the resonance sensitivity index , multiplying it by a preset proportion coefficient to obtain an initial frequency-up amplitude;

[0030] calculating the relative deviation between the current compressor operating frequency and the rated frequency ; if the relative deviation exceeds a preset tolerance threshold, the initial frequency-up amplitude is suppressed to slow down the frequency-up speed;

[0031] generating a frequency-up suggestion amplitude according to the initial frequency-up amplitude and the relative deviation, combining the current compressor operating frequency to obtain a frequency-up target frequency.

[0032] Further, the S3 further comprises:

[0033] triggering a linkage control flag according to the resonance sensitivity index, which is used to indicate whether linkage control of the fan speed or electronic expansion valve is needed, specifically:

[0034] when the resonance sensitivity index is greater than a preset threshold, the system will output , linkage control flag = 1, indicating that the expansion valve / fan needs to be adjusted simultaneously.

[0035] Further, the S4 specifically comprises:

[0036] obtaining a frequency-up suggestion amplitude to perform equal-step frequency-up with a preset minimum frequency step , the direction of which is determined by the current frequency-up direction , and the step is set by the PWM controller;

[0037] after each frequency update, returning to S2 to reconstruct the latest frequency spectrum and calculate a new resonance sensitivity index ; if the new resonance sensitivity index is detected to exceed a set re-resonance judgment threshold after two consecutive updates, reverse fine tuning is entered.

[0038] Further, the reverse fine tuning specifically comprises:

[0039] based on the new resonance sensitivity index , performing reverse frequency-up by changing the current frequency-up direction A proportionally scaled reverse fine adjustment is applied to the new resonance boundary to obtain a reverse fine adjustment, which is used to exit the new resonance boundary; wherein the reverse fine adjustment is triggered immediately when an abnormal resonance response is detected, and the new frequency is written into the controller register, and the process returns to S1.

[0040] Further, the controller continuously collects and processes the original acoustic signal at a sampling rate of 8000 Hz, and performs sliding window frame processing on the original acoustic signal with a frame length of 128 ms, each frame containing 1024 sampling points, and the frame overlap is 768 points, i.e., the frame shift is 256 points, and the corresponding time interval is 32 ms, which ensures continuous coverage of spectral information on the time axis.

[0041] The beneficial technical effects of the present application are at least as follows:

[0042] The present application provides a variable frequency compressor resonance avoidance control method based on acoustic sensing and disturbance response identification, which adopts a phased execution and dynamic feedback mechanism to construct a complete closed-loop control logic. First, the acoustic sensor installed in the compressor cavity collects running acoustic wave data and extracts spectral features, while synchronously reading the current compressor frequency, and real-time constructing the feature frequency band related to 2 times and 3 times frequency. Subsequently, the system applies a small disturbance above and below the current frequency, collects the acoustic spectrum change data before and after the disturbance, constructs the disturbance response asymmetry index, which is used to judge whether the current is in the resonance critical region, and outputs the clear frequency modulation direction. The frequency modulation strategy generation introduces a disturbance intensity and direction linkage mechanism, and combines the relative deviation between the current frequency and the rated frequency to compress the frequency modulation amplitude, preventing the frequency from jumping into the non-design working area. The frequency modulation execution stage adopts a step frequency modulation method based on the main loop period, and continuously calculates the real-time resonance sensitivity index during the frequency modulation process. If resonance is identified again, a fine adjustment action is triggered to quickly exit the risk frequency band, realizing a "identification-adjustment-verification-correction" closed-loop process. This method replaces the traditional amplitude surge method with disturbance-driven spectral identification, and replaces the fixed frequency modulation rule with strategy compression and real-time feedback, so that the control system has higher identification accuracy, better resonance avoidance strategy and stronger robustness under complex working conditions, and is suitable for variable frequency compressors of multiple types and structures. BRIEF DESCRIPTION OF DRAWINGS

[0043] The present application is further illustrated by the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present application. For ordinary skilled in the art, other drawings can be obtained without creative labor according to the following drawings.

