A multi-channel distributed active noise control system for transformer noise

Abstract

The present invention provides a multi-channel distributed active noise control system for transformer noise. This system uses a distributed algorithm to perform active noise control on the outdoor free-field transformer noise. This system includes: The transformer noise reference signal module includes multiple reference microphones that exist independently and are used to provide reference input signals to the system; The network node noise reduction component includes multiple network node noise reduction modules; Each network node noise reduction module calculates an output signal with a phase opposite to and an amplitude the same as the reference transformer noise at the current network node noise reduction module according to multiple reference input signals, and superimposes it on the transformer noise signal, so as to achieve the purpose of noise reduction at the current network node noise reduction module. The solution proposed by the present invention is used for transformer noise reduction in the outdoor free field. By reducing the ambient noise around the transformer, it brings a healthy living environment to residents and has broad expansibility and practicability.

Description

Technical Field

[0001] The present invention belongs to the field of transformers, and particularly relates to a multi-channel distributed active noise control system for transformer noise. Background Art

[0002] With the rapid development of China's economy, the electricity demand has soared rapidly, resulting in more and more power transformers being built near commercial areas and residential areas in cities, making the noise pollution problem in China increasingly serious. During the operation of transformers, low-frequency noise will be generated. These low-frequency noises have strong penetration power, strong diffraction ability, and a very wide propagation range. They will not only affect the normal operation of other electronic devices, but also damage people's physical and mental health.

[0003] Traditional passive noise control methods are difficult to solve the low-frequency noise problem by using methods such as sound absorption and blocking. Adopting the active noise control method is the main and most fundamental means to solve low-frequency noise.

[0004] The noise of outdoor transformers belongs to the noise in free space, and a multi-channel active noise control system must be adopted. In terms of control strategy, there are a large number of secondary channel couplings and complex matrix operations in the centralized multi-channel active noise control. In practical applications of the centralized multi-channel active noise control, the controller needs to bear a huge amount of computation, resulting in a very unsatisfactory noise reduction effect of the system. Using an ordinary multi-channel one-to-one distributed noise reduction system, due to ignoring the coupling between a large number of speakers and error microphones, the overall noise reduction effect will be much worse than that of the multi-channel centralized noise reduction system and cannot achieve an effective noise reduction effect. Summary of the Invention

[0005] To solve the above technical problems, the present invention proposes a technical solution for a multi-channel distributed active noise control system for transformer noise.

[0006] The first aspect of the present invention discloses a multi-channel distributed active noise control system for transformer noise, and the system includes: a transformer noise reference signal module and a network node noise reduction component;

[0007] The transformer noise reference signal module includes a plurality of reference microphones, which exist independently of each other. Each reference microphone only collects the reference transformer noise at its own position, and converts the reference transformer noise from analog quantity to digital quantity to obtain a reference input signal; the transformer noise reference signal module outputs a plurality of reference input signals to a plurality of network node noise reduction modules;

[0008] The network node noise reduction component includes multiple network node noise reduction modules; each network node noise reduction module calculates an output signal with a phase opposite to and an amplitude equal to that of the reference transformer noise at the current network node noise reduction module based on multiple reference input signals, and the output signal is superimposed on the transformer noise signal at the current network node noise reduction module, so as to achieve the purpose of noise reduction at the current network node noise reduction module.

[0009] For the system according to the first aspect of the present invention, multiple network node noise reduction modules form a ring topology network structure, and the ring topology network structure is deployed in three directions of the transformer; the k-th network node noise reduction module in the ring topology network structure is only connected to the k-1-th and k+1-th network node noise reduction modules, and the last network node noise reduction module is connected to the 1st network node noise reduction module, so that the topology network structure forms a ring.

[0010] For the system according to the first aspect of the present invention, the network node noise reduction module includes: a secondary speaker, an error microphone, multiple analog secondary channels, and a controller;

[0011] The controller is a finite impulse response filter; the multiple reference input signals are input into the controller for filtering to obtain a filtered signal; the controller then outputs the filtered signal to the secondary speaker, and the secondary speaker outputs an output signal with a phase opposite to and an amplitude equal to that of the reference transformer noise at the current network node noise reduction module;

[0012] The output of the error microphone is the difference signal between the output of the secondary speaker and the noise at the current network node noise reduction module;

[0013] The secondary channel refers to the sound channel between the secondary speaker and the error microphone;

[0014] The analog secondary channel mimics the influence of the secondary channel on the outputs of the secondary speaker and the controller;

[0015] The multiple reference input signals are input into the multiple analog secondary channels and the controller to generate a secondary input signal;

[0016] The controller updates the weight coefficients of the finite impulse response filter through the error signal and the secondary input signal.

