CN116381604A - Sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization method and device - Google Patents

Sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization method and device Download PDF

Info

Publication number
CN116381604A
CN116381604A CN202310331057.9A CN202310331057A CN116381604A CN 116381604 A CN116381604 A CN 116381604A CN 202310331057 A CN202310331057 A CN 202310331057A CN 116381604 A CN116381604 A CN 116381604A
Authority
CN
China
Prior art keywords
array
microphone array
array element
group
element group
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202310331057.9A
Other languages
Chinese (zh)
Inventor
黎敏
陈岩
冯道方
潘薇
石有泰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Science and Technology Beijing USTB
Original Assignee
University of Science and Technology Beijing USTB
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by University of Science and Technology Beijing USTB filed Critical University of Science and Technology Beijing USTB
Priority to CN202310331057.9A priority Critical patent/CN116381604A/en
Publication of CN116381604A publication Critical patent/CN116381604A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
    • G01S5/18—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using ultrasonic, sonic, or infrasonic waves
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/04—Constraint-based CAD
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Computer Hardware Design (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Circuit For Audible Band Transducer (AREA)

Abstract

The invention discloses a sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization method and device, and relates to the technical field of microphone array signal measurement. Comprising the following steps: generating a microphone array primary array element group through improved Sine chaotic mapping; constructing a mixed fitness function, and calculating the fitness of the microphone array to a sound source positioning and sound field reconstruction algorithm; and obtaining the optimized three-dimensional microphone array for measuring the multi-objective synergy of the sound field by using a whale elite optimization algorithm according to the mixed fitness function. The invention can solve the problem that the primary array element group cannot be uniformly distributed in the whole space, improves the sound source positioning-oriented and sound field reconstruction algorithm of the microphone array, has better adaptability, and realizes the optimal distribution of the three-dimensional microphone array elements and the excitation weight of the microphone array elements. Through the fusion of the three parts, the three-dimensional microphone array with high adaptability to beam forming and harmony holographic algorithm is finally optimized, and the output response characteristic of the microphone array can be effectively improved.

Description

Sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization method and device
Technical Field
The invention relates to the technical field of microphone array signal measurement, in particular to a sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization method and device.
Background
The microphone array is formed by arranging a plurality of acoustic sensor units according to a certain structure, and obtains sound source distribution by carrying out correlation analysis on measured sound signals. The structure of the microphone array affects the spatial resolution of sound source recognition and the accuracy of sound field reconstruction, so that factors affecting the performance of the microphone array need to be analyzed in order to find the most reasonable microphone array structure. The geometric parameters of the microphone array mainly include the aperture size of the microphone array, the microphone spacing, the microphone space position, the number of microphones and the like. Wherein, the larger the aperture of the microphone array, the smaller the sound source frequency can be measured; the smaller the microphone array aperture, the lower the spatial resolution of sound source identification; the microphone spacing determines the range of frequencies within which the microphone array can identify the sound source; the spatial position of the microphones determines that the microphone array has different main lobe widths and side lobe orders.
Sound source localization and sound field reconstruction are two research hot spots in the acoustic field, and are widely applied to noise identification in recent years, the sound source localization processes measured sound signals by using a localization algorithm, and the arrival direction and distance of a sound source point relative to a microphone can be obtained, so that the method is mainly used for detection of ships and vehicles, localization of main noise sources in machines, target selection and interference suppression in communication equipment or voice identification processing, and state monitoring of a mechanical system, and the most commonly used method is a beam forming algorithm. The sound field reconstruction can accurately acquire the amplitude and the spatial sound pressure distribution of a sound source through a reconstruction algorithm, and can provide references for noise evaluation and sound insulation and noise reduction through noise distribution and clear sound propagation paths of different spatial areas, and the most commonly used method is a sound holographic algorithm.
In the beam forming method, the method has excellent high-frequency space resolution capability, the beam forming is to calculate the sound source position based on the phase of the sound signal, and the larger the array element distance is, the narrower the beam bandwidth is, and the higher the positioning precision is. However, when the array element spacing is too small, the beam bandwidth is too wide, and the positioning accuracy is not high enough; the too large array element spacing causes the increase of delay among array elements, which causes the deviation of phase difference and the generation of grating lobes, thereby causing the positioning blurring phenomenon.
In the acoustic hologram algorithm, the method has excellent low-frequency space resolution capability, the acoustic hologram is used for reversely calculating the sound source position based on the amplitude of the sound signal, and as the array element distance is increased, the reconstruction position deviation is increased. Therefore, by changing the array element spacing and the array element excitation weight, the beam forming algorithm and the acoustic holographic algorithm sound source positioning result accuracy show antagonism.
In the traditional microphone array method, various optimization methods are provided, one of two problems of the positioning precision of a microphone array beam forming algorithm or the sound field reconstruction precision of an acoustic holographic algorithm can be solved, and the traditional optimization method can not effectively solve the different requirements of the positioning of a sound source and the reconstruction of the sound field on the measurement of the microphone array.
Disclosure of Invention
Aiming at the problems that in the traditional microphone array method, various optimization methods can only solve one of the two problems of the positioning precision of a microphone array beam forming algorithm or the sound field reconstruction precision of an acoustic holographic algorithm, and the traditional optimization method can not effectively solve the problems facing to the different requirements of the positioning of a sound source and the reconstruction of the sound field on the measurement of the microphone array.
In order to solve the technical problems, the invention provides the following technical scheme:
in one aspect, the invention provides a sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization method, which is implemented by electronic equipment and comprises the following steps:
s1, acquiring geometric parameters of a traditional array element group of the microphone array, and generating a primary array element group of the microphone array through improved Sine chaotic mapping.
S2, constructing a mixed fitness function.
And S3, obtaining the optimized three-dimensional microphone array for multi-objective coordination of sound field measurement according to the microphone array primary array element group, the mixed fitness function and the whale elite optimization algorithm.
Optionally, the step S1 of obtaining geometric parameters of a traditional group of array elements of the microphone array, and generating a primary array element group of the microphone array through improved fine chaotic mapping includes:
S11, acquiring geometric parameters of a traditional group array element group of the microphone array; the geometric parameters comprise the number of array elements, the initial scale of the traditional array element group, the aperture of the microphone array, the lower limit of the constraint threshold of the array element spacing and the upper limit of the constraint threshold of the array element spacing.
