A method for optimizing and matching low-noise operating parameters of a refrigerator compressor

Abstract

This invention discloses a method for optimizing and matching low-noise operating parameters of a refrigerator compressor, relating to the field of compressor noise control technology. By acquiring compressor operating parameter data and broadband acoustic fingerprint data, quantum analysis technology is used to process the acoustic fingerprint information. A parameter reference model is constructed by combining physical constraint modeling and quantum computing to accelerate the solution. Parameter feature data is determined through data statistical processing and incremental edge learning models. Furthermore, by combining broadband acoustic fingerprint feedback, an optimization algorithm is used to determine the optimal operating parameter matching scheme. This scheme solves the problems of insufficient data processing accuracy, poor real-time performance of parameter models, and low efficiency in integrating multi-source acoustic fingerprint data in traditional optimization methods. It provides an efficient and accurate parameter matching strategy for low-noise operation of refrigerator compressors, improves user experience, and expands the application scenarios of noise reduction technology for smart home devices.

Description

Technical Field

[0001] This invention relates to the field of compressor noise control technology, and more specifically, to a method for optimizing and matching low-noise operating parameters of a refrigerator compressor. Background Technology

[0002] During refrigerator use, compressor noise has always been a significant factor affecting user experience. Traditional methods for controlling refrigerator compressor noise have many limitations. Early methods mostly relied on optimizing the mechanical structure, such as adding shock-absorbing pads. However, this approach often has limited effectiveness and is difficult to adapt to different operating conditions. With technological advancements, people have begun to focus on the impact of operating parameters on noise, attempting to reduce noise by optimizing these parameters. However, existing methods may not be comprehensive or accurate enough in acquiring operating parameter data, failing to fully utilize quantum analysis technology to process acoustic fingerprint data, resulting in an inaccurate understanding of noise characteristics. In terms of parameter modeling, traditional modeling methods are slow to solve and cannot reflect compression noise in a timely manner. The current technology lacks sufficient processing and analysis capabilities for real-time data, failing to effectively combine historical operating parameters with current parameters for in-depth comparative analysis to uncover key parameter characteristics. Consequently, it is difficult to determine the optimal operating parameter matching scheme to achieve low-noise operation. Furthermore, the statistical analysis of multiple sets of operating parameter data is not detailed enough, failing to fully consider the distribution characteristics of data in each dimension and the dynamic characteristics of data changes over time. This makes it impossible to fully consider the timeliness and variability of data during parameter optimization. In terms of utilizing voiceprint data, existing technologies have failed to effectively integrate data from multiple voiceprint acquisition devices and have not weighted the voiceprint data from different locations, resulting in inaccurate localization and assessment of noise sources.

[0003] Existing technologies suffer from inaccurate acquisition of operating parameter data and voiceprint processing, slow modeling and solving with a lack of real-time performance, and insufficient multi-dimensional data statistical analysis and voiceprint data integration and utilization, making it difficult to adapt parameter optimization to low-noise requirements. Summary of the Invention

[0004] To overcome the problems of inaccurate acquisition of operating parameter data and voiceprint processing, slow modeling and solving with a lack of real-time performance, and insufficient multi-dimensional data statistical analysis and voiceprint data integration and utilization in existing technologies, this invention discloses an optimized matching method for low-noise operating parameters of a refrigerator compressor, which can effectively solve the above-mentioned technical problems.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] A method for optimizing and matching low-noise operating parameters of a refrigerator compressor includes the following steps:

[0007] The refrigerator compressor acquires multiple sets of operating parameter data and corresponding broadband acoustic data within a first time period. The broadband acoustic data is acquired based on broadband acoustic data acquisition devices set around the refrigerator compressor and processed by quantum analysis technology. The operating parameter data includes physical parameters in multiple dimensions.

[0008] Physical constraint modeling is performed on the operating parameter data, and quantum computing is used to accelerate the solution to obtain a parameter reference model of the refrigerator compressor in the first time period.

[0009] The operating parameter data and broadband acoustic signature data of the refrigerator compressor at the current moment are obtained; the current moment is the moment after the first time period, and the operating parameter data at the current moment includes physical parameters of multiple dimensions;

[0010] Based on the operating parameter data within the first time period and the operating parameter data at the current moment, data statistical processing is performed on the physical parameters in each dimension to obtain the data statistical value corresponding to each dimension;

[0011] Based on the statistical values ​​of the data corresponding to each dimension and the operating parameter data at the current moment, the parameter feature data of the refrigerator compressor are determined using an incremental edge learning model.

[0012] Based on the parameter feature data and the parameter reference model, and combined with the feedback from broadband acoustic data, the optimal operating parameter matching scheme for the refrigerator compressor at the current moment is determined.

[0013] Preferably, the step of performing physical constraint modeling on the operating parameter data and combining it with quantum computing to accelerate the solution, to obtain the parameter reference model of the refrigerator compressor during the first time period, includes:

[0014] Based on the physical characteristics and working principle of the refrigerator compressor, a mathematical model containing multiple physical constraints is established.

[0015] The operating parameter data is mapped to the quantum state space, and the physical constraint model is solved more quickly using quantum algorithms to obtain the parameter weight coefficients for each dimension.

[0016] Based on the parameter weight coefficients of each dimension, a parameter reference model of the refrigerator compressor is constructed during the first time period.

[0017] Preferably, the data statistics include the average value and variance of each dimension; the parameter characteristic data for determining the refrigerator compressor includes:

[0018] Based on the physical parameters of each dimension in the current operating parameter data, and the average value of the data in the corresponding dimension, determine the parameter difference for each dimension;

[0019] The parameter difference of each dimension and the data variance of the corresponding dimension are used as inputs and processed by an incremental edge learning model to obtain the parameter feature data of the refrigerator compressor in each dimension.