[0017] For the system according to the first aspect of the present invention, the method by which the controller updates the weight coefficients of the finite impulse response filter through the error signal and the secondary input signal, that is, the update formula for the weight coefficients of the finite impulse response filter is:

[0018]

[0019] where \(w(n)\) is the weight coefficient of the finite impulse response filter at time \(n\), \(w(n - 1)\) is the weight coefficient of the finite impulse response filter at time \(n - 1\), \(\mu\) is the update step size, \(v\) k (n) is the secondary input signal, \(e\) k (n) is the error signal, and \(N\) is the number of network node noise reduction modules.

[0020] For the system according to the first aspect of the present invention, the analog secondary channel is obtained through offline identification. The specific method includes:

[0021] The formula for the analog secondary channel is:

[0022]

[0023] where is the weight coefficient of the analog secondary channel, \(x(n - i)\) is the input of the analog secondary channel, and \(M\) is the length of the analog secondary channel;

[0024] The weight coefficient of the analog secondary channel is updated using the least mean square algorithm. Through algorithm iteration, the error signal is made to reach a preset value, and the weight coefficient of the identified analog secondary channel is obtained.

[0025] For the system according to the first aspect of the present invention, the system further includes: a controller communication component;

[0026] The controller communication component includes a plurality of controller communication modules; the number of controller communication modules is the same as the number of network node noise reduction modules;

[0027] The \((k - 1)\)-th controller communication module transfers the weight coefficient of the finite impulse response filter of the controller of the \((k - 1)\)-th network node noise reduction module to the \(k\)-th controller communication module; the \(k\)-th controller communication module adjusts the weight coefficient of the finite impulse response filter of the controller of the \(k\)-th network node noise reduction module using the weight coefficient of the finite impulse response filter of the controller of the \((k - 1)\)-th network node noise reduction module.

[0028] For the system according to the first aspect of the present invention, the method by which the \(k\)-th controller communication module adjusts the weight coefficient of the finite impulse response filter of the controller of the \(k\)-th network node noise reduction module using the weight coefficient of the finite impulse response filter of the controller of the \((k - 1)\)-th network node noise reduction module includes:

[0029] \(w\) k (n)=w k-1 (n - 1)-\(\mu v\) k (n)e k (n), \(1\leq k\leq N\)

[0030] where \(w\)k The weight coefficient of the finite impulse response filter of the controller of the k-th network node noise reduction module adjusted at the n-th moment is w k-1 The weight coefficient of the finite impulse response filter of the controller of the (k - 1)-th network node noise reduction module at the (n - 1)-th moment is w(n - 1).

[0031] According to the system of the first aspect of the present invention, 16 network node noise reduction modules are deployed in a ring topology network structure in three directions of the transformer;

[0032] Among them, 8 network node noise reduction modules are selected and connected in the first direction as the first line, and the minimum distance between each network node noise reduction module of the first line is 1.8 meters; the second direction and the third direction are at 90° to the first line and are respectively located on the left and right sides of the first line; 4 network node noise reduction modules are selected in the second direction as the second line, and the minimum distance between each network node noise reduction module of the second line is 1.8 meters; 4 network node noise reduction modules are selected in the third direction as the third line, and the minimum distance between each network node noise reduction module of the third line is 1.8 meters; the placement distance between the secondary speaker and the error microphone of each network node noise reduction module is 50 centimeters; the error microphones and speakers of all network node noise reduction modules are flush with the midpoint height of the outdoor transformer.

[0033] The second aspect of the present invention provides an electronic device, which includes a memory and a processor. A computer program is stored on the memory. When the computer program is executed by the processor, it executes the method in a multi-channel distributed active noise control system for transformer noise as described in the first aspect of the present invention.

[0034] The third aspect of the present invention provides a storage medium. The computer program stored on the storage medium can be executed by one or more processors and can be used to implement the method in a multi-channel distributed active noise control system for transformer noise as described in the first aspect of the present invention.