[0044] Figure 1 The present application is further illustrated by the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present application. For ordinary skilled in the art, other drawings can be obtained without creative labor according to the following drawings. DETAILED DESCRIPTION

[0045] Embodiments of the present application are described below in detail with reference to the accompanying drawings, wherein the same or similar components or components having the same or similar functions are denoted by the same or similar reference numerals throughout. The embodiments described below by reference to the accompanying drawings are exemplary and are intended only for explanation of the present application, and cannot be understood as limiting the present application.

[0046] As shown in Figure 1 The method for resonance avoidance control of a variable frequency compressor based on acoustic sensing provided by the embodiments of the present application comprises the following steps S1-S4:

[0047] S1: Collecting an operating acoustic signal through an acoustic sensor of a compressor cavity, generating a frequency spectrum vector through frame division and Fourier transform, extracting a 2nd harmonic frequency and a 3rd harmonic frequency feature band of the current frequency spectrum vector, and synchronously obtaining a current compressor operating frequency.

[0048] Specifically, the target of this step is to extract high-quality acoustic frequency spectrum data that can be used for resonance judgment from a variable frequency compressor in operation. The frequency spectrum data not only accurately reflects the acoustic state of the current compressor cavity, but also can extract the amplitude variation of the frequency band related to the 2nd harmonic frequency and the 3rd harmonic frequency according to the current operating frequency of the compressor.

[0049] Further, the collection of the acoustic signal is completed through a capacitive electret microphone module installed outside the compressor cavity, which is selected as a standard model (such as Knowles EK-23132-P07), and the frequency response range thereof is 20 Hz to 10 kHz, and the sensitivity thereof is about -42 dB. The microphone is connected to the ADC channel of the main controller via an analog front end (including a preamplifier and an anti-aliasing filter), and the ADC sampling rate is fixedly set to 8000 Hz to ensure the sampling accuracy of the 0-500 Hz low frequency band. In order to eliminate the additional effects of the shell resonance and mechanical noise, the microphone is installed on the side of the compressor shell through a magnetic base, avoiding the air outlet and the mechanical connection part.

[0050] The time series signal collected is denoted as The signal is the original acoustic data collected by the capacitive microphone installed on the surface of the cavity during the operation of the compressor, and the controller continuously collects and processes it at a sampling rate of 8000 Hz. In order to facilitate subsequent frequency spectrum analysis, the controller performs sliding window frame division processing on the one-dimensional time series signal with 128 milliseconds as one frame, each frame contains 1024 sampling points, the frame overlap is 768 points, that is, the frame shift is 256 points, and the corresponding time interval is 32 milliseconds, which ensures the continuous coverage of the frequency spectrum information on the time axis.

[0051] During each frame processing, the controller first multiplies the current frame data by a standard Hamming window function to reduce the influence of discontinuity at the window edges on the spectral analysis. After the windowing is completed, the controller performs a Fast Fourier Transform (FFT) operation on the frame data to convert the time-domain signal into a frequency-domain representation. The resulting spectral vector is , which represents the acoustic energy amplitudes of each frequency component inside the compressor cavity at the current frame . Since the resonance characteristics of interest in this patent are concentrated in the low frequency range, especially the 2nd and 3rd harmonics of the operating frequency, only the spectral data within the 0-500 Hz range is retained, and the rest of the high frequency part is not processed.

[0052] To avoid the influence of different installation environments, gain settings, or microphone sensitivity differences on the spectral amplitude, the vector is uniformly normalized to the maximum value. Normalization ensures that the system can still compare the energy distribution between different frequency bands under different hardware configurations. According to the system configuration parameters, a combination of 1024-point FFT and 8000 Hz sampling rate is used, resulting in a frequency resolution of approximately 7.8 Hz. Therefore, within the 0-500 Hz range, a vector composed of 64 discrete frequency points can be obtained. For example, the point with a frequency index of 6 corresponds to approximately 47 Hz, the point with an index of 12 corresponds to 93 Hz, and the point with an index of 18 corresponds to 140 Hz, which are the most critical feature points in the resonance judgment process.