S12, setting control parameters and initial parameters of the improved Sine chaotic map.
S13, generating a sparse initial microphone array position according to the geometric parameters and the improved fine chaotic mapping, and obtaining a microphone array initial array element group.
Optionally, constructing the hybrid fitness function in S2 includes:
s21, constructing a peak sidelobe level PSLL as an optimization target for evaluating the suitability of the microphone array for a beam forming algorithm in sound source positioning.
S22, constructing a peak signal-to-noise ratio PSNR as an optimization target for evaluating the suitability of the microphone array to the acoustic holographic algorithm in sound field reconstruction.
S23, constructing a mixing function according to the sidelobe level and the peak signal-to-noise ratio, and further constructing a mixing fitness function.
Optionally, the mixing function is represented by the following formula (1):
Figure BDA0004154991280000031
wherein alpha is 1 ,α 2 E (0, 1), w represents the iteration sequence value of the improved Sine chaotic map, and d represents the improvementInitial parameters of the Sine chaotic map.
Optionally, the fitness function is mixed as shown in the following formula (2):
Figure BDA0004154991280000032
wherein fitness represents a mixing function, BW -3dB Representing the main lobe width, BW thr Represents the main lobe width constraint threshold, d ij Represents the spacing of any two array elements, i, j=1, 2,3 … N P ，N P The traditional array element group size of the microphone array is set; d, d min Representing the lower limit of the array element spacing constraint threshold; d, d max Representing the upper limit of the array element spacing constraint threshold.
Optionally, obtaining the optimized three-dimensional microphone array for sound field measurement multi-objective cooperation according to the microphone array primary array element group, the mixed fitness function and the whale elite optimization algorithm in S3, including:
s31, according to the primary array element group of the microphone array, an excellent array element file with the lowest value of the mixed fitness function is established, and a leading array element of the microphone array is selected to generate a deep search group array element group of the microphone array.
S32, updating the positions and the array element excitation weights of the traditional array element groups of the microphone array through a whale elite optimization algorithm, and generating a new traditional array element group of the microphone array.
S33, updating the excellent array element file according to the new traditional array element group of the microphone array, selecting a new leading array element of the microphone array, and generating a new deep search array element group of the microphone array.
S34, judging whether the array element group size of the new microphone array deep search array element group reaches the maximum value of the preset array element group size, if so, adjusting the number of individuals of the new microphone array traditional array element group and the new microphone array deep search array element group, and executing the step S35; if not, step S35 is performed.
S35, updating the excellent array element file according to the traditional array element group of the adjusted microphone array, and updating the deep search array element group.
S36, judging whether the preset iteration times are reached, if so, obtaining the position of the optimal array element group and the excitation weight of the array element, and obtaining the optimized three-dimensional microphone array for the multi-objective coordination of sound field measurement; if not, go to execute step S32.
Optionally, in S31, an excellent array element file with the lowest value of the mixed fitness function is established according to the microphone array primary array element group, and a microphone array leading array element is selected to generate a microphone array deep search group array element group, which includes:
s311, calculating a mixed fitness function value of each individual in the traditional array element group of the microphone array, and sequencing the individuals from small to large according to the mixed fitness function value.
S312, build scale N A Selecting the individuals with the lowest fitness function values for archiving.
S313, selecting a microphone array leading array element according to a deep hunting group scheme, and generating a deep searching group array element group nearby the microphone array leading array element.
Optionally, updating the position and the element excitation weight of the traditional group element group of the microphone array through a whale elite optimization algorithm in S32 to generate a new traditional group element group of the microphone array, including:
s321, carrying out iterative optimization on the positions and the array element excitation weights of the traditional group array element groups of the microphone array through surrounding the prey, high-level spiral shrinkage surrounding and high-level searching the prey, and obtaining updated traditional group array element groups of the microphone array.
S322, detecting whether an individual exceeding the aperture of the microphone array exists in the updated traditional array element group of the microphone array, and if so, randomly generating a new individual to replace the exceeding individual in the aperture of the microphone array to generate a new traditional array element group of the microphone array.
Optionally, adjusting the new microphone array legacy group element group and the new microphone array deep search group element group individual number in S34 includes:
According to the deep prey group scheme, increasing or decreasing the number of individuals of the new traditional group element group of the microphone array and the new deep search group element group of the microphone array by N 0 The total array element group is composed of a new traditional array element group of the microphone array and a new deep search array element group of the microphone array.
On the other hand, the invention provides a sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization device, which is applied to realizing a sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization method, and comprises the following steps:
the acquisition module is used for acquiring the geometric parameters of the traditional array element group of the microphone array and generating the primary array element group of the microphone array through improved Sine chaotic mapping.
And the function construction module is used for constructing a mixed fitness function.
And the output module is used for obtaining the optimized three-dimensional microphone array for measuring multi-objective coordination in the sound field according to the microphone array primary array element group, the mixed fitness function and the whale elite optimization algorithm.
Optionally, the acquiring module is further configured to:
s11, acquiring geometric parameters of a traditional group array element group of the microphone array; the geometric parameters comprise the number of array elements, the initial scale of the traditional array element group, the aperture of the microphone array, the lower limit of the constraint threshold of the array element spacing and the upper limit of the constraint threshold of the array element spacing.
S12, setting control parameters and initial parameters of the improved Sine chaotic map.
S13, generating a sparse initial microphone array position according to the geometric parameters and the improved fine chaotic mapping, and obtaining a microphone array initial array element group.
Optionally, the function construction module is further configured to:
s21, constructing a peak sidelobe level PSLL as an optimization target for evaluating the suitability of the microphone array for a beam forming algorithm in sound source positioning.
S22, constructing a peak signal-to-noise ratio PSNR as an optimization target for evaluating the suitability of the microphone array to the acoustic holographic algorithm in sound field reconstruction.
S23, constructing a mixing function according to the sidelobe level and the peak signal-to-noise ratio, and further constructing a mixing fitness function.