[0035] The solution proposed by the present invention is used for noise reduction of transformers in free fields, which reduces environmental noise and brings a healthy living environment to residents, and has broad scalability and practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0037] Figure 1 Layout diagram of the noise reduction module of the network node of a multi-channel distributed active noise control system for transformer noise according to an embodiment of the present invention;

[0038] Figure 2 Schematic diagram of the structure of a finite impulse response filter according to an embodiment of the present invention

[0039] Figure 3 Structure diagram of an electronic device according to an embodiment of the present invention. Detailed implementation manners

[0040] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are only a part rather than all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0041] The first aspect of the present invention discloses a multi-channel distributed active noise control system for transformer noise, Figure 1 Structure diagram of a multi-channel distributed active noise control system for transformer noise according to an embodiment of the present invention, specifically as Figure 1 shown. The system includes: a transformer noise reference signal module and a network node noise reduction component;

[0042] The transformer noise reference signal module includes a plurality of reference microphones, which are independent of each other. Each reference microphone only collects the reference transformer noise at its own position, and converts the reference transformer noise from analog quantity to digital quantity to obtain a reference input signal; the transformer noise reference signal module outputs a plurality of reference input signals to a plurality of network node noise reduction modules;

[0043] The network node noise reduction component includes a plurality of network node noise reduction modules; each network node noise reduction module calculates an output signal with a phase opposite to and an amplitude the same as the reference transformer noise at the current network node noise reduction module according to a plurality of reference input signals, and the output signal is superimposed on the transformer noise signal at the current network node noise reduction module, so as to achieve the purpose of noise reduction at the current network node noise reduction module.

[0044] The transformer noise reference signal module includes multiple reference microphones, which exist independently of each other. Each reference microphone only collects the reference transformer noise at its own position, converts the reference transformer noise from analog to digital, and obtains a reference input signal. The transformer noise reference signal module outputs multiple reference input signals to multiple network node noise reduction modules.

[0045] Specifically, the function of the transformer noise reference signal module is to collect the transformer reference noise signal, convert the transformer reference noise signal into a reference input signal, and transmit the reference input signal to all network node noise reduction modules. The transformer noise reference signal module consists of multiple reference microphones, which perform original acquisition of the transformer noise near the transformer and convert it into a reference input signal. The multiple reference microphones exist independently of each other. Each reference microphone only collects the reference transformer noise at its own position and converts this analog signal of the reference transformer noise into a digital reference input signal through an analog-to-digital conversion (A / D) hardware device. The transformer noise reference signal module will output multiple reference input signals generated by the multiple reference microphones to each network node noise reduction module. The reference input signal is not the noise source signal generated by the transformer, but the signal obtained after the noise source signal generated by the transformer passes through the environment between the noise source and the transformer noise reference signal module. The function of the reference input signal is to provide the reference transformer noise at the position of the transformer noise reference signal module to the network node noise reduction module, and the network node noise reduction module calculates the output signal at the position of its own network node noise reduction module through the reference input signal.

[0046] The transformer noise reference signal module respectively collects the 220KV and 380KV outdoor transformers. The centralized frequency range of the obtained reference input signal is 100 - 500Hz, so it can be concluded that the reference input signal is a narrowband signal. At the same time, through spectrum observation, it can be obtained that the reference input signal is a non-steady signal, that is, a signal whose distribution law of amplitude and frequency changes with time. The transformer noise reference signal module inputs the reference input signal to all network node noise reduction modules, and the network node noise reduction modules perform noise reduction.

[0047] The network node noise reduction component includes multiple network node noise reduction modules; each network node noise reduction module calculates an output signal with a phase opposite to and an amplitude the same as the reference transformer noise at the current network node noise reduction module based on multiple reference input signals, and the output signal is superimposed on the transformer noise signal at the current network node noise reduction module, so as to achieve the purpose of noise reduction at the current network node noise reduction module.

[0048] In some embodiments, such as Figure 1As shown, multiple network node noise reduction modules are deployed in a ring topology network structure in three directions of the transformer; in the ring topology network structure, the k-th network node noise reduction module is only connected to the (k-1)-th and (k+1)-th network node noise reduction modules, and the last network node noise reduction module is connected to the 1st network node noise reduction module, making the topology network structure into a ring.

[0049] In some embodiments, the network node noise reduction module includes: a secondary speaker, an error microphone, multiple analog secondary channels, and a controller;

[0050] The controller is a finite impulse response filter; the multiple reference input signals are input into the controller for filtering to obtain a filtered signal; the controller then outputs the filtered signal to the secondary speaker, and the secondary speaker outputs an output signal with a phase opposite to and an amplitude equal to that of the reference transformer noise at the current network node noise reduction module;

[0051] The output of the error microphone is the difference signal between the output of the secondary speaker and the noise at the current network node noise reduction module, that is, the error signal;

[0052] The secondary channel refers to the sound channel between the secondary speaker and the error microphone;

[0053] The analog secondary channel mimics the influence of the secondary channel on the outputs of the secondary speaker and the controller;

[0054] The multiple reference input signals are input into the multiple analog secondary channels and the controller to generate a secondary input signal;

[0055] The controller updates the weight coefficients of the finite impulse response filter through the error signal and the secondary input signal.