[0020] Preferably, the optimal operating parameter matching scheme for the refrigerator compressor at the current moment includes:

[0021] The product of the parameter feature data of each dimension and the weight coefficient of the corresponding dimension in the parameter reference model is calculated to obtain the weighted parameter value of the refrigerator compressor in each dimension.

[0022] The weighted parameter values ​​corresponding to each dimension are summed to obtain the total parameter feature value of the refrigerator compressor at the current moment;

[0023] Analyze the broadband acoustic data to extract noise-related feature information;

[0024] By combining the total value of the parameter characteristics and the noise characteristic information of the broadband acoustic data, the optimal operating parameter matching scheme of the refrigerator compressor at the current moment is determined by an optimization algorithm.

[0025] Preferably, the number of broadband voiceprint acquisition devices is multiple; the broadband voiceprint data within the first time period includes multiple sets of data corresponding to each broadband voiceprint acquisition device; the broadband voiceprint data at the current moment includes multiple sets of data corresponding to each broadband voiceprint acquisition device; the method includes:

[0026] Determine the measurement weight data for each broadband acoustic signature acquisition device; the measurement weight data is determined based on the measurement location of the broadband acoustic signature acquisition device around the refrigerator compressor;

[0027] Calculate the noise characteristic value corresponding to each broadband acoustic signature acquisition device;

[0028] Based on the measurement weight data of each broadband acoustic signature acquisition device and the noise characteristic value corresponding to the broadband acoustic signature acquisition device, the weighted noise characteristic value of each broadband acoustic signature acquisition device is determined;

[0029] Based on the weighted noise characteristic value of each broadband acoustic signature acquisition device and combined with the operating parameter data, the optimal operating parameter matching scheme for the refrigerator compressor is determined.

[0030] Preferably, the first time period includes multiple detection time nodes; after acquiring the operating parameter data and broadband acoustic signature data of the refrigerator compressor at the current moment, the method includes:

[0031] Based on the operating parameter data and broadband voiceprint data corresponding to the current moment, as well as the operating parameter data and broadband voiceprint data corresponding to the adjacent multiple detection time nodes before the current moment, the operating parameter data and broadband voiceprint data within the first time period are updated to obtain the updated operating parameter data and broadband voiceprint data; the number of detection time nodes corresponding to the updated operating parameter data and broadband voiceprint data is consistent with the number of detection time nodes included in the first time period.

[0032] Preferably, after acquiring the operating parameter data and broadband acoustic signature data of the refrigerator compressor at the current moment, the method includes:

[0033] Calculate the minimum distance between the operating parameter data within the first time period and the operating parameter data at the current moment in each dimension;

[0034] If the minimum distance in at least one dimension is greater than a preset distance threshold, the operating parameter data and parameter reference model for the first time period are corrected.

[0035] Preferably, the step of correcting the operating parameter data and parameter reference model within the first time period when the minimum distance in at least one dimension is greater than a preset distance threshold includes:

[0036] Calculate the physical parameters of the current running parameter data in each dimension, and the difference between the physical parameters of the corresponding dimension in the previous detection time node at the current time, to obtain the data offset value corresponding to each dimension;

[0037] Based on the physical parameters of the operating parameter data in each dimension during the first time period, and the data offset value corresponding to each dimension, the operating parameter data during the first time period is corrected.

[0038] The parameter reference model is reconstructed based on the revised operating parameter data.

[0039] An electronic device includes: a memory and at least one processor, the memory storing instructions, wherein at least one processor invokes the instructions in the memory to cause the device to perform various steps of the optimized matching method as described above.

[0040] A computer-readable storage medium storing instructions that, when executed by a processor, implement the steps of the optimized matching method as described above.

[0041] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention acquires broadband acoustic data around the refrigerator compressor using a broadband acoustic data acquisition device and processes it using quantum analysis technology. Compared with traditional acoustic data processing methods, quantum analysis technology can analyze the subtle features in the acoustic data more deeply, improving the accuracy and richness of the acoustic data. Simultaneously, the acquired operating parameter data covers multiple dimensions of physical parameters, comprehensively reflecting the compressor's operating status and ensuring the completeness and accuracy of the acquired operating parameter data. Based on the physical characteristics and working principle of the refrigerator compressor, a mathematical model containing multiple physical constraints is constructed, and the operating parameter data is mapped to a quantum state space. Quantum algorithms are used to accelerate the solution, not only improving the solution speed but also obtaining more accurate parameter weight coefficients in a short time, thereby quickly constructing a parameter reference model. By acquiring data at multiple detection time points within the first time period and dynamically updating it in conjunction with the current data, the changing trends of the compressor's operating parameters can be reflected in a timely manner, ensuring that the parameter reference model accurately describes the real-time operating status of the compressor. This invention addresses the issues of slow modeling and lack of real-time performance in existing technologies. Based on the operating parameter data within the first time period and the current operating parameter data, it performs statistical processing on each dimension to obtain statistical values ​​such as the average value and variance. Then, it uses an incremental edge learning model to determine parameter feature data, fully mining the feature information contained in the multi-dimensional data to achieve in-depth analysis of the compressor's operating status. Simultaneously, it determines the measurement weight data for multiple broadband acoustic fingerprint acquisition devices and calculates weighted noise feature values ​​accordingly. It integrates acoustic fingerprint data from different locations to more accurately reflect the actual noise situation of the compressor. Finally, combining the parameter feature data, parameter reference model, and feedback from broadband acoustic fingerprint data, it determines the optimal operating parameter matching scheme through optimization algorithms. This overcomes the shortcomings of insufficient multi-dimensional data statistical analysis and acoustic fingerprint data integration in existing technologies, enabling the optimization of refrigerator compressor operating parameters to accurately adapt to low-noise requirements, improving the low-noise operation performance of the refrigerator compressor, and providing users with a guarantee for reducing the overall noise level. Attached Figure Description

[0042] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other embodiments can be derived from the provided drawings without creative effort.