Optionally, the mixing function is represented by the following formula (1):
Figure BDA0004154991280000061
wherein alpha is 1 ,α 2 E (0, 1), w represents an iteration sequence value of the improved fine chaotic map, and d represents an initial parameter of the improved fine chaotic map.
Optionally, the fitness function is mixed as shown in the following formula (2):
Figure BDA0004154991280000062
wherein fitness represents a mixing function, BW -3dB Representing the main lobe width, BW thr Represents the main lobe width constraint threshold, d ij Represents the spacing of any two array elements, i, j=1, 2,3 … N P ，N P The traditional array element group size of the microphone array is set; d, d min Representing the lower limit of the array element spacing constraint threshold; d, d max Representing the upper limit of the array element spacing constraint threshold.
Optionally, the output module is further configured to:
s31, according to the primary array element group of the microphone array, an excellent array element file with the lowest value of the mixed fitness function is established, and a leading array element of the microphone array is selected to generate a deep search group array element group of the microphone array.
S32, updating the positions and the array element excitation weights of the traditional array element groups of the microphone array through a whale elite optimization algorithm, and generating a new traditional array element group of the microphone array.
S33, updating the excellent array element file according to the new traditional array element group of the microphone array, selecting a new leading array element of the microphone array, and generating a new deep search array element group of the microphone array.
S34, judging whether the array element group size of the new microphone array deep search array element group reaches the maximum value of the preset array element group size, if so, adjusting the number of individuals of the new microphone array traditional array element group and the new microphone array deep search array element group, and executing the step S35; if not, step S35 is performed.
S35, updating the excellent array element file according to the traditional array element group of the adjusted microphone array, and updating the deep search array element group.
S36, judging whether the preset iteration times are reached, if so, obtaining the position of the optimal array element group and the excitation weight of the array element, and obtaining the optimized three-dimensional microphone array for the multi-objective coordination of sound field measurement; if not, go to execute step S32.
Optionally, the output module is further configured to:
s311, calculating a mixed fitness function value of each individual in the traditional array element group of the microphone array, and sequencing the individuals from small to large according to the mixed fitness function value.
S312, build scale N A Selecting the individuals with the lowest fitness function values for archiving.
S313, selecting a microphone array leading array element according to a deep hunting group scheme, and generating a deep searching group array element group nearby the microphone array leading array element.
Optionally, the output module is further configured to:
s321, carrying out iterative optimization on the positions and the array element excitation weights of the traditional group array element groups of the microphone array through surrounding the prey, high-level spiral shrinkage surrounding and high-level searching the prey, and obtaining updated traditional group array element groups of the microphone array.
S322, detecting whether an individual exceeding the aperture of the microphone array exists in the updated traditional array element group of the microphone array, and if so, randomly generating a new individual to replace the exceeding individual in the aperture of the microphone array to generate a new traditional array element group of the microphone array.
Optionally, the output module is further configured to:
according to the deep prey group scheme, increasing or decreasing the number of individuals of the new traditional group element group of the microphone array and the new deep search group element group of the microphone array by N 0 The total array element group is composed of a new traditional array element group of the microphone array and a new deep search array element group of the microphone array.
In one aspect, an electronic device is provided, the electronic device includes a processor and a memory, the memory stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the above-mentioned sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization method.
In one aspect, a computer readable storage medium having stored therein at least one instruction loaded and executed by a processor to implement the above-described sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization method is provided.
Compared with the prior art, the technical scheme has at least the following beneficial effects:
according to the scheme, the microphone array is optimized by the mixed fitness function, and the microphone array with the optimized design can meet the two functional requirements of sound source positioning and sound field reconstruction at the same time, so that a scheme capable of being used as a reference is provided for the design of the multifunctional microphone array.
According to the invention, the improved Sine chaotic mapping is introduced to randomly generate the array element positions of the three-dimensional sparse microphone array, so that the primary array element groups are uniformly distributed in space, the chaos characteristic is more obvious, and the subsequent optimization convergence speed is accelerated; introducing a whale elite optimization algorithm, weakening the influence of the previous generation population in the optimization process of the array element position and the array element excitation weight, expanding the array element search range, improving the development and detection capability, getting rid of local optimization in the later stage, and iteratively calculating the array element position and the array element excitation weight through faster local convergence speed, higher convergence precision and lower calculation complexity to obtain the three-dimensional sparse microphone array with higher adaptability to the beam forming and the harmony hologram algorithm.
The method is suitable for optimizing the three-dimensional microphone array, and the traditional optimization method is mainly aimed at two-dimensional array design, so that the traditional two-dimensional microphone array is difficult to capture time, frequency and space distribution information of a complex sound field in a refined manner. According to the invention, the high-performance three-dimensional microphone array with a wide dynamic response range is obtained by optimizing the array element positions and the excitation weights of the three-dimensional microphone array, so that the problem that the spatial random distribution of noise sources is difficult to accurately measure is solved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic flow diagram of a sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization method provided by an embodiment of the invention;
fig. 2 is a schematic structural diagram of a three-dimensional microphone array optimization method for sound field measurement multi-objective cooperation provided by the embodiment of the invention;
FIG. 3 is a diagram of an optimized three-dimensional microphone array versus multiple sound source signal measurement arrangement provided by an embodiment of the present invention;
fig. 4 is a block diagram of a sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization device provided by an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. All other embodiments, which can be made by a person skilled in the art without creative efforts, based on the described embodiments of the present invention fall within the protection scope of the present invention.
As shown in fig. 1, the embodiment of the invention provides a sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization method, which can be implemented by electronic equipment. The process flow of the method for optimizing the sound field measurement-oriented multi-objective collaborative three-dimensional microphone array as shown in fig. 1 can comprise the following steps:
s1, acquiring geometric parameters of a traditional array element group of the microphone array, and generating a primary array element group of the microphone array through improved Sine chaotic mapping.
Optionally, the step S1 may include the following steps S11 to S13:
s11, acquiring geometric parameters of the traditional group array element group of the microphone array.
The geometric parameters may include the number N of array elements, the initial size of the conventional array element group, the microphone array aperture, the lower limit of the array element spacing constraint threshold, the upper limit of the array element spacing constraint threshold, and the like.