[0056] The method by which the controller updates the weight coefficients of the finite impulse response filter through the error signal and the secondary input signal includes:

[0057] The formula of the finite impulse response filter is:

[0058]

[0059] where y(n) is the output of the finite impulse response filter, x(n-l) is the input of the finite impulse response filter, l = 0, 1,..., L-1, L is the length of the finite impulse response filter, and w(n) is the weight coefficient of the finite impulse response filter at time n;

[0060] The update formula of the weight coefficients of the finite impulse response filter is:

[0061]

[0062] Among them, w(n) is the weight coefficient of the finite impulse response filter at time n, w(n - 1) is the weight coefficient of the finite impulse response filter at time n - 1, μ is the update step size, and v k (n) is the secondary input signal, and e k (n) is the error signal, and N is the number of network node noise reduction modules.

[0063] w(n) is the vector form of the weight coefficient w(n) of the controllers of all network node noise reduction modules;

[0064] The simulated secondary channel is obtained through offline identification. The specific method includes:

[0065] The formula of the simulated secondary channel is:

[0066]

[0067] Among them, is the weight coefficient of the simulated secondary channel, x(n - i) is the input of the simulated secondary channel, and M is the length of the simulated secondary channel;

[0068]

[0069]

[0070] The output of the error microphone is the difference signal between the output of the secondary speaker and the noise at the current network node noise reduction module, that is, the error signal;

[0071] The least mean square algorithm is used to update the weight coefficient of the simulated secondary channel. Through algorithm iteration, the error signal reaches a preset value, and the weight coefficient of the identified simulated secondary channel is obtained.

[0072] Specifically, the network node noise reduction component includes multiple network node noise reduction modules; each network node noise reduction module calculates an output signal with the opposite phase and the same amplitude as the reference transformer noise at the current network node noise reduction module according to multiple reference input signals. The output signal is superimposed on the transformer noise signal at the current network node noise reduction module, so as to achieve the purpose of noise reduction at the current network node noise reduction module.

[0073] As Figure 1 shown, multiple network node noise reduction modules are deployed in a ring topology network structure in three directions of the transformer; the kth network node noise reduction module in the ring topology network structure is only connected to the k - 1th and k + 1th network node noise reduction modules, and the last network node noise reduction module is connected to the 1st network node noise reduction module, making the topology network structure into a ring.

[0074] The function of the network node noise reduction module is to perform noise reduction at the current network node noise reduction module. The input of the network node noise reduction module is the transformer noise reference signal module. The output of the network node noise reduction module is an output signal with the same amplitude but opposite phase to the transformer noise signal at the network node noise reduction module after calculation using the reference input signal provided by the transformer noise reference signal module. The output signal of the network node noise reduction module is superimposed on the transformer noise signal at the current network node noise reduction module. Since the two signals have opposite phases and the same amplitude, they can be superimposed and cancelled, thus achieving the purpose of noise reduction at this node. Each network node noise reduction module consists of a secondary speaker, an error microphone, multiple analog secondary channels, and a controller.

[0075] The function of the secondary speaker of the network node noise reduction module is to output the output signal. The input of the secondary speaker is the digital output signal of the controller, and the secondary speaker converts it into an analog output signal through a digital-to-analog conversion (D / A) hardware device. The analog output signal is superimposed on the transformer noise signal at the current network node noise reduction module, thus achieving the purpose of noise reduction at this node.

[0076] The output of the error microphone is the difference signal between the output of the secondary speaker and the noise at the current network node noise reduction module, that is, the error signal.

[0077] The secondary channel refers to the sound channel between the secondary speaker and the error microphone. The secondary channel is crucial for the noise reduction effect of the network node noise reduction module and is often not allowed to be ignored. The function of the analog secondary channel of the network node noise reduction module is to mathematically model the secondary channel between the secondary speaker and the error microphone and load it into a finite impulse response (FIR) filter, that is, the controller. That is, channel identification is performed on the secondary channel, and the channel identification result is set in the finite impulse response filter. The input of the analog secondary channel is the reference input signal output by the transformer noise reference signal module, and the output of the analog secondary channel is the secondary input signal. The analog secondary channel outputs the secondary input signal to the controller. The network node noise reduction module contains multiple analog secondary channels. Because in a multi-channel system, secondary channels will be generated by the secondary speakers and error microphones of each network node noise reduction module. The multiple analog secondary channels in the network node noise reduction module are the analog secondary channels between all the secondary speakers of the network node noise reduction modules and the error microphone in the current network node noise reduction module.