[0053] To support the subsequent identification of the compressor harmonic resonance characteristics, this step needs to simultaneously extract the sound spectrum and obtain the current compressor operating frequency . This parameter is obtained by communicating with the compressor variable frequency drive module through the interface in the controller mainboard. Specifically, the frequency setting register value corresponding to the PWM control signal is read. The data reading period is set to 100 milliseconds to ensure sufficient time resolution and response synchronization during system operation. Among them, is a floating-point scalar with a unit of Hz, representing the actual operating frequency of the compressor at the current time. This value typically varies dynamically between 20 Hz and 100 Hz, changing with load demand and temperature control algorithm adjustments.

[0054] According to the physical principle of resonance identification, the structural resonance of the compressor tends to occur at integer multiples of its operating frequency, especially the 2nd and 3rd harmonics. Therefore, to provide more accurate and targeted spectral input for the subsequent steps, this step immediately locates two key frequency bands in the spectral vector based on the value of : the first frequency band is centered on the 2nd harmonic of the current frequency, with a range of ±5 Hz; the second frequency band is centered on the 3rd harmonic, also with a range of ±5 Hz. These two frequency bands together form the variable , is used to represent the characteristic frequency region in the acoustic spectrum that is most relevant to octave resonance.

[0055] At the implementation level, the controller will Extract all amplitude data points corresponding to the two frequency intervals mentioned above to form variables. This information is used for further resonance assessment or disturbance response analysis. For example, at a certain moment, if... If the value is 48Hz, then the second harmonic is 96Hz, and the third harmonic is 144Hz. Therefore, the system will... Extract the amplitude values ​​of all frequency points between 91Hz and 101Hz and between 139Hz and 149Hz to form the corresponding frame. This method avoids the inefficiency of blindly searching for resonance signs across the entire spectrum, and can accurately focus on the region where resonance may occur. This not only improves the real-time performance of the algorithm, but also more closely reflects the resonance mechanism of the compressor's physical structure.

[0056] The two variables output in this step will be directly used as input for the perturbation analysis in the next step:

[0057] : The complete 0–500Hz amplitude normalized spectrum vector, with a dimension of 64;

[0058] The current operating frequency of the compressor is provided by the variable frequency drive control logic; it is a scalar value.

[0059] S2: The 2nd and 3rd harmonic characteristic frequency bands of the current spectrum vector are subjected to a small ±1Hz disturbance above and below the current compressor operating frequency to calculate the disturbance spectrum. Combined with the current compressor operating frequency, an energy change is generated. Subsequently, the controller statistically analyzes the high-frequency spectrum fluctuation of the current frame of the system to generate an anti-interference term. Combined with the energy change, a resonance sensitivity index is obtained, and a frequency modulation direction suggestion is generated; wherein the frequency modulation direction includes frequency boosting and frequency downsampling.

[0060] Specifically, this step involves processing the spectrum vector output in step one. and current compressor operating frequency The combined use of [the two methods] allows the system to perform a "structural resonance sensitivity test" based on small disturbance excitation to dynamically detect whether the current operating frequency is approaching the octave resonance region, and determine whether frequency avoidance control needs to be activated, and in which direction the frequency should be adjusted. This method is particularly suitable for situations where the compressor's octave resonance frequency drifts with load, installation status, and refrigerant fluctuations, and has higher real-time performance and adaptability than traditional threshold detection.

[0061] Furthermore, the system operates at the current compressor frequency. On the basis of the two disturbance frequency points: the upper disturbance and the lower disturbance are automatically generated. Each disturbance lasts 200 ms, which is achieved by the controller modifying the drive frequency through the frequency converter. New acoustic spectrum data is collected in real time during the disturbance process, denoted as and . Subsequently, the system extracts two key frequency multiplication feature segments: ±5 Hz ranges centered at and , denoted as and , which is a key step in modeling the resonance characteristic structure in this invention, specifically matching the compressor frequency multiplication resonance mechanism. The above two frequency bands are combined into the total frequency band , and within this interval, the average energy difference between the disturbance spectrum and the current spectrum is compared to construct the energy change quantity after positive and negative disturbance:

[0062] ;

[0063] In this formula, represents the frequency spectrum vector after the upper or lower disturbance, is the original frequency spectrum, is the number of frequency points in the frequency band (typically 14 to 16, depending on the spectral resolution). This energy change quantity measures the difference in acoustic response in the frequency multiplication related region before and after the disturbance, reflecting the system's sensitivity to frequency disturbance, which is the first criterion for identifying whether it is at the resonance boundary in this step.