In a possible implementation manner, the flow of the invention is shown in fig. 2, wherein, as shown in fig. 3, the three-dimensional microphone array is two orthogonally distributed planar microphone arrays, 28 electret free field microphones are distributed on each planar array, and the multichannel synchronous acquisition of the microphone arrays is completed by utilizing a data acquisition system based on an NI-PXIe bus, and the sampling frequency is 44100Hz.
The sparse initial microphone array position is generated by adopting improved Sine chaotic mapping and is used as a microphone array primary array element group, so that the array element positions are uniformly distributed in space.
S12, setting control parameters and initial parameters of the improved Sine chaotic map.
S13, generating a sparse initial microphone array position according to the geometric parameters and the improved fine chaotic mapping, and obtaining a microphone array initial array element group.
In a possible implementation manner, according to the number N of array elements and the initial scale N of the traditional group array element group of the microphone array P0 The aperture D of the microphone array and the array element spacingLower constraint threshold d min Upper limit d of array element spacing constraint threshold max The control parameter μ and the initial parameter d (1) =rand, e (1) =rand of the improved fine chaotic map are set, and rand is a random real number generating [0,1 ].
Generating initial position of microphone array by adopting improved Sine chaotic mapping, namely initial position X of traditional array element group of microphone array traditional ＝{x 1 ,x 2 ,…,x NP0 }
S2, constructing a mixed fitness function.
Optionally, the step S2 may include the following steps S21 to S23:
s21, constructing a peak sidelobe level PSLL as an optimization target for evaluating the suitability of the microphone array for a beam forming algorithm in sound source positioning.
S22, constructing a peak signal-to-noise ratio PSNR as an optimization target for evaluating the suitability of the microphone array to the acoustic holographic algorithm in sound field reconstruction.
S23, constructing a mixing function according to the sidelobe level and the peak signal-to-noise ratio, and further constructing a mixing fitness function.
In a feasible implementation mode, the beam forming algorithm and the acoustic holographic algorithm reach the best effect by carrying out multi-objective optimization on the array element positions and the array element excitation weights of the microphone array in face of different requirements of sound source positioning and sound field reconstruction on the measurement microphone array, and a mixed fitness function of a side lobe level and a peak signal to noise ratio is constructed.
Specifically, to evaluate the suitability of a microphone array for a beam forming algorithm in sound source localization, a microphone array response function is first generated. If the microphone array is focused to any point x in space, the response function of the microphone array can be written as the following equation (1):
Figure BDA0004154991280000101
in which x is s For the sound source position, the sound source coordinates are (x, y, z), and the microphone array center coordinates are (x 0 ,y 0 ,z 0 ) M channelMicrophone coordinate is (x m ,y m ,z m )，r 0 And r m The distances of the sound source to the center of the microphone array and the m-channel microphones,
Figure BDA0004154991280000102
Figure BDA0004154991280000103
r is the distance from the sound source to any point in space, w m Is the corresponding weighting factor of the M channel microphone, which can be used to adjust the loudness of the microphone array, k is the wave number, j is the complex unit, and M is the number of channels. Relative to the source point x 0 The loudness of the microphone array for any point focus is the following equation (2):
Figure BDA0004154991280000104
taking the ratio of the side lobe level to the main lobe level (i.e., peak side lobe level psll=max (dB (x)) as an optimization target, the smaller the PSLL value, the higher the applicability of the microphone array to the beamforming algorithm.
Further, in order to evaluate the suitability of the microphone array for the acoustic holographic algorithm in the sound field reconstruction, the larger the peak signal-to-noise ratio (PSNR) is, the smaller the absolute error of the reconstruction is, which means that the suitability of the microphone array for the acoustic holographic algorithm is higher, and the mathematical expression is as follows (3):
Figure BDA0004154991280000111
wherein MAX I Is the maximum value of the sound field in the region, F rec Is a reconstructed sound field, RMSE (F, F rec ) Is the root mean square error of the original sound field and the reconstructed sound field.
In order to simultaneously optimize the applicability of the microphone array to beam forming and acoustic hologram algorithms, a mixed function of sidelobe levels and peak signal to noise ratios is used as an optimization target, as shown in the following formula (4):
Figure BDA0004154991280000112
wherein alpha is 1 And alpha 2 For sidelobe level and peak signal to noise ratio weights, alpha 1 ,α 2 E (0, 1), w represents an iteration sequence value of the improved fine chaotic map, and d represents an initial parameter of the improved fine chaotic map.
The lower the fitness value, the more excellent the microphone array performance. Influencing the main lobe width BW -3dB The main factors of the method are microphone array aperture, meanwhile, the reasonable arrangement problem of the microphone array aperture and array element spacing is considered, and the mixing fitness function is constructed as shown in the following formula (5):
Figure BDA0004154991280000113
wherein fitness represents a mixing function, BW -3dB Representing the main lobe width, BW thr Represents the main lobe width constraint threshold, d ij Represents the spacing of any two array elements, i, j=1, 2,3 … N P ，N P The traditional array element group size of the microphone array is set; d, d min Representing the lower limit of the array element spacing constraint threshold; d, d max Representing the upper limit of the array element spacing constraint threshold.
And S3, obtaining the optimized three-dimensional microphone array for multi-objective coordination of sound field measurement according to the microphone array primary array element group, the mixed fitness function and the whale elite optimization algorithm.
Optionally, the step S3 may include the following steps S31 to S36:
s31, according to the primary array element group of the microphone array, an excellent array element file with the lowest value of the mixed fitness function is established, and a leading array element of the microphone array is selected to generate a deep search group array element group of the microphone array.
Optionally, the step S31 may include the following steps S311 to S313:
s311, calculating a mixed fitness function value of each individual in the traditional array element group of the microphone array, and sequencing the individuals from small to large according to the mixed fitness function value.
In one possible embodiment, the WEOA (Whale Elite Optimization Algorithm ) divides the search array element group into two parts, one part being the traditional group array element group X of the microphone array traditional Iterative optimization by encircling the prey, high-level spiral shrinkage encircling and high-level search of the prey, and deep search of group element group X of microphone array deep Iterative optimization will be performed according to the deep hunter method.
Calculating the primary array element group X of the traditional group traditional ＝{x 1 ,x 2 ,…,x NP Fitness function fitness and order the values from small to large.