[0078] The network node noise reduction module uses a finite impulse response filter as the controller. The structure diagram of the finite impulse response filter is as Figure 2 shown. At time n, where x(n) is the reference input signal, y(n) is the output signal, w lis the controller weight coefficient, and the length of the finite impulse response filter is L, where l = 0, 1, ..., L - 1, z -1 represents a unit time delay. The controller has three inputs, namely all the reference input signals of the transformer noise reference signal module, the error signal of the current network node noise reduction module, and all the secondary input signals of the current network node noise reduction module. The reference input signals are filtered by the controller to obtain an output signal, and the output signal is transmitted to the secondary speaker of the current network node noise reduction module. The error signal of the current network node noise reduction module and all the secondary input signals of the current network node noise reduction module are used to update the controller weight coefficient. The controller uses the filter-X least mean square algorithm to update the controller weight coefficient in the finite impulse response filter structure. The error signal of the current network node noise reduction module changes with the transformer noise and the output signal, and all the secondary input signals of the network node noise reduction module also change with the reference input signals. To meet the system noise reduction requirements, the controller weight coefficient must be updated to generate an output signal with a phase opposite to and an amplitude equal to that of the transformer noise signal at this network node noise reduction module.

[0079] The network node noise reduction module uses the distributed incremental cooperation strategy filter-X least mean square algorithm to complete the noise reduction of the current network node. When executing the entire algorithm, the analog secondary channels of all network node noise reduction modules at the previous moment must be known. Since the secondary channels of the network node noise reduction module change little, the analog secondary channels of the network node noise reduction module are solved by the offline secondary channel identification method. N network node noise reduction modules will bring N error microphones, and N secondary speakers will form N×N secondary channels. The solution algorithm uses the least mean square (LMS) algorithm for the adaptive identification algorithm. A white noise generator is used to identify the analog secondary channels.

[0080] In the specific implementation, it consists of 16 error microphones and 16 secondary speakers, forming 196 secondary channels. All these secondary channels need to be identified offline.

[0081] The analog secondary channels are obtained through offline identification. The specific method includes:

[0082] The formula for the analog secondary channels is:

[0083]

[0084] where is the weight coefficient of the analog secondary channel, x(n - i) is the input of the analog secondary channel (when identifying the analog secondary channel offline, the input of the analog secondary channel is white noise), and M is the length of the analog secondary channel;

[0085]

[0086]

[0087] The output of the error microphone is the difference signal between the output of the secondary speaker and the noise at the current network node noise reduction module, that is, the error signal;

[0088] The least mean square algorithm is used to update the weight coefficients of the simulated secondary channel. Through algorithm iteration, the error signal reaches a preset value, and the identified weight coefficients of the simulated secondary channel are obtained.

[0089] This weight coefficient can be set as a secondary channel of [M×1] matrix and provided for use in a multi-channel distributed active noise control system. For example, the [M×1] matrix can represent the simulated secondary channel between the secondary speaker of the jth network node noise reduction module and the error microphone of the kth network node noise reduction module.

[0090] The controller adopts a distributed incremental cooperation strategy filter-X least mean square algorithm. This distributed algorithm can reduce the computational amount of processing units between network node noise reduction modules. In terms of convergence speed and final residual noise, the performance of this distributed algorithm is the same as that of the centralized algorithm.

[0091] It is composed of N network node noise reduction modules, and the transformer noise reference signal module provides I reference input signals. At time n, the reference input signal provided by the ith reference microphone is represented by x i (n). The distributed incremental cooperation strategy filter-X least mean square algorithm cascades the controllers of N network node noise reduction modules into a [ILN×1] controller vector w(n), where L is the length of the controller vector of the kth network node noise reduction module, and L = 150.

[0092] w(n) = [w1 T (n), w2 T (n),..., w N T (n)] T

[0093] Among them, the vector at the controller of the kth network node noise reduction module is represented by [IL×1] w k (n), where k = 1,..., N. And the controller of each network node noise reduction module is w(n). Therefore, each network node noise reduction module contains the vectors at the controllers of all network node noise reduction modules, where (·) T represents the transpose of a matrix or vector.

[0094] w k (n) = [w 1k T (n), w 2k T (n),..., w Ik T (n)] T

[0095] The analog secondary channel formed by the k-th network node noise reduction module error microphone in the k-th network node noise reduction module and the secondary speaker of the j-th network node noise reduction module is represented by an [M×1] matrix. The length of the analog secondary channel is M = 256, where j = 1,..., N. Where v jk (n) represents that this analog secondary channel filters the [IL×1] vector of all reference input signals.