[0043] Figure 1 A step-by-step diagram illustrating a method for optimizing and matching low-noise operating parameters of a refrigerator compressor. Detailed Implementation

[0044] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this patent.

[0045] To better illustrate this embodiment, some parts in the accompanying drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions;

[0046] It will be understood by those skilled in the art that certain well-known structures and their descriptions may be omitted in the accompanying drawings.

[0047] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0048] Example

[0049] A method for optimizing and matching low-noise operating parameters of a refrigerator compressor includes the following steps:

[0050] The refrigerator compressor acquires multiple sets of operating parameter data and corresponding broadband acoustic data within a first time period. The broadband acoustic data is acquired based on broadband acoustic data acquisition devices set around the refrigerator compressor and processed by quantum analysis technology. The operating parameter data includes physical parameters in multiple dimensions.

[0051] Physical constraint modeling is performed on the operating parameter data, and quantum computing is used to accelerate the solution to obtain a parameter reference model of the refrigerator compressor in the first time period.

[0052] The operating parameter data and broadband acoustic signature data of the refrigerator compressor at the current moment are obtained; the current moment is the moment after the first time period, and the operating parameter data at the current moment includes physical parameters of multiple dimensions;

[0053] Based on the operating parameter data within the first time period and the operating parameter data at the current moment, data statistical processing is performed on the physical parameters in each dimension to obtain the data statistical value corresponding to each dimension;

[0054] Based on the statistical values ​​of the data corresponding to each dimension and the operating parameter data at the current moment, the parameter feature data of the refrigerator compressor are determined using an incremental edge learning model.

[0055] Based on the parameter feature data and the parameter reference model, and combined with the feedback from broadband acoustic data, the optimal operating parameter matching scheme for the refrigerator compressor at the current moment is determined.

[0056] The step of performing physical constraint modeling on the operating parameter data and combining it with quantum computing to accelerate the solution, to obtain the parameter reference model of the refrigerator compressor during the first time period, includes:

[0057] Based on the physical characteristics and working principle of the refrigerator compressor, a mathematical model containing multiple physical constraints is established.

[0058] The operating parameter data is mapped to the quantum state space, and the physical constraint model is solved more quickly using quantum algorithms to obtain the parameter weight coefficients for each dimension.

[0059] Based on the parameter weight coefficients of each dimension, a parameter reference model of the refrigerator compressor is constructed during the first time period.

[0060] The data statistics include the average value and variance of each dimension; the parameter characteristic data for determining the refrigerator compressor includes:

[0061] Based on the physical parameters of each dimension in the current operating parameter data, and the average value of the data in the corresponding dimension, determine the parameter difference for each dimension;

[0062] The parameter difference of each dimension and the data variance of the corresponding dimension are used as inputs and processed by an incremental edge learning model to obtain the parameter feature data of the refrigerator compressor in each dimension.

[0063] The optimal operating parameter matching scheme for the refrigerator compressor at the current moment includes:

[0064] The product of the parameter feature data of each dimension and the weight coefficient of the corresponding dimension in the parameter reference model is calculated to obtain the weighted parameter value of the refrigerator compressor in each dimension.

[0065] The weighted parameter values ​​corresponding to each dimension are summed to obtain the total parameter feature value of the refrigerator compressor at the current moment;

[0066] Analyze the broadband acoustic data to extract noise-related feature information;

[0067] By combining the total value of the parameter characteristics and the noise characteristic information of the broadband acoustic data, the optimal operating parameter matching scheme of the refrigerator compressor at the current moment is determined by an optimization algorithm.

[0068] The number of broadband voiceprint acquisition devices is multiple; the broadband voiceprint data within the first time period includes multiple sets of data corresponding to each broadband voiceprint acquisition device; the broadband voiceprint data at the current moment includes multiple sets of data corresponding to each broadband voiceprint acquisition device; the method includes:

[0069] Determine the measurement weight data for each broadband acoustic signature acquisition device; the measurement weight data is determined based on the measurement location of the broadband acoustic signature acquisition device around the refrigerator compressor;

[0070] Calculate the noise characteristic value corresponding to each broadband acoustic signature acquisition device;

[0071] Based on the measurement weight data of each broadband acoustic signature acquisition device and the noise characteristic value corresponding to the broadband acoustic signature acquisition device, the weighted noise characteristic value of each broadband acoustic signature acquisition device is determined;

[0072] Based on the weighted noise characteristic value of each broadband acoustic signature acquisition device and combined with the operating parameter data, the optimal operating parameter matching scheme for the refrigerator compressor is determined.

[0073] The first time period includes multiple detection time nodes; after acquiring the operating parameter data and broadband acoustic signature data of the refrigerator compressor at the current moment, the method includes:

[0074] Based on the operating parameter data and broadband voiceprint data corresponding to the current moment, as well as the operating parameter data and broadband voiceprint data corresponding to the adjacent multiple detection time nodes before the current moment, the operating parameter data and broadband voiceprint data within the first time period are updated to obtain the updated operating parameter data and broadband voiceprint data; the number of detection time nodes corresponding to the updated operating parameter data and broadband voiceprint data is consistent with the number of detection time nodes included in the first time period.