[0064] However, due to the presence of a large amount of non-structural noise in the compressor system, such as sudden changes in refrigerant flow and expansion valve switching actions, these will cause non-typical spectral amplitude changes in the high frequency band, which may mislead the system's judgment. Therefore, an anti-interference term is introduced as a regular modification of the resonance recognition criterion, which is constructed by statistically calculating the mean square fluctuation in the 200-500 Hz frequency band (200-500 Hz), specifically:

[0065] ;

[0066] where is the average spectral amplitude in the 200-500 Hz frequency band, is the number of frequency points in this frequency band, is the adjustment term weight (usually set to 0.5 to 0.8), which can be statically configured in the controller. The physical meaning of the anti-interference term is a penalty for non-structural spectral fluctuations in the disturbance background, to prevent mistaking mechanical noise or fluid disturbance as structural resonance.

[0067] After the extraction of the sound spectrum changes before and after the frequency disturbance, the system further constructs a resonance sensitivity index for determining whether it is in the critical region of structural resonance . The index is constructed based on the difference in response amplitude of the two disturbances, and an anti-interference term is introduced to exclude the interference of non-structural noise. Specifically, the system first obtains the frequency spectrum change by forward disturbance (i.e., increasing the frequency by 1 Hz) and the change by reverse disturbance (i.e., reducing the frequency by 1 Hz), which respectively reflect the difference in response strength of the system in the 2 times frequency and 3 times frequency characteristic frequency bands. Subsequently, the controller performs statistics on the sound spectrum fluctuation degree of the current frame of the system in the high frequency band (e.g., 200 Hz to 500 Hz) to generate an anti-interference term for correcting the judgment result, which can be regarded as a quantitative estimate of the non-resonant high-frequency activity inside the system. Based on the three, the system calculates to measure the asymmetric amplitude of the disturbance response, and the larger the value, the more sensitive the current frequency is to the disturbance, and the closer it is to the resonance critical region. Its calculation formula is:

[0068] ;

[0069] wherein and represent the response amplitude changes of the forward disturbance and the reverse disturbance to the frequency spectrum respectively, and γ(t) is the anti-interference term. The larger the value, the more sensitive the current frequency is to the disturbance, and the closer it is to the resonance critical region. Through this index, the system can accurately determine whether the compressor is close to the resonance frequency band and provide a basis for subsequent frequency modulation control.

[0070] To ensure that the judgment standard has universality and controllability, the system sets an empirical threshold , for example, with a value of 0.12. When is greater than the threshold, it means that the current frequency has obvious resonance risk, and it is recommended to enter the frequency modulation avoidance process. The frequency modulation direction is determined by the comparison relationship of the two disturbance response amplitudes: if is less than , it means that the response amplitude in the frequency increasing direction is smaller, i.e., it is more likely to move away from the resonance region after frequency increasing, and the system sets the frequency modulation direction to +1; otherwise, it is set to -1. Taking an actual running case as an example, when the current frequency is 47 Hz, the system detects that is 0.09, is 0.27, and the anti-interference term is 0.04, and the calculation gives is 0.14, which is higher than the threshold 0.12, the system confirms that it is in the resonance boundary area, and since the influence of frequency increase is small, it is recommended to increase the frequency, that is, The resonance judgment and frequency modulation direction identification process are all completed locally by the controller, and the typical execution period does not exceed 500 milliseconds, which can be embedded in an embedded MCU or DSP system with a main frequency of 100 MHz or more.

[0071] This step outputs two core variables:

[0072] : resonance sensitivity index, indicating the response strength of the current frequency to small perturbations;

[0073] : frequency modulation direction, indicates frequency increase, indicates frequency decrease.

[0074] S3: Determine the initial frequency modulation amplitude based on the resonance sensitivity index, and suppress the amplitude combined with the deviation of the current compressor operating frequency from the rated frequency to generate the frequency modulation suggestion amplitude, frequency modulation target frequency and linkage control flag.