[0096]

[0097] Matrix X(n) is the vertical concatenation of matrix X i (n), i = 1,..., I. Matrix X i (n) is the cyclic permutation matrix of the last M + L samples of the reference input signal x i (n). Matrix X(n) contains the last L + M samples of all reference input signals x i (n).

[0098]

[0099] From the centralized filter-X least mean square algorithm controller weight coefficient update equation, the controller weight coefficient update equation in the distributed incremental cooperation strategy filter-X least mean square algorithm can be deduced.

[0100]

[0101] In the ring topology network structure, the controller in the k-th network node noise reduction module can only be calculated by the k-th network node noise reduction module. Only the error signal e k (n) provided by the error microphone of the network node noise reduction module and the analog secondary channel affect the controller update. μ is the fixed step size of the controller formula iteration, μ = 0.001.

[0102] This algorithm processes the update of the controller weight coefficients in the network node noise reduction module according to an incremental strategy. At time n, a complete cycle is executed along the ring topology network, where each network node noise reduction module calculates the w(n) formula, sets it as the controller vector of this network node noise reduction module, and passes it to the next network node noise reduction module in incremental order.

[0103] w(n) = w(n - 1) - μv1(n)e1(n) - μv1(n)e1(n) -... - μv N (n)e N (n) The local version of the controller vector w(n) at the k-th network node noise reduction module is defined as w k (n).

[0104]

[0105] w k (n) is defined as the local version of w(n) at the k-th network node noise reduction module.

[0106] In some embodiments, the system further includes: a controller communication component;

[0107] The controller communication component includes a plurality of controller communication modules; the number of controller communication modules is the same as the number of network node noise reduction modules;

[0108] The (k - 1)-th controller communication module passes the weight coefficients of the finite impulse response filter of the controller of the (k - 1)-th network node noise reduction module to the k-th controller communication module; the k-th controller communication module adjusts the weight coefficients of the finite impulse response filter of the controller of the k-th network node noise reduction module by applying the weight coefficients of the finite impulse response filter of the controller of the (k - 1)-th network node noise reduction module.

[0109] The method by which the k-th controller communication module adjusts the weight coefficients of the finite impulse response filter of the controller of the k-th network node noise reduction module by applying the weight coefficients of the finite impulse response filter of the controller of the (k - 1)-th network node noise reduction module includes:

[0110] w k (n) = w k-1 (n - 1) - μv k (n)e k (n), 1 ≤ k ≤ N

[0111] where, w k (n) is the weight coefficient of the finite impulse response filter of the controller of the k-th network node noise reduction module adjusted at time n, w k-1(n - 1) is the weight coefficient of the finite impulse response filter of the controller of the (k - 1)-th network node noise reduction module at the (n - 1)-th moment.

[0112] Specifically, the main function of the controller communication module is to enable information transfer between controllers. In a general multi-channel active noise control structure, each controller exists independently and cannot transfer information to each other. With the controller communication module, information can be transferred between controllers. The controller communication module is only an abstract concept in terms of structure. In the hardware system, the multi-channel buffered serial port (McBSP) of the digital signal processing chip produced by Texas Instruments is used for information transfer between controllers. The number of controller communication modules is equal to the number of network node noise reduction modules. The system consists of N network node noise reduction modules, and there should also be N controller communication modules.

[0113] In the ring topology network structure, the controller communication module completes the communication between the controllers of N network node noise reduction modules through the incremental cooperation strategy. The incremental cooperation strategy uses the controller communication module to update the controllers of all network node noise reduction modules in the ring topology network. And the information transfer of the incremental cooperation strategy is from the controller of the (k - 1)-th network node noise reduction module to the controller of the k-th network node noise reduction module, and cannot be transferred in the reverse direction, but the controller of the N-th network node noise reduction module transfers to the controller of the first network node noise reduction module. The input of the controller communication module is the controller weight coefficient of the (k - 1)-th network node noise reduction module, and the output is given to the controller of the k-th network node noise reduction module, which is used for the k-th network node noise reduction module to update its own controller weight coefficient.