S312, build scale N A Selecting the individuals with the lowest fitness function values for archiving.
In one possible implementation, an optimal individual is established for the archival record, and is used to update the location of the new individual. File size N A . According to the traditional group of primary array element groups X traditional The mixed fitness function fitness of the system is changed from small to large, and the individual with the lowest value of the current mixed fitness function is selected for archiving.
S313, selecting a microphone array leading array element according to a deep hunting group scheme, and generating a deep searching group array element group nearby the microphone array leading array element.
In a possible embodiment, the leading element X is selected according to a deep prey group scheme leader And generating deep search group array element group X near the leading array element deep To speed up local searches. Setting deep search group array element group X deep Number N F . Wherein N is F +N P ＝N。
The deep prey group adjustment principle is utilized to form the traditional group element group X traditional And deep search group element group X deep The size of the array element group is increased or reduced based on the array element fitness value 0 The total array element group remains unchanged in units. To avoid X traditional Too small to ensure array element group diversity, X deep Upper limit of (2) is set to NF max And initial X deep Is NF (NF) 0 。
Group element group X is searched according to the first generation deep layer deep Fitness function fitness update archive records and record first generation best individuals X * (1)。
S32, updating the positions and the array element excitation weights of the traditional array element groups of the microphone array through a whale elite optimization algorithm, and generating a new traditional array element group of the microphone array.
Optionally, the step S32 may include the following steps S321 to S322:
s321, carrying out iterative optimization on the positions and the array element excitation weights of the traditional group array element groups of the microphone array through surrounding the prey, high-level spiral shrinkage surrounding and high-level searching the prey, and obtaining updated traditional group array element groups of the microphone array.
S322, detecting whether an individual exceeding the aperture of the microphone array exists in the updated traditional array element group of the microphone array, and if so, randomly generating a new individual to replace the exceeding individual in the aperture of the microphone array to generate a new traditional array element group of the microphone array.
In one possible embodiment, the group X of conventional array elements is surrounded by three parts, namely surrounding the prey, high-level helical shrinkage surrounding and high-level search for the prey traditional And performing iterative optimization of the position and the excitation weight. Group X of traditional group array elements traditional And (3) finishing the position updating, detecting whether an individual exceeds the aperture, and randomly generating a new individual replacement in the aperture if the individual exceeding the aperture exists.
Further, the traditional group element group X of the generation is calculated traditional Is ordered after the fitness function fitness and updates the records.
Further, a new leading array element X is selected leader And generating deep search group array element group X deep Group X is searched according to deep layer deep The fitness function fitness of the file records is updated again; based on the size of the deep hunting group array element groupAdjusting strategy for updating traditional group array element group X traditional And deep search group element group X deep Is the number of (3); recording this generation of optimal individuals X * (t)。
S33, updating the excellent array element file according to the new traditional array element group of the microphone array, selecting a new leading array element of the microphone array, and generating a new deep search array element group of the microphone array.
S34, judging whether the array element group size of the new microphone array deep search array element group reaches the maximum value of the preset array element group size, if so, adjusting the number of individuals of the new microphone array traditional array element group and the new microphone array deep search array element group, and executing the step S35; if not, step S35 is performed.
Optionally, adjusting the new microphone array legacy group element group and the new microphone array deep search group element group individual number in S34 includes:
according to the deep prey group scheme, increasing or decreasing the number of individuals of the new traditional group element group of the microphone array and the new deep search group element group of the microphone array by N 0 The total array element group is composed of a new traditional array element group of the microphone array and a new deep search array element group of the microphone array.
S35, updating the excellent array element file according to the traditional array element group of the adjusted microphone array, and updating the deep search array element group.
S36, judging whether the preset iteration times are reached, if so, obtaining the position of the optimal array element group and the excitation weight of the array element, and obtaining the optimized three-dimensional microphone array for the multi-objective coordination of sound field measurement; if not, go to execute step S32.
In a possible implementation manner, the microphone array element group is circularly iterated to finally obtain the optimal element group position distribution and the element excitation weight, and the final element group X is obtained by outputting f And final optimization result P f 。
Specifically, a maximum number of iterations T is set max Repeating the iteration to the position of the array element group and the excitation of the array elementOptimizing the excitation weight, and recording the optimal individual X of each generation * (t) finally, obtaining the optimal array element group position distribution and the array element excitation weight, and outputting to obtain the final array element group X f And final optimization result P f And obtaining the sparse sensor three-dimensional microphone array with high adaptability to beam forming and harmony hologram algorithm.
Firstly, determining a microphone capable of meeting the sound pressure level and the frequency response range according to the sound pressure level requirement of a sound signal; then, a three-dimensional sparse microphone array is built, a multi-channel data acquisition device of NI-PXIe is connected through a BNC cable, and data acquisition and storage of multi-channel acoustic signals are realized through labview software.
Then, generating a sparse microphone array by improving the fine chaotic mapping, constructing a mixed fitness function as a primary array element group of the microphone array, and calculating an individual fitness function of a traditional array element group of the microphone array; establishing an excellent array element file with the lowest value of the fitness function, selecting a leading array element, generating a microphone array deep search group array element group and updating the file; updating the traditional group of array elements by using a whale elite optimization algorithm to generate a next generation of array element group; updating the deep search group array element group according to the new leading array element, adjusting the quantity of the traditional group array element group and the deep search group array element group, and recording the optimal individual of the first generation array element group by archives; setting the maximum iteration times to perform iterative updating on the array element group positions to obtain the final optimal microphone array element group, and obtaining the optimized microphone array positions and the excitation weights of the microphones.
And drawing a reconstruction cloud image according to the solved sound pressure matrix P, so as to determine the position of the noise source. Meanwhile, mathScript program modules can be added into labview acquisition software to realize real-time calculation of acquired data, so that dynamic positioning of sound source positions is achieved.
In this embodiment, two planar orthogonal sparse microphone arrays are used to form a three-dimensional microphone array, and sound signal measurement is performed on two sound sources in space. The aperture of the microphone array is 0.5m, 28 electret free field microphones are distributed on the microphone array, the multichannel synchronous acquisition of the microphone array is completed by utilizing a data acquisition system based on an NI-PXIe bus, and the sampling frequency is 44100Hz.