[0075] Specifically, combined with the typical response characteristics of the air conditioner compressor during operation, this step constructs a dynamic amplitude calculation mechanism based on the perturbation identification result when calculating the frequency modulation amplitude. Specifically, the system first multiplies the resonance sensitivity index identified in the last stage by a preset proportional coefficient to obtain an initial frequency modulation suggestion amplitude. This coefficient is usually between 4 and 6, and the system can be configured according to the device model. The initial suggestion amplitude and the frequency modulation direction jointly determine whether to increase or decrease the frequency. This calculation structure can quickly adjust the frequency modulation strength according to the change trend of , realize the response enhancement of high-sensitive frequency points, and the micro-avoidance of boundary frequency points. In order to ensure that the frequency modulation behavior does not cause the compressor to deviate from the design operating condition area, the system further reads the current compressor operating frequency and compares it with the rated frequency of the device to calculate the relative deviation. If the deviation exceeds the set threshold (such as 10%), the above frequency modulation suggestion amplitude is scaled and suppressed, and only a part of the original suggestion amplitude is executed to slow down the frequency dispersion trend. This setting is to prevent abnormal refrigerant distribution, current fluctuation or energy efficiency decline caused by frequency modulation in the high or low frequency band. The final output frequency modulation amplitude is the result of S1 and S2 calculation and suppression, which is called by the controller execution module in the next step.

[0076] Further, to ensure that the frequency modulation behavior is always within the physically allowed range, the system introduces a dynamic suppression mechanism to limit the rate of change of the frequency adjustment amplitude under certain conditions when calculating the recommended frequency modulation amplitude. The core idea of this mechanism is: if the current compressor operating frequency is very close to or significantly deviates from the rated frequency set by the device , the frequency modulation amplitude should be reduced to avoid entering the non-design operating range and prevent control stability problems caused by compressor efficiency decline, sudden current load change or abnormal refrigerant flow rate.

[0077] In specific implementation, the system first calculates the relative deviation between the current compressor operating frequency and the rated frequency , i.e. the difference between the two relative to the rated frequency. If the relative deviation exceeds the pre-set tolerance threshold , such as in the range of 10% to 15%, the system will suppress the initial frequency modulation amplitude calculated previously and only keep a portion of it (such as 60%) to slow down the frequency modulation speed. This proportional suppression is achieved by setting a regulation factor , which decreases with the increase of the deviation, thereby automatically reducing the intensity of the frequency modulation action when the frequency is close to the boundary; if the frequency is near the rated range, the system does not scale the frequency modulation amplitude and executes the original value.

[0078] The calculation formula of the regulation factor is as follows:

[0079] ;

[0080] If exceeds the threshold , the regulation factor is set to 0.6, indicating strong suppression; if is less than or equal to the threshold , the regulation factor is set to 1, indicating no suppression. The final frequency modulation amplitude of the system can be obtained by the following formula:

[0081]

[0082] wherein is the control of the frequency modulation amplitude, is the relative deviation between the current frequency and the rated frequency, is the initial frequency modulation amplitude calculated according to the resonance sensitivity and other factors, is the final frequency modulation amplitude.

[0083] Finally, the system will adjust the initial frequency modulation recommendation multiplied by the compression factor to get the final frequency modulation actually to be executed . The final frequency modulation will be added to the current compressor operating frequency to get the recommended new target frequency for the controller to set the new compressor operating frequency value in the next execution step. Through such a mechanism, the frequency modulation suggestion can not only reflect the avoidance intention based on the disturbance intensity and direction in the previous step, but also automatically consider whether the current frequency is close to the physical boundary, thereby maintaining the actual effectiveness of the vibration avoidance action and the stability of the equipment operation.

[0084] To further improve the smoothness of strategy execution and the overall comfort of the system, this step also introduces an optional control flag variable triggered by to indicate whether the fan speed or electronic expansion valve needs to be linked and controlled. When ( is the preset linkage control threshold, such as 0.2), the system will output , indicating that the fan speed should be appropriately increased and the expansion valve opening should be increased at the same time to ensure that the refrigerant flow and heat exchange capacity can compensate for the change in compressor power caused by the frequency modulation action.