[0114] The multi-channel ring topology network structure combined by each network node noise reduction module can replace the general multi-channel active noise control structure. The controller weight coefficient calculated by each network node noise reduction module will be transferred to the controller of the next network node noise reduction module, and the entire ring topology network structure has the characteristic of a ring, which can allow information to be circulated between nodes, rather than in a general multi-channel active noise control structure where each controller exists independently and cannot effectively transfer information to each other. In the ring topology network structure strategy, the distance between each network node noise reduction module and the distance between the secondary speaker and the error microphone of all network node noise reduction modules are important relevant parameters that affect the noise reduction of the entire ring topology network structure, and are also one of the important influencing parameters for the real-time performance and noise reduction effectiveness of the overall multi-channel active noise control system.

[0115] Starting from the local network node noise reduction module for analysis, when the current network node noise reduction module k = 1 and the time is n, w 1 (n) update equation can be obtained.

[0116] w 1w(n) = w(n - 1) - μv1(n)e1(n)

[0117] The w at the n - 1 moment of the noise reduction module controller of the first network node 1 , is passed to the noise reduction module of the second network node, and this network node noise reduction module updates the local version of the controller to the new w 2 .

[0118] w 2 w(n) = w 1 (n - 1) - μv2(n)e2(n)

[0119] The second network node noise reduction module passes its local version of the controller to the third network node noise reduction module, and so on until a complete round of the ring - topology network is completed.

[0120] w k w(n) = w k-1 (n - 1) - μv k (n)e k (n), 1 ≤ k ≤ N

[0121] The relationship between the n - th moment and the (n - 1) - th moment can be obtained from w 0 w(n) = w N (n - 1) = w(n - 1). When the network node noise reduction module is updated from 1 to N, this system has obtained the updated controller vector w(n), which is equal to the controller vector of the last network node noise reduction module w(n) = w N (n).

[0122] At the n - th moment, the controller output signal y k (n) of the k - th network node noise reduction module

[0123] y k (n) = w k T (n)[X(n)] (：，1)

[0124] The relationship between the controller w k (n) and w(n - 1) at the k - th network node noise reduction module is w k (n) = [w(n - 1)] (IL(k-1)+1：ILk) , that is, the controller for the output signal is the controller coefficient updated at the (n - 1) - th moment. Where [X(n)] (：，1) is the vector of the first column of X(n) of [IL×1].[[]]

[0125] In summary, compared with the traditional multi-channel centralized active noise control system, its advantages lie in that while considering the transformer noise as narrowband non-stationary noise, the use of a ring topology network structure greatly reduces the computational load of each node, while not changing the overall computational complexity and the effect of residual noise. Compared with the centralized multi-channel noise reduction system where only one controller can be used for all channels, each network node in the distributed system can adopt a controller. Compared with the one-to-one distributed multi-channel noise reduction system that does not consider coupling, the distributed incremental cooperation strategy filter-X least mean square algorithm takes into account the coupling relationship between each node, and the noise reduction effect is much better than that of the distributed multi-channel coupling system without consideration.

[0126] The specific method of this system adopts a multi-channel distributed noise reduction system with N = 16 to process the 220KV and 380KV free-field transformer noise. On this basis, some expansions can also be made. For example, the number of channels can be increased to make N = 32, 64, 96,...; different fixed step sizes μ of the controller can be set; the length L of the controller can be changed; the length M of the analog secondary channel can be changed; the spacing between each node can be adjusted; the spacing between the node speaker and the error microphone can be adjusted; different parameters can be set to find the optimal parameters.

[0127] The second aspect of the present invention discloses an electronic device, including a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the steps of the method in a multi-channel distributed active noise control system for transformer noise according to any one of the first aspects disclosed in the present invention are implemented.

[0128] Figure 3 For the structural diagram of an electronic device according to an embodiment of the present invention, as Figure 3 shown, the electronic device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the electronic device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be implemented through WIFI, a carrier network, near-field communication (NFC), or other technologies. The display screen of the electronic device can be a liquid crystal display screen or an electronic ink display screen. The input device of the electronic device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the electronic device, or an external keyboard, a touchpad, or a mouse, etc.

[0129] Those skilled in the art can understand, Figure 3The structure shown is only a structural diagram of the part related to the technical solution of the present disclosure, and does not constitute a limitation on the electronic device to which the solution of the present application is applied. The specific electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0130] The third aspect of the present invention discloses a storage medium, specifically related to a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the steps of the method in a multi-channel distributed active noise control system for transformer noise according to any one of the first aspect of the present invention are implemented.