[0075] After acquiring the operating parameter data and broadband acoustic signature data of the refrigerator compressor at the current moment, the method includes:

[0076] Calculate the minimum distance between the operating parameter data within the first time period and the operating parameter data at the current moment in each dimension;

[0077] If the minimum distance in at least one dimension is greater than a preset distance threshold, the operating parameter data and parameter reference model for the first time period are corrected.

[0078] The step of correcting the operating parameter data and parameter reference model within the first time period when the minimum distance value in at least one dimension is greater than a preset distance threshold includes:

[0079] Calculate the physical parameters of the current running parameter data in each dimension, and the difference between the physical parameters of the corresponding dimension in the previous detection time node at the current time, to obtain the data offset value corresponding to each dimension;

[0080] Based on the physical parameters of the operating parameter data in each dimension during the first time period, and the data offset value corresponding to each dimension, the operating parameter data during the first time period is corrected.

[0081] The parameter reference model is reconstructed based on the revised operating parameter data.

[0082] An electronic device includes: a memory and at least one processor, the memory storing instructions, wherein at least one processor invokes the instructions in the memory to cause the device to perform various steps of the optimized matching method as described above.

[0083] A computer-readable storage medium storing instructions that, when executed by a processor, implement the steps of the optimized matching method as described above.

[0084] For specific implementation details, please refer to [link / reference]. Figure 1 Five high-precision broadband acoustic signature acquisition devices are evenly distributed around the refrigerator compressor at key locations such as the top, sides, and rear of the compressor. These devices have a wide frequency response range (generally 20Hz-20kHz) and can accurately capture sound signals of various frequencies generated by the compressor during operation. Each device has independent calibration data to ensure the accuracy and consistency of the acquired sound signals. At the same time, each acquisition device is assigned a unique identifier.

[0085] Install appropriate sensors on key parts of the refrigerator compressor to monitor operating parameter data. These physical parameters include, but are not limited to:

[0086] The compressor speed is monitored in real time by a speed sensor with an accuracy of ±1 rpm.

[0087] The exhaust pressure is measured using a pressure sensor with an accuracy of ±0.01 MPa.

[0088] Intake pressure is also measured using a pressure sensor, with the same accuracy as the exhaust pressure sensor.

[0089] The operating current is obtained through a current transformer with an accuracy of ±0.01A.

[0090] Operating voltage is monitored using a voltage sensor with an accuracy of ±0.1V.

[0091] The sensor transmits the real-time data to the data acquisition system. The sampling frequency of the data acquisition system is set to 1Hz, which means that the data values ​​of each operating parameter are collected once per second. The collected data is then subjected to preliminary filtering to remove high-frequency noise interference and ensure the accuracy and reliability of the data.

[0092] The first time period is set as the first 100 hours of continuous operation of the refrigerator compressor. The time period should cover the operation of the refrigerator compressor under various working conditions, including different stages such as start-up, normal operation, and load changes, so as to obtain comprehensive operating parameter data and wideband acoustic data.

[0093] During the first time period, the data acquisition system simultaneously initiates operational parameter monitoring and broadband acoustic fingerprint acquisition. Every second, it records the physical parameter values ​​such as rotational speed, exhaust pressure, intake pressure, operating current, and operating voltage collected by various operational parameter sensors, as well as the sound signal data acquired by five broadband acoustic fingerprint acquisition devices. The operational parameter data collected at the same time are combined into a set of operational parameter data records, while the broadband acoustic fingerprint data are stored with corresponding timestamps and acquisition device identifiers for subsequent data association. During the acquisition process, the acquired sound signal data is processed using quantum analysis technology, with the specific steps as follows:

[0094] The analog sound signal is converted into a digital signal through an analog-to-digital converter, and the sampling frequency is set to 44.1kHz to meet the requirements of lossless digitization of audio signals.

[0095] The quantum Fourier transform algorithm is used to perform frequency domain transformation on digital signals, converting time-domain signals into frequency-domain signals, thereby obtaining the frequency components of sound signals and their corresponding amplitude information. Compared with the traditional Fourier transform, the quantum Fourier transform has a significant speedup advantage when processing large-scale data, and can quickly and accurately extract key frequency features in sound signals.

[0096] Quantum noise reduction is performed on the transformed frequency domain signal. By utilizing the superposition and entanglement properties of quantum states, noise components in the signal are identified and removed, thereby improving the signal-to-noise ratio and making the broadband acoustic data more clearly and accurately reflect the actual operating noise of the refrigerator compressor.

[0097] During the first 100-hour period, a total of 360,000 sets of operational parameter data records were collected, 3,600 sets per hour, for a total of 100 hours, as well as a large amount of corresponding broadband acoustic data. These data will provide a data foundation for physical constraint modeling and the construction of parameter reference models.

[0098] Based on the physical characteristics and working principle of refrigerator compressors, this paper analyzes in depth the interrelationships between various operating parameters and their impact mechanisms on noise generation. Taking compressor speed as an example, the speed directly affects the mechanical vibration and gas flow inside the compressor, thus generating varying degrees of noise. Specifically, a mathematical model is established that includes the following physical constraints:

[0099] The relationship between speed and exhaust pressure is constrained. Based on the compressor's pressure ratio characteristic curve, a nonlinear equation between speed and exhaust pressure is established, expressed as Pd = f1(N), where Pd is the exhaust pressure, N is the speed, and f1 is a functional relationship obtained from the actual performance of the compressor.

[0100] The relationship between suction pressure and operating current is constrained by considering the heat load of the evaporator and the operating efficiency of the compressor in the refrigeration system. The relationship between suction pressure and operating current is established as I = f2(Ps), where I is the operating current, Ps is the suction pressure, and f2 is also a function obtained by fitting actual measurement data.