[0085] The final output result includes two core variables:

[0086] : the final frequency modulation, in Hz;

[0087] : the frequency modulation target frequency for subsequent controller adjustment.

[0088] Optional output variables are:

[0089] : linkage control flag, indicating that the expansion valve / fan needs to be adjusted at the same time.

[0090] S4: execute the frequency modulation target frequency in stages by the minimum step size, and monitor the resonance sensitivity index in real time. If a new resonance is detected, trigger the reverse fine tuning to generate a new frequency modulation;

[0091] Specifically, this step will calculate the final frequency modulation and the frequency modulation target frequency The actual control command is converted to drive the variable frequency compressor to change the operating frequency, thereby avoiding the resonance frequency band. In order to improve the controllability and system stability of the execution process, the frequency modulation process is divided into two stages: initial adjustment stage and dynamic monitoring stage. In the initial stage, the controller first judges the difference between the current compressor operating frequency and the target frequency of frequency modulation , and performs equal step frequency modulation with a preset minimum frequency step . The direction is determined by , and the step is set by the PWM controller, usually 0.5Hz or 1Hz. The controller main loop period is usually 100ms, and the frequency is updated once per round. The execution logic is as follows:

[0092] ;

[0093] Where:

[0094] represents the updated operating frequency;

[0095] is the preset minimum frequency step of the controller;

[0096] is a sign function, which ensures that the frequency modulation direction is consistent with the recommended direction.

[0097] After each frequency update, the system immediately triggers a sound spectrum acquisition and disturbance response analysis, reconstructs the latest frequency spectrum and uses the disturbance sensitivity calculation mechanism of step two to obtain a new resonance sensitivity index . If the exceeds the set resonance re-determination threshold after two consecutive updates, the system considers that the current frequency modulation path may enter a new resonance frequency band.

[0098] To deal with such situations, the controller introduces a reverse fine-tuning mechanism, which applies a proportionally scaled fine-tuning amount in the opposite direction of the current frequency modulation direction to exit the new resonance boundary. The reverse fine-tuning amount is defined as follows:

[0099] ;

[0100] Where:

[0101] is the fine-tuning coefficient, usually set to 0.3 to 0.5, used to control the amplitude of the fine-tuning action;

[0102] is the latest disturbance response index;

[0103] is a resonance determination threshold value;

[0104] Still for maintaining directionality.

[0105] The reverse fine-tuning action is triggered immediately when an abnormal resonance reaction is detected, and a new frequency is written into the controller register to re-enter the next round of frequency modulation period. Through this mechanism, the system has the dynamic correction ability of "first tuning-determination-then tuning", which can greatly improve the control accuracy under the conditions of identification error, disturbance lag and working condition mutation in the frequency modulation process. The entire frequency modulation process is realized through the PWM control interface between the controller and the variable frequency drive module. Usually the frequency control word is transmitted through CAN or UART bus, and the frequency modulation response period is controlled between 0.5 seconds and 2 seconds, depending on the equipment model and system inertia dynamic adjustment.

[0106] In addition to the main frequency control channel, this step also responds to the linkage control flag output by the third step When , the system synchronously adjusts the fan speed and the electronic expansion valve opening during the frequency modulation process to match the changes of the compressor refrigerant flow and the heat exchange capacity. The specific execution logic is: according to the direction and size of the adjustment fan PWM signal (increase 10%~20% duty cycle) and expansion valve step opening (adjust 5~10 steps) to improve the stability of the refrigerant and the accuracy of the comfort control. After the frequency modulation execution is completed, the system needs to mark the execution result. If reaches , and in the last two periods are lower than , it is considered that the frequency modulation anti-vibration is successful, and the frequency modulation state flag is set; otherwise, it is determined that the frequency modulation fails, , and the controller will re-trigger the strategy module to generate a new frequency modulation suggestion.

[0107] It should be noted that the above-described workflow is only illustrative and does not limit the scope of protection of the present application. In actual application, those skilled in the art can select part or all of them to achieve the purpose of the embodiment scheme according to actual needs, which is not limited here.