[0131] Please note that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification. The above embodiments only represent several implementation manners of the present application. Their descriptions are relatively specific and detailed, but they should not be construed as a limitation on the scope of the invention patent. It should be pointed out that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A multi-channel distributed active noise control system for transformer noise, characterized in that, The system includes: a transformer noise reference signal module and a network node noise reduction component; The transformer noise reference signal module includes a plurality of reference microphones, which are independent of each other. Each reference microphone only collects the reference transformer noise at its own position, converts the reference transformer noise from analog to digital to obtain a reference input signal; the transformer noise reference signal module outputs the plurality of reference input signals to a plurality of network node noise reduction modules; The network node noise reduction component includes a plurality of network node noise reduction modules; each network node noise reduction module calculates an output signal with a phase opposite to and an amplitude equal to that of the reference transformer noise at the current network node noise reduction module according to the plurality of reference input signals, and the output signal is superimposed on the transformer noise signal at the current network node noise reduction module, so as to achieve the purpose of noise reduction at the current network node noise reduction module; The network node noise reduction module includes: a secondary speaker, an error microphone, a plurality of analog secondary channels and a controller; The controller is a finite impulse response filter; the plurality of reference input signals are input into the controller for filtering to obtain a filtered signal; the controller then outputs the filtered signal to the secondary speaker, and the secondary speaker outputs an output signal with a phase opposite to and an amplitude equal to that of the reference transformer noise at the current network node noise reduction module; The output of the error microphone is the difference signal between the output of the secondary speaker and the noise at the current network node noise reduction module; The secondary channel refers to the actual sound channel between the secondary speaker and the error microphone; The analog secondary channel is to mathematically model the secondary channel between the secondary speaker and the error microphone and load it into a finite impulse response (FIR) filter; The plurality of reference input signals are input into the plurality of analog secondary channels and the controller to generate a secondary input signal; The controller updates the weight coefficients of the finite impulse response filter through the difference signal and the secondary input signal.

2. The multi-channel distributed active noise control system for transformer noise according to claim 1, characterized in that, A plurality of network node noise reduction modules form a ring topology network structure, and the ring topology network structure is deployed in three directions of the transformer; in the ring topology network structure, the k-th network node noise reduction module is only connected to the (k - 1)-th and (k + 1)-th network node noise reduction modules, and the last network node noise reduction module is connected to the 1st network node noise reduction module, so that the topology network structure forms a ring.

3. The multi-channel distributed active noise control system for transformer noise according to claim 1, characterized in that, The method for the controller to update the weight coefficients of the finite impulse response filter through the difference signal and the secondary input signal, that is, the update formula for the weight coefficients of the finite impulse response filter is: Among them, \(w(n)\) is the weight coefficient of the finite impulse response filter at time \(n\), \(w(n - 1)\) is the weight coefficient of the finite impulse response filter at time \(n-1\), \(\mu\) is the update step size, and \(v\) k (n) is the secondary input signal of the \(k\)-th network node noise reduction module, and \(e\) k (n) is the difference signal of the \(k\)-th network node noise reduction module, and \(N\) is the number of network node noise reduction modules.

4. The multi-channel distributed active noise control system for transformer noise according to claim 3, characterized in that, The analog secondary channel is obtained through offline identification, and the specific method includes: The formula of the analog secondary channel is: At the nth moment, is the weight coefficient of the analog secondary channel, x(n - i) is the input of the analog secondary channel, and M is the length of the analog secondary channel; The least mean square algorithm is used to obtain the weight coefficients of the identified analog secondary channel.

5. The multi-channel distributed active noise control system for transformer noise according to claim 3, characterized in that, The system further includes: a controller communication component; The controller communication component includes a plurality of controller communication modules; the number of controller communication modules is the same as the number of network node noise reduction modules; The (k - 1)-th controller communication module transfers the weight coefficients of the finite impulse response filter of the controller of the (k - 1)-th network node noise reduction module to the k-th controller communication module; the k-th controller communication module adjusts the weight coefficients of the finite impulse response filter of the controller of the k-th network node noise reduction module by using the weight coefficients of the finite impulse response filter of the controller of the (k - 1)-th network node noise reduction module.

6. The multi-channel distributed active noise control system for transformer noise according to claim 5, characterized in that, The method by which the k-th controller communication module adjusts the weight coefficients of the finite impulse response filter of the controller of the k-th network node noise reduction module by using the weight coefficients of the finite impulse response filter of the controller of the (k - 1)-th network node noise reduction module includes: w k y(n) = w k-1 y(n - 1)-μv k y(n)e k (n), 1 ≤ k ≤ N where, w k (n) is the weight coefficient of the finite impulse response filter of the controller of the k-th network node noise reduction module adjusted at the n-th moment, w k-1 (n - 1) is the weight coefficient of the finite impulse response filter of the controller of the (k - 1)-th network node noise reduction module at the (n - 1)-th moment.

Images