S1, generating a sparse initial microphone array position by adopting improved Sine chaotic mapping, and taking the sparse initial microphone array position as a microphone array initial array element group.
Specifically, according to the number N of array elements and the initial scale N of the traditional array element group of the microphone array P0 The microphone array aperture D and the array element spacing constraint threshold value lower limit D min Upper limit d of array element spacing constraint threshold max The control parameter μ and the initial parameter d (1) =rand, e (1) =rand of the improved fine chaotic map are set, and rand is a random real number generating [0,1 ].
Generating initial position X of microphone array by adopting improved Sine chaotic mapping, namely initial position X of traditional group array elements traditional ＝{x 1 ,x 2 ,…,x NP }. The improved Sine chaotic mapping formula comprises the following steps:
Figure BDA0004154991280000151
wherein μ and w are control parameters and iteration sequence values of the improved one-dimensional fine chaotic map, respectively.
S2, constructing a mixed fitness function, calculating individual fitness functions of the array element groups, and sequencing the function values from small to large.
Specifically, to evaluate the suitability of a microphone array for a beam forming algorithm in sound source localization, a microphone array response function is first generated. If the microphone array is focused to any point x in space, the response function of the microphone array can be written as:
Figure BDA0004154991280000152
in which x is s For the sound source position, the sound source coordinates are (x, y, z), and the microphone array center coordinates are (x 0 ,y 0 ,z 0 ) The m-channel microphone coordinates are (x m ,y m ,z m )，r 0 And r m The distances of the sound source to the center of the microphone array and the m-channel microphones,
Figure BDA0004154991280000153
Figure BDA0004154991280000154
r is the distance from the sound source to any point in space, w m Is the corresponding weighting factor of the M channel microphone, which can be used to adjust the loudness of the microphone array, k is the wave number, j is the complex unit, and M is the number of channels. Relative to the source point x 0 The loudness of the microphone array for any point focus is:
Figure BDA0004154991280000155
taking the ratio of the side lobe level to the main lobe level (i.e., peak side lobe level psll=max (dB (x)) as an optimization target, the smaller the PSLL value, the higher the applicability of the microphone array to the beamforming algorithm.
In order to evaluate the suitability of the microphone array for the acoustic holographic algorithm in the sound field reconstruction, the larger the peak signal-to-noise ratio (PSNR) is, the smaller the absolute error of the reconstruction is, which means that the higher the suitability of the microphone array for the acoustic holographic algorithm is, and the mathematical expression is as follows:
Figure BDA0004154991280000161
wherein MAX I Is the maximum value of the sound field in the region, F rec Is a reconstructed sound field, RMSE (F, F rec ) Is the root mean square error of the original sound field and the reconstructed sound field.
In order to simultaneously optimize the applicability of the microphone array to beam forming and acoustic hologram algorithms, a mixed fitness function of sidelobe levels and peak signal-to-noise ratios is constructed:
Figure BDA0004154991280000162
the lower the fitness value, the more excellent the microphone array performance.Wherein alpha is 1 And alpha 2 For sidelobe level and peak signal to noise ratio weights, alpha 1 ,α 2 E (0, 1). Influencing the main lobe width BW -3dB The main factors of the method are microphone array aperture, and meanwhile, the reasonable arrangement problem of the microphone array aperture and the array element distance is considered, and the mixing fitness function is constructed as follows:
Figure BDA0004154991280000163
wherein BW is thr Representing a main lobe width constraint threshold; d, d ij Representing the distance between any two array elements; d, d min Representing the lower limit of the array element spacing constraint threshold; d, d max Representing the upper limit of the array element spacing constraint threshold; i, j=1, 2,3 … N P (N P For the traditional group size).
S3, establishing excellent array element files records with lowest fitness function value, and selecting leading whales X leader Generating a microphone array deep search group array element group X deep And updates the file.
Specifically, the whale elite optimization algorithm divides the search array element group into two parts, wherein one part is the traditional group array element group X of the microphone array traditional Iterative optimization by encircling the prey, high-level spiral shrinkage encircling and high-level search of the prey, and deep search of group element group X of microphone array deep Iterative optimization will be performed according to the deep hunter method.
Calculating the primary array element group X of the traditional group traditional ＝{x 1 ,x 2 ,…,x NP0 Fitness function fitness and order the values from small to large.
And establishing an optimal individual of the file record history, and applying the optimal individual to updating the position of a new individual. File size N A . According to the traditional group of primary array element groups X traditional The mixed fitness function fitness of the system is changed from small to large, and the individual with the lowest value of the current mixed fitness function is selected for archiving.
Selecting leading array element X according to deep hunting group scheme leader And is combined withGenerating deep search group element group X near the leading element deep To speed up local searches. Setting deep search group array element group X deep Number N F . Wherein N is F +N P =n. The specific method comprises the following steps:
(a) Deep hunt group location updating method.
First, the most excellent individuals of the current generation are selected as leading array elements, and the expression is:
X leader (t)＝X * (t) (12)
Wherein X is * And (t) is the position of the current optimal array element.
Then, several deep search set elements are generated near the leading element to accelerate the local search:
X p (t)X leader (t)+k(m-0.5)(X leader (t)-X records (t)) (13)
where p=1, 2, …, NF, m is a random number between 0 and 1, and k is a constant.
(b) And (5) a strategy for adjusting the size of array elements of the deep hunting group.
In order to guarantee the usability of the proposed deep search group, an adjustment principle is used. By this method, the conventional array element group X traditional ＝{x 1 ,x 2 ,…,x NP Matrix element group X is searched for in } and deep deep ＝{x 1 ,x 2 ,…,x NF Increasing or decreasing the size of the array element group based on the member fitness value thereof 0 The total array element group remains unchanged in units. To avoid X traditional Too small to ensure array element group diversity, X deep Upper limit of (2) is set to NF max And initial X deep Is NF (NF) 0 。
Group element group X is searched according to the first generation deep layer deep Fitness function fitness update archive records and record first generation best individuals X * (1)。
S4, updating the positions of the traditional group of array elements and the excitation weights of the array elements of the microphone array through a whale elite optimization algorithm to generate a next generation array element group; and updating the microphone array deep search group array element group according to the new microphone array leading array element, adjusting the traditional group array element group and the deep search group array element group number of the microphone array, and recording the optimal individual of the first generation array element group by archives.