[0101] The power constraints of operating voltage and current follow the basic principles of electricity, and establish the power relationship between operating voltage V and current I as P = V×I, where P is the input power of the compressor. This constraint is used to ensure that each operating parameter meets the requirements of power balance during the modeling process.

[0102] The correlation constraints between noise and various operating parameters were investigated through experimental research and theoretical analysis. A multiple linear regression model was established between noise level (represented by sound pressure level in broadband acoustic data) and multiple operating parameters such as speed N, exhaust pressure Pd, intake pressure Ps, operating current I, and operating voltage V. The model is L = a×N + b×Pd + c×Ps + d×I + e×V + f, where L is the noise sound pressure level and a, b, c, d, e, and f are regression coefficients. These coefficients were obtained by fitting data under different operating conditions. The above physical constraints were integrated into a mathematical model containing multiple equations to describe the intrinsic relationship between the operating parameters and noise of the refrigerator compressor.

[0103] The large amount of operational parameter data collected in the first time period is normalized to make its numerical range adaptable to the representation requirements of quantum states. Then, the processed operational parameter data is mapped to the quantum state space, with each operational parameter dimension corresponding to a qubit. Complex data relationships are represented by the superposition and entanglement states of the qubits. Quantum algorithms, such as quantum annealing or quantum genetic algorithms, are used to accelerate the solution of the physical constraint model. Taking quantum annealing as an example, it can be regarded as the process of finding the lowest energy state (i.e., the corresponding optimal solution) in the energy landscape of the quantum system.

[0104] Initialize the state of the qubits by setting the qubits corresponding to each operating parameter to a superposition state, representing all possible combinations of parameter values.

[0105] Construct the Hamiltonian of the system so that its energy function corresponds to the objective function of the physical constraint model, such as minimizing the noise level. At the same time, introduce the physical constraints as a penalty term into the Hamiltonian to ensure that the solution satisfies the physical constraints.

[0106] By gradually adjusting the quantum annealing parameters of the system, such as annealing time and quantum fluctuation intensity, the quantum system gradually evolves from the initial high temperature and high energy state to the low temperature and low energy state. During this process, the state of the qubit is continuously adjusted and eventually converges to a state that minimizes the energy of the Hamiltonian, thus obtaining the parameter weight coefficients of each operating parameter dimension.

[0107] Quantum computing accelerates the solution process, significantly reducing the solution time compared to traditional methods. The computation time, which might have taken several days, is now reduced to within a few hours. This efficiently obtains the parameter weight coefficients for each dimension, providing a crucial basis for constructing the parameter reference model.

[0108] Based on the obtained parameter weight coefficients for each dimension, a parameter reference model for the refrigerator compressor in the first time period is constructed using a weighted summation method. The parameters for the five dimensions—speed, exhaust pressure, intake pressure, operating current, and operating voltage—are multiplied by their corresponding weight coefficients, and then the results for each dimension are summed to obtain a comprehensive parameter characteristic value. This value comprehensively reflects the performance status of the refrigerator compressor under different combinations of operating parameters and its correlation with noise levels. The parameter reference model can be expressed as:

[0109] Mr = w1×N + w2×Pd + w3×Ps + w4×I + w5×V

[0110] Where Mr is the comprehensive parameter feature value of the parameter reference model, and w1, w2, w3, w4, and w5 are the weight coefficients corresponding to each dimension of the operating parameter, respectively. These weight coefficients are obtained by quantum computing to accelerate the solution. They reflect the importance of each operating parameter in affecting the performance and noise level of the refrigerator compressor. This parameter reference model will serve as an important reference standard for evaluating and optimizing the operating parameters of the refrigerator compressor at the current moment.

[0111] After the refrigerator compressor completes the first 100-hour period of operation, any subsequent operating time is defined as the current time. For example, the current time is the 100th hour and 5th minute of the subsequent operation. At this time, the refrigerator compressor is still running normally.

[0112] Following the same data acquisition method as the first time period, the operating parameter data and broadband acoustic fingerprint data of the refrigerator compressor are collected at the current moment. That is, the instantaneous values ​​of speed, exhaust pressure, intake pressure, operating current and operating voltage are obtained through various operating parameter sensors to form the operating parameter data record at the current moment. At the same time, the sound signal data at this moment is obtained using 5 broadband acoustic fingerprint acquisition devices, and is also processed using quantum analysis technology to obtain the broadband acoustic fingerprint feature data at the corresponding moment, including the frequency domain components of the sound signal and its amplitude.

[0113] The physical parameters of the operating parameters in each dimension (speed, exhaust pressure, intake pressure, operating current, and operating voltage) are statistically processed separately for the operating parameter data in the first time period and the operating parameter data at the current time, and the average value and variance of the data for each dimension are calculated:

[0114] For the speed dimension, the average speed value μN and the variance σ²N of the speed data are calculated among all 360,000 sets of operating parameter data in the first time period. The average speed value μN is obtained by adding all the speed values ​​and dividing by the number of data sets, while the variance σ²N is the sum of the squares of the differences between each speed value and the average speed value and then divided by the number of data sets.

[0115] Similarly, the average value μPd and variance σ²Pd of the exhaust pressure dimension, the average value μPs and variance σ²Ps of the intake pressure dimension, the average value μI and variance σ²I of the operating current dimension, and the average value μV and variance σ²V of the operating voltage dimension are calculated respectively. These statistical values ​​reflect the average level and data fluctuation of each operating parameter dimension in the first time period, providing a statistical basis for determining the parameter characteristic data of the refrigerator compressor.