[0108] In addition, technical details not described in detail in this embodiment can be referred to the parameter operation method provided by any embodiment of the present application, which will not be repeated here.

[0109] It should be noted that, in this document, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises a" does not, without more constraints, exclude the presence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0110] The above-mentioned embodiment numbers of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0111] From the above description of the embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be realized by means of software and necessary general hardware platforms, and of course can also be realized by hardware, but in many cases the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a magnetic disk, or an optical disk) and includes a number of instructions for making a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device) execute the methods described in the embodiments of the present application.

[0112] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent flow transformation made by using the content of the specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A resonance avoidance control method for variable frequency compressors based on acoustic sensing, characterized in that, The method includes: S1: Acoustic signals are collected by acoustic sensors in the compressor cavity, and a spectrum vector is generated by framing and Fourier transform. The 2nd and 3rd harmonic characteristic frequency bands of the current spectrum vector are extracted, and the current compressor operating frequency is obtained synchronously. S2: For the 2nd and 3rd harmonic characteristic frequency bands of the current spectrum vector, apply a small ±1Hz perturbation above and below the current compressor operating frequency to calculate the perturbation spectrum. Combined with the current compressor operating frequency, an energy change is generated. Subsequently, the controller statistically analyzes the high-frequency spectrum fluctuation of the current frame of the system to generate an anti-interference term. Combined with the energy change, a resonance sensitivity index is obtained, and a frequency modulation direction suggestion is generated; wherein the frequency modulation direction includes frequency boosting and frequency downing. S3: Accept the resonance sensitivity index and multiply it by a preset proportional coefficient to obtain the initial frequency modulation amplitude; calculate the relative deviation between the current compressor operating frequency and the rated frequency; if the relative deviation exceeds a preset tolerance threshold, suppress the initial frequency modulation amplitude to slow down the frequency modulation speed; generate a suggested frequency modulation amplitude based on the initial frequency modulation amplitude and the relative deviation, and obtain the target frequency modulation frequency by combining it with the current compressor operating frequency. The resonance sensitivity index triggers a linkage control flag, indicating whether linkage control of the fan speed and electronic expansion valve is required. Specifically: When the resonance sensitivity index is greater than the preset threshold, the system will output a linkage control flag = 1, indicating that the electronic expansion valve and the fan need to be adjusted simultaneously. S4: Execute the target frequency modulation in stages according to the minimum step size, and monitor the resonance sensitivity index in real time. If a new resonance is detected, trigger reverse fine-tuning to generate a new frequency modulation amplitude. When the linkage control flag is 1, a frequency modulation action is performed according to the suggested frequency modulation amplitude. The fan PWM signal and electronic expansion valve are adjusted according to the direction and magnitude of the suggested frequency modulation amplitude. After the frequency modulation is completed, an updated operating frequency is generated, and the execution result is marked with a status. If the updated operating frequency reaches the target frequency of frequency modulation, and the resonance sensitivity index is lower than the preset threshold for two consecutive cycles, then the frequency modulation vibration isolation is considered successful, and the frequency modulation status flag is set to 1; otherwise, the frequency modulation is determined to have failed, the frequency modulation status flag is set to 0, and the process returns to S3 to generate a new frequency modulation suggestion; the new frequency modulation suggestion includes the frequency modulation suggestion amplitude, the target frequency of frequency modulation, and the linkage control flag.

2. The resonance avoidance control method for variable frequency compressors based on acoustic sensing according to claim 1, characterized in that, The acoustic sensor is a capacitive electret microphone module, a standard model, with a frequency response range of 20Hz to 10kHz and a sensitivity of -42dB. The capacitive electret microphone module is connected to the ADC channel of the main controller via an analog front end, and the ADC sampling rate is fixed at 8000Hz.