Specifically, the conventional group element group X is formed by basically encircling the prey, high-level spiral shrinkage encircling and high-level searching the prey traditional And performing iterative optimization of the position and the excitation weight. The specific method comprises the following steps:
(a) Surrounding and surrounding the prey:
assuming that the current leading array element is positioned as X * (t) the position of the individual array element is X (t), the next position of the individual array element X (t) under the influence of the leading array element is X (t+1), then:
X k (t+1)＝X * (t)-A·D,p<0.5 and |A|<1 (14)
D＝C·X * (t)-X k (t),k＝1,2,…,NP (15)
Wherein: t is the current iteration number; x is X k (t) is a position vector of an array element individual; d is the distance between the array element individual and the prey; multiplication element by element; a and C are coefficient vectors for controlling the motion mode of the array element, and are:
C＝2r,A＝a(2r-1) (16)
a＝2-2t/T max (t＝1,2,…,T max ) (17)
where r is a random number between 0 and 1, it is apparent that a gradually decreases from 2 to 0 with the search process.
(b) Advanced helical contraction surrounds:
X k (t+1)＝w·X * (t)＝D′e bl′ cos(2πl′) (18)
D′＝X * (t)-X t (t),k＝1,2,…,NP (19)
w＝w 1 -(w 1 -w 2 )×(t/T max ) 1/t (20)
l′＝2i/NP(i＝1,2,…,NP) (21)
to better exploit this mechanism, the parameter l is replaced by l 'to ensure that individuals with better fitness values correspond to smaller l', and vice versa. Furthermore, the nonlinear variable w derives from itMaximum value w 1 The minimum w is less than or equal to 0 2 <w 1 And is less than or equal to 1. Therefore, with the development of the optimization process, the influence of the previous generation is weakened, and the search range is expanded, so that the local optimization is eliminated in the later stage.
(c) Advanced search prey:
X k (t+1)＝X records (t)′-A·D records ′,p<0.5 and |A| is not less than 1 (22)
D records ′＝C·X records (t)′-X k (t),k＝1,2,…,NP (23)
Wherein X is records (t)' is selected from the best individuals in the profile, making the exploration search process more purposeful and efficient.
Group X of traditional group array elements traditional And (3) finishing the position updating, detecting whether an individual exceeds the aperture, and randomly generating a new individual replacement in the aperture if the individual exceeding the aperture exists.
And updating the deep search group array element group according to the new leader, adjusting the quantity of the traditional group array element group and the deep search group array element group, and recording the optimal individual of the generation array element group by the file.
According to the step S2, the traditional group element group X of the generation is calculated traditional Is ordered after the fitness function fitness and updates the records.
According to the step S3, selecting the individual with the minimum new fitness function fitness as the leading array element X leader And generating deep search group array element group X deep Calculating deep search group array element group X deep The fitness function fitness of the file records is updated again; updating the traditional group array element group X according to the deep hunting group array element group scale adjustment strategy traditional And deep search group element group X deep Is the number of (3); recording this generation of optimal individuals X * (t)。
And S5, repeating the iteration S4 according to the maximum iteration times, and carrying out iterative updating on the array element group of the microphone array to obtain the individual position and the excitation weight of the optimal array element group, thereby obtaining the optimized microphone array.
Specifically, a maximum number of iterations T is set max Repeating iteration S4, optimizing the array element group position and the array element excitation weight, and recording each generation of optimal individual X * (t) finally, obtaining the optimal array element group position distribution and the array element excitation weight, and outputting to obtain the final array element group X f And final optimization result P f And obtaining the sparse sensor three-dimensional microphone array with high adaptability to beam forming and harmony hologram algorithm.
In the embodiment of the invention, the microphone array is optimized by providing the mixed fitness function, and the microphone array with the optimized design can simultaneously meet the two functional requirements of sound source positioning and sound field reconstruction, thereby providing a borrowable scheme for the design of the multifunctional microphone array.
According to the invention, the improved Sine chaotic mapping is introduced to randomly generate the array element positions of the three-dimensional sparse microphone array, so that the primary array element groups are uniformly distributed in space, the chaos characteristic is more obvious, and the subsequent optimization convergence speed is accelerated; introducing a whale elite optimization algorithm, weakening the influence of the previous generation population in the optimization process of the array element position and the array element excitation weight, expanding the array element search range, improving the development and detection capability, getting rid of local optimization in the later stage, and iteratively calculating the array element position and the array element excitation weight through faster local convergence speed, higher convergence precision and lower calculation complexity to obtain the three-dimensional sparse microphone array with higher adaptability to the beam forming and the harmony hologram algorithm.
The method is suitable for optimizing the three-dimensional microphone array, and the traditional optimization method is mainly aimed at two-dimensional array design, so that the traditional two-dimensional microphone array is difficult to capture time, frequency and space distribution information of a complex sound field in a refined manner. According to the invention, the high-performance three-dimensional microphone array with a wide dynamic response range is obtained by optimizing the array element positions and the excitation weights of the three-dimensional microphone array, so that the problem that the spatial random distribution of noise sources is difficult to accurately measure is solved.
As shown in fig. 4, an embodiment of the present invention provides a sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization apparatus 400, where the apparatus 400 is applied to implement a sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization method, and the apparatus 400 includes:
the obtaining module 410 is configured to obtain geometric parameters of a traditional array element group of the microphone array, and generate a primary array element group of the microphone array through improved fine chaotic mapping.
The function construction module 420 is configured to construct a hybrid fitness function.
And the output module 430 is configured to obtain an optimized three-dimensional microphone array for sound field measurement multi-objective coordination according to the microphone array primary array element group, the mixed fitness function and the whale elite optimization algorithm.
Optionally, the obtaining module 410 is further configured to:
s11, acquiring geometric parameters of a traditional group array element group of the microphone array; the geometric parameters comprise the number of array elements, the size of a traditional array element group, the aperture of a microphone array, the lower limit of an array element spacing constraint threshold and the upper limit of an array element spacing constraint threshold.