[0116] Based on the physical parameter values ​​in each dimension of the current operating parameter data, and the previously calculated average and variance of the corresponding dimensions, the parameter feature data of the refrigerator compressor are determined using an incremental edge learning model:

[0117] For each dimension, calculate the difference between the physical parameter value at the current moment and the average value of the data in that dimension, i.e., the parameter difference. For example, in the rotational speed dimension, the parameter difference is ΔN = Ncurrent - μN, where Ncurrent is the rotational speed value at the current moment.

[0118] The parameter differences and corresponding data variances of each dimension are used as input features and fed into the incremental edge learning model. The incremental edge learning model is an efficient machine learning algorithm that can quickly learn and update new data based on the existing model without retraining the entire model. In this embodiment, the incremental edge learning model was trained in advance, using a portion of the data in the first time period as the training set, so that the model can learn the mapping relationship between the parameter differences and variances of each operating parameter dimension and the performance status of the refrigerator compressor.

[0119] The incremental edge learning model, based on the input feature data, performs internal neural network calculations, including signal transmission and processing in the input layer, hidden layer, and output layer, to output parameter feature data of the refrigerator compressor in each dimension. These parameter feature data comprehensively reflect the degree of deviation of each operating parameter dimension of the refrigerator compressor at the current moment from the average operating state in the first time period, as well as the potential performance change trend. For example, the parameter feature data may be represented by a numerical value, the absolute value of which indicates the degree of deviation, and the positive or negative sign indicates the direction of deviation (whether it is increasing or decreasing relative to the average state).

[0120] Calculate the product between the parameter feature data of each dimension and the weight coefficient of the corresponding dimension in the parameter reference model to obtain the weighted parameter value for each dimension. That is, for the speed dimension, the weighted parameter value is FN = fN×w1, where fN is the parameter feature data of the speed dimension and w1 is the weight coefficient of the speed dimension in the parameter reference model. Similarly, calculate the weighted parameter values ​​FPd, FPs, FI, and FV for each dimension of exhaust pressure, intake pressure, operating current, and operating voltage.

[0121] The weighted parameter values ​​of each dimension are added together to obtain the total parameter characteristic value of the refrigerator compressor at the current moment, Ftotal = FN + FPd + FPs + FI + FV. The total parameter characteristic value Ftotal comprehensively reflects the degree of difference between the overall operating parameter status of the refrigerator compressor at the current moment and the reference status in the first time period, as well as the potential impact on the noise level.

[0122] The current broadband voiceprint data is analyzed to extract noise-related feature information. Specifically, by analyzing the spectral characteristics of the broadband voiceprint data, the energy distribution of different frequency components in the sound signal is determined, and the dominant frequency components and their energy levels related to refrigerator compressor noise are identified. For example, if the energy in a specific frequency range, such as 1000Hz-2000Hz, is found to be significantly higher than that of other frequency components, this may be related to mechanical vibration or gas flow noise inside the compressor. The energy value in this frequency range is extracted as the noise feature value Feature_noise.

[0123] By combining the total parameter feature value Ftotal and the noise feature information Feature_noise from broadband acoustic signature data, a comprehensive evaluation function is constructed to assess the noise level of the refrigerator compressor under different combinations of operating parameters. The comprehensive evaluation function can be expressed as:

[0124] Evaluation =α×Ftotal +β×Feature_noise

[0125] Here, α and β are weighting coefficients, set based on practical experience and the assessment of the importance of the total value of parameter characteristics and noise characteristic information. For example, α is set to 0.6 and β to 0.4, reflecting their relative contributions to noise assessment. Through optimization algorithms, such as particle swarm optimization or genetic algorithms, the combination of operating parameters that minimizes the comprehensive evaluation function (Evaluation) is searched within the allowable operating parameter range of the refrigerator compressor. This determines the optimal operating parameter matching scheme for the refrigerator compressor at the current moment. The optimization algorithm continuously adjusts the values ​​of each operating parameter (speed, exhaust pressure, intake pressure, operating current, and operating voltage) through an iterative process, calculates the corresponding Evaluation value, and gradually approaches the optimal solution, ultimately obtaining the operating parameter settings that minimize the noise of the refrigerator compressor at the current moment.

[0126] A detailed analysis was conducted on the measurement locations of five broadband acoustic signature acquisition devices deployed around the refrigerator compressor. Considering the sensitivity of each location to the main noise sources of the refrigerator compressor, such as compressor vibration, motor noise, and refrigerant flow noise, as well as factors such as sound propagation paths, the measurement weight data for each broadband acoustic signature acquisition device was determined through a combination of experimental and numerical simulation methods. For example:

[0127] The acquisition device located at the top center of the compressor is closer to the core vibration source of the compressor and experiences less obstruction during sound propagation. Therefore, it is more direct and sensitive to the response of compressor vibration and motor noise, and is thus assigned a higher measurement weight, such as 0.3.

[0128] The acquisition devices located on the sides and rear are assigned different weights based on their relative positions to the main noise source and the reflection and absorption of sound by the surrounding structures. For example, the two acquisition devices on the sides are each assigned a weight of 0.25, and the acquisition device at the rear is assigned a weight of 0.2. The sum of all weights is 1, ensuring that the data from each acquisition device can be reasonably included in the calculation according to its importance during the comprehensive evaluation.

[0129] For each broadband acoustic signature acquisition device, in the broadband acoustic signature data acquired at the current moment, the noise feature information is extracted according to the above method, such as analyzing the spectral features to obtain the energy value within a specific frequency range, and the noise feature value corresponding to each device is calculated. For example, the noise feature value of the top acquisition device is L1, the noise feature values ​​of the two side acquisition devices are L2 and L3 respectively, and the noise feature value of the rear acquisition device is L4.