3. The resonance avoidance control method for variable frequency compressors based on acoustic sensing according to claim 1, characterized in that, S1 specifically includes: Collect raw acoustic data, denoted as ; During each frame processing, the controller first multiplies the current frame data by a standard Hamming window function to reduce the impact of discontinuities at the edges of the window function on the spectrum analysis. After windowing is completed, the controller performs a fast Fourier transform on the frame data to convert the time-domain signal into a frequency-domain representation. The spectrum vector obtained after conversion , indicating in the current frame At any given moment, the acoustic energy amplitude of each frequency component inside the compressor cavity; For the spectrum vector Maximum value normalization is performed to ensure that the system can still compare energy distribution between different frequency bands under different hardware configurations; Synchronously acquire the current compressor operating frequency, and based on the current compressor operating frequency in the spectrum vector Two key frequency bands are identified: the first band is centered at twice the current compressor operating frequency, extending upwards and downwards by 5Hz each; the second band is centered at a third harmonic, similarly extending within a ±5Hz range; the first and second frequency bands together form a variable... , is used to represent the characteristic frequency region in the acoustic spectrum that is most relevant to octave resonance.

4. The resonance avoidance control method for variable frequency compressors based on acoustic sensing according to claim 1, characterized in that, S2 specifically includes: Based on the current compressor operating frequency, two disturbance frequency points are generated: the upper disturbance. and lower disturbance During the disturbance process, new acoustic spectrum data are acquired in real time to generate a disturbance spectrum vector, including the upper disturbance spectrum vector. With the lower perturbation spectrum vector ; by and Within a ±5Hz range centered on the center, two key harmonic characteristic segments are extracted, denoted as... and merged into a total frequency band and in the total frequency band Within this framework, the average energy difference between the disturbance spectrum vector and the current compressor operating frequency is compared to construct the energy changes after positive and negative disturbances; Statistical analysis of mean square fluctuations in the high-frequency band of the 0–500Hz spectrum was used to construct an anti-interference term. The resonance sensitivity index is calculated based on the anti-interference term and energy change, which is used to measure the asymmetric amplitude of the disturbance response. The larger the value, the more sensitive the current frequency is to the disturbance response, and the closer it is to the resonance critical region.

5. The resonance avoidance control method for variable frequency compressors based on acoustic sensing according to claim 4, characterized in that, The recommended frequency modulation direction is generated as follows: Set an experience threshold ; When the resonance sensitivity index is greater than the empirical threshold When this occurs, it can be determined that the current frequency has a resonance risk, and the frequency tuning avoidance process begins, including: FM direction The comparison between the upper and lower disturbance response amplitudes of the energy change is then determined: if the upper disturbance response amplitude... Less than the amplitude of the lower disturbance response This indicates a smaller response amplitude in the up-frequency direction, meaning it's more likely to escape the resonant region after up-frequency adjustment. The system sets the frequency modulation direction. Set it to +1; otherwise, set it to -1.

6. The resonance avoidance control method for variable frequency compressors based on acoustic sensing according to claim 1, characterized in that, S4 specifically includes: Obtain the suggested frequency modulation amplitude, using a preset minimum frequency step size. Perform constant-step frequency modulation, with the direction changing from the current frequency modulation direction. The step size is determined by the PWM controller. After each frequency update, return to S2 to reconstruct the latest spectrum. And calculate the new resonance sensitivity index. If a new resonance sensitivity index is detected after two consecutive updates; If the set re-resonance threshold is exceeded, reverse fine-tuning will begin.

7. The resonance avoidance control method for variable frequency compressors based on acoustic sensing according to claim 6, characterized in that, The reverse fine-tuning specifically refers to: Based on the new resonance sensitivity index By using the current frequency modulation direction A proportionally scaled fine-tuning amount is applied in the reverse direction to obtain a reverse fine-tuning amount, which is used to exit the new resonance boundary; wherein the reverse fine-tuning is triggered immediately when an abnormal resonance response is detected, and the new frequency is written to the controller register, returning to S1.

8. The resonance avoidance control method for variable frequency compressors based on acoustic sensing according to claim 1, characterized in that, The controller continuously acquires and processes the original acoustic signal at a sampling rate of 8000Hz. The original acoustic signal is divided into frames by a sliding window with each frame consisting of 128 milliseconds. Each frame contains 1024 sampling points, with an inter-frame overlap of 768 points, i.e., a frame shift of 256 points, corresponding to a time interval of 32 milliseconds, ensuring continuous coverage of the spectral information on the time axis.

Images