[0130] Based on the measurement weight data of each broadband acoustic signature acquisition device, the weighted noise characteristic value of each device is calculated. That is, the weighted noise characteristic value of the top acquisition device is 0.3×L1, the weighted noise characteristic values ​​of the two side acquisition devices are 0.25×L2 and 0.25×L3 respectively, and the weighted noise characteristic value of the rear acquisition device is 0.2×L4.

[0131] The weighted noise feature values ​​of all broadband acoustic signature acquisition devices are summed to obtain a comprehensive weighted noise feature total value Ltotal = 0.3×L1 + 0.25×L2 + 0.25×L3 + 0.2×L4. This value more accurately reflects the actual noise level of the refrigerator compressor at the current moment because it integrates noise information collected from different locations and performs weighted processing according to the importance of each location.

[0132] The total weighted noise feature value Ltotal is combined with the operating parameter data and incorporated into the comprehensive evaluation function to further optimize the operating parameter matching scheme of the refrigerator compressor. That is, when determining the optimal combination of operating parameters, not only the total parameter feature value Ftotal should be considered, but also the real noise situation reflected by the broadband acoustic data after weighted processing should be fully considered. Through the search of the optimization algorithm, the operating parameter settings that minimize the comprehensive evaluation function are found to ensure that the refrigerator compressor reaches the best low-noise operating state at the current moment.

[0133] After obtaining the current operating parameter data and broadband acoustic data, based on the current operating parameter data and broadband acoustic data, as well as the operating parameter data and broadband acoustic data corresponding to multiple adjacent detection time nodes before the current time, such as the operating parameter data and broadband acoustic data corresponding to the previous 10 hours, the operating parameter data and broadband acoustic data within the first time period are updated. The specific operation is as follows:

[0134] The data at the current moment replaces the earliest data record collected in the first time period. This is done using a rolling update method, keeping the total number of data records in the first time period constant and always covering the operating data of the most recent 100 hours. For example, if the current moment is 100 hours and 5 minutes, then the data at the current moment replaces the original data from hour 0 to 1. At the same time, the data from the subsequent hours 1 to 100 are shifted forward sequentially to form new operating parameter data and broadband acoustic data for the first time period. This updated data will continue to be used for subsequent physical constraint modeling and parameter reference model construction, ensuring that the model can reflect the recent changes in the operating status of the refrigerator compressor in a timely manner.

[0135] The minimum distance between the operating parameter data in the first time period and the operating parameter data at the current time in each dimension is calculated. Here, Euclidean distance is used as the metric. For each operating parameter dimension (speed, exhaust pressure, intake pressure, operating current, operating voltage), the absolute value of the difference between the physical parameter value at the current time and the physical parameter value of that dimension in all data records in the first time period is calculated. Then, the minimum difference in each dimension is found, which is the minimum distance in that dimension.

[0136] Preset distance thresholds are set, for example, for the speed dimension, the preset distance threshold is 50 rpm; for the exhaust pressure dimension, it is 0.03 MPa; for the intake pressure dimension, it is 0.02 MPa; for the operating current dimension, it is 0.2 A; and for the operating voltage dimension, it is 0.3 V. After calculating the minimum distance value for each dimension, it is determined whether there is at least one minimum distance value for a dimension that is greater than the corresponding preset distance threshold. For example, if the minimum distance value for the speed dimension is 60 rpm, which is greater than the preset threshold of 50 rpm, then the correction condition is met.

[0137] When the minimum distance in at least one dimension is greater than the preset distance threshold, it is necessary to correct the operating parameter data and parameter reference model in the first time period. First, calculate the difference between the physical parameter value of the operating parameter data in each dimension at the current moment and the physical parameter value of the corresponding dimension at the previous detection time node (i.e., 1 hour before the current moment) to obtain the data offset value corresponding to each dimension. For example, the data offset value of the rotation speed dimension is ΔNshift = Ncurrent - Nprevious, where Nprevious is the rotation speed value 1 hour before the current moment.

[0138] Based on the physical parameter values ​​of each dimension of the original operating parameter data within the first time period, and the calculated data offset values ​​corresponding to each dimension, the operating parameter data within the first time period is corrected. Specifically, the correction method can be to add the corresponding data offset value to the physical parameter values ​​of each dimension in each data record within the first time period to obtain the corrected operating parameter data. For example, for the speed dimension, the corrected speed value is Ncorrected = Noriginal + ΔNshift, where Noriginal is the speed value at a certain moment within the first time period before correction.

[0139] Based on the revised operating parameter data, a new parameter reference model is constructed by re-establishing the mathematical model that includes physical constraints and accelerating the solution using quantum computing, following the previous physical constraint modeling and quantum computing acceleration steps. This involves re-establishing a mathematical model that includes physical constraints, mapping the revised operating parameter data to the quantum state space, and using quantum algorithms to re-solve the parameter weight coefficients for each dimension. The new parameter reference model is then constructed to adapt to changes in the operating state of the refrigerator compressor, ensuring that subsequent optimization matching methods can be based on the accurate and updated model, thereby continuously and effectively optimizing the low-noise operating parameters of the refrigerator compressor.

[0140] The same or similar labels correspond to the same or similar parts;

[0141] The terms used to describe positional relationships in the accompanying drawings are for illustrative purposes only and should not be construed as limiting this patent.

[0142] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all implementation methods here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the claims of the present invention.

Claims

1. A method for optimizing matching of low noise operation parameters of a refrigerator compressor, characterized in that, The method comprises the following steps: obtaining a plurality of sets of operating parameter data of a refrigerator compressor in a first time period and corresponding broadband voiceprint data; the broadband voiceprint data is obtained based on a broadband voiceprint collection device arranged around the refrigerator compressor and is processed by quantum analysis technology; the operating parameter data comprises physical parameters in multiple dimensions; physically constrained modeling is performed on the operating parameter data, and quantum calculation is used to accelerate solving, to obtain a parameter reference model of the refrigerator compressor in the first time period; obtaining operating parameter data and broadband voiceprint data of the refrigerator compressor at a current time; the current time is a time after the first time period, and the operating parameter data at the current time comprises physical parameters in multiple dimensions; based on the physical parameters in each dimension of the operating parameter data in the first time period and the operating parameter data at the current time, data statistical processing is performed to obtain data statistical values corresponding to each dimension; based on the data statistical values corresponding to each dimension and the operating parameter data at the current time, an incremental edge learning model is used to determine parameter feature data of the refrigerator compressor; based on the parameter feature data and the parameter reference model, and in combination with feedback of the broadband voiceprint data, an optimal operating parameter matching scheme of the refrigerator compressor at the current time is determined.

2. The method of claim 1, wherein, The physically constrained modeling of the operating parameter data, in combination with quantum calculation for accelerated solving, to obtain the parameter reference model of the refrigerator compressor in the first time period comprises: a mathematical model comprising a plurality of physical constraint conditions is established according to the physical characteristics and working principle of the refrigerator compressor; the operating parameter data is mapped to a quantum state space, and a quantum algorithm is used to accelerate solving of the physically constrained model to obtain parameter weight coefficients in each dimension; based on the parameter weight coefficients in each dimension, the parameter reference model of the refrigerator compressor in the first time period is constructed.

3. The method of claim 1, wherein, The data statistical values comprise data average values and data variances in each dimension; and the determination of the parameter feature data of the refrigerator compressor comprises: based on the physical parameters in each dimension of the operating parameter data at the current time and the data average values in the corresponding dimensions, parameter difference values corresponding to each dimension are determined; the parameter difference values in each dimension and the data variances in the corresponding dimensions are taken as inputs, and an incremental edge learning model is used to process the inputs to obtain parameter feature data of the refrigerator compressor corresponding to each dimension.

4. The method of claim 3, wherein, The determination of the optimal operating parameter matching scheme of the refrigerator compressor at the current time comprises: a product of the parameter feature data in each dimension and the weight coefficients of the corresponding dimensions in the parameter reference model is calculated to obtain weighted parameter values corresponding to each dimension of the refrigerator compressor; the weighted parameter values corresponding to each dimension are summed to obtain a parameter feature total value of the refrigerator compressor at the current time; noise-related feature information is extracted from the broadband voiceprint data; in combination with the parameter feature total value and the noise feature information of the broadband voiceprint data, an optimal operating parameter matching scheme of the refrigerator compressor at the current time is determined by an optimization algorithm.

5. The method of claim 1, wherein, The number of the wideband voiceprint collection devices is multiple; the wideband voiceprint data in the first time period includes multiple groups of data corresponding to each wideband voiceprint collection device; the wideband voiceprint data at the current moment includes multiple groups of data corresponding to each wideband voiceprint collection device; the method comprises: determining the measurement weight data of each wideband voiceprint collection device; the measurement weight data is determined based on the measurement position of the wideband voiceprint collection device around the refrigerator compressor; calculating the noise feature value corresponding to each wideband voiceprint collection device; determining the weighted noise feature value of each wideband voiceprint collection device based on the measurement weight data of each wideband voiceprint collection device and the noise feature value corresponding to the wideband voiceprint collection device; determining the optimal operation parameter matching scheme of the refrigerator compressor based on the weighted noise feature value of each wideband voiceprint collection device and the operation parameter data.

6. The method of claim 1, wherein, The first time period includes multiple detection time nodes; after obtaining the operation parameter data and wideband voiceprint data of the refrigerator compressor at the current moment, the method comprises: updating the operation parameter data and wideband voiceprint data in the first time period based on the operation parameter data and wideband voiceprint data corresponding to the current moment and the operation parameter data and wideband voiceprint data corresponding to the adjacent multiple detection time nodes before the current moment, to obtain updated operation parameter data and wideband voiceprint data; the number of detection time nodes corresponding to the updated operation parameter data and wideband voiceprint data is consistent with the number of detection time nodes included in the first time period.

7. The method of claim 1, wherein, After obtaining the operation parameter data and wideband voiceprint data of the refrigerator compressor at the current moment, the method comprises: calculating the minimum distance value of the operation parameter data in the first time period and the operation parameter data at the current moment in each dimension; in the case that the minimum distance value of at least one dimension is greater than a preset distance threshold, correcting the operation parameter data in the first time period and the parameter reference model.

8. The method of claim 7, wherein, In the case that the minimum distance value of at least one dimension is greater than a preset distance threshold, the method for correcting the operation parameter data in the first time period and the parameter reference model comprises: calculating the physical parameter of the operation parameter data at the current moment in each dimension, and the difference between the physical parameter of the corresponding dimension at the current moment and the physical parameter of the corresponding dimension at the last detection time node before the current moment, to obtain the data offset value corresponding to each dimension; correcting the operation parameter data in the first time period based on the physical parameter of the operation parameter data in each dimension in the first time period and the data offset value corresponding to each dimension; reconstructing the parameter reference model according to the corrected operation parameter data.

9. An electronic device, comprising: comprises: a memory and at least one processor, the memory having instructions stored therein, and the at least one processor invoking the instructions in the memory to cause the device to perform the steps of the optimization matching method according to any one of claims 1-8.

10. A computer-readable storage medium having stored thereon instructions, the instructions comprising, The instructions, when executed by the processor, implement the various steps of the method for optimizing matching of claim 1-8.

Images