S12, setting control parameters and initial parameters of the improved Sine chaotic map.
S13, generating a sparse initial microphone array position according to the geometric parameters and the improved fine chaotic mapping, and obtaining a microphone array initial array element group.
Optionally, the function construction module 420 is further configured to:
s21, constructing a peak sidelobe level PSLL as an optimization target for evaluating the suitability of the microphone array for a beam forming algorithm in sound source positioning.
S22, constructing a peak signal-to-noise ratio PSNR as an optimization target for evaluating the suitability of the microphone array to the acoustic holographic algorithm in sound field reconstruction.
S23, constructing a mixing function according to the sidelobe level and the peak signal-to-noise ratio, and further constructing a mixing fitness function.
Optionally, the mixing function is represented by the following formula (1):
Figure BDA0004154991280000201
wherein alpha is 1 ,α 2 ∈(0,1)，w represents an iteration sequence value of the improved fine chaotic map, and d represents an initial parameter of the improved fine chaotic map.
Optionally, the fitness function is mixed as shown in the following formula (2):
Figure BDA0004154991280000202
wherein fitness represents a mixing function, BW -3dB Representing the main lobe width, BW thr Represents the main lobe width constraint threshold, d ij Represents the spacing of any two array elements, i, j=1, 2,3 … N P ，N P The traditional array element group size of the microphone array is set; d, d min Representing the lower limit of the array element spacing constraint threshold; d, d max Representing the upper limit of the array element spacing constraint threshold.
Optionally, the output module 430 is further configured to:
s31, according to the primary array element group of the microphone array, an excellent array element file with the lowest value of the mixed fitness function is established, and a leading array element of the microphone array is selected to generate a deep search group array element group of the microphone array.
S32, updating the positions and the array element excitation weights of the traditional array element groups of the microphone array through a whale elite optimization algorithm, and generating a new traditional array element group of the microphone array.
S33, updating the excellent array element file according to the new traditional array element group of the microphone array, selecting a new leading array element of the microphone array, and generating a new deep search array element group of the microphone array.
S34, judging whether the array element group size of the new microphone array deep search array element group reaches the maximum value of the preset array element group size, if so, adjusting the number of individuals of the new microphone array traditional array element group and the new microphone array deep search array element group, and executing the step S35; if not, step S35 is performed.
S35, updating the excellent array element file according to the traditional array element group of the adjusted microphone array, and updating the deep search array element group.
S36, judging whether the preset iteration times are reached, if so, obtaining the position of the optimal array element group and the excitation weight of the array element, and obtaining the optimized three-dimensional microphone array for the multi-objective coordination of sound field measurement; if not, go to execute step S32.
Optionally, the output module 430 is further configured to:
s311, calculating a mixed fitness function value of each individual in the traditional array element group of the microphone array, and sequencing the individuals from small to large according to the mixed fitness function value.
S312, build scale N A Selecting the individuals with the lowest fitness function values for archiving.
S313, selecting a microphone array leading array element according to a deep hunting group scheme, and generating a deep searching group array element group nearby the microphone array leading array element.
Optionally, the output module is further configured to:
s321, carrying out iterative optimization on the positions and the array element excitation weights of the traditional group array element groups of the microphone array through surrounding the prey, high-level spiral shrinkage surrounding and high-level searching the prey, and obtaining updated traditional group array element groups of the microphone array.
S322, detecting whether an individual exceeding the aperture of the microphone array exists in the updated traditional array element group of the microphone array, and if so, randomly generating a new individual to replace the exceeding individual in the aperture of the microphone array to generate a new traditional array element group of the microphone array.
Optionally, the output module 430 is further configured to:
according to the deep prey group scheme, increasing or decreasing the number of individuals of the new traditional group element group of the microphone array and the new deep search group element group of the microphone array by N 0 The total array element group is composed of a new traditional array element group of the microphone array and a new deep search array element group of the microphone array.
In the embodiment of the invention, the microphone array is optimized by providing the mixed fitness function, and the microphone array with the optimized design can simultaneously meet the two functional requirements of sound source positioning and sound field reconstruction, thereby providing a borrowable scheme for the design of the multifunctional microphone array.
According to the invention, the improved Sine chaotic mapping is introduced to randomly generate the array element positions of the three-dimensional sparse microphone array, so that the primary array element groups are uniformly distributed in space, the chaos characteristic is more obvious, and the subsequent optimization convergence speed is accelerated; introducing a whale elite optimization algorithm, weakening the influence of the previous generation population in the optimization process of the array element position and the array element excitation weight, expanding the array element search range, improving the development and detection capability, getting rid of local optimization in the later stage, and iteratively calculating the array element position and the array element excitation weight through faster local convergence speed, higher convergence precision and lower calculation complexity to obtain the three-dimensional sparse microphone array with higher adaptability to the beam forming and the harmony hologram algorithm.
The method is suitable for optimizing the three-dimensional microphone array, and the traditional optimization method is mainly aimed at two-dimensional array design, so that the traditional two-dimensional microphone array is difficult to capture time, frequency and space distribution information of a complex sound field in a refined manner. According to the invention, the high-performance three-dimensional microphone array with a wide dynamic response range is obtained by optimizing the array element positions and the excitation weights of the three-dimensional microphone array, so that the problem that the spatial random distribution of noise sources is difficult to accurately measure is solved.
Fig. 5 is a schematic structural diagram of an electronic device 500 according to an embodiment of the present invention, where the electronic device 500 may have a relatively large difference due to different configurations or performances, and may include one or more processors (central processing units, CPU) 501 and one or more memories 502, where at least one instruction is stored in the memories 502, and the at least one instruction is loaded and executed by the processors 501 to implement the following method for optimizing a sound field measurement multi-objective collaborative three-dimensional microphone array:
s1, acquiring geometric parameters of a traditional array element group of the microphone array, and generating a primary array element group of the microphone array through improved Sine chaotic mapping.
S2, constructing a mixed fitness function.
And S3, obtaining the optimized three-dimensional microphone array for multi-objective coordination of sound field measurement according to the microphone array primary array element group, the mixed fitness function and the whale elite optimization algorithm.
In an exemplary embodiment, a computer readable storage medium, such as a memory comprising instructions executable by a processor in a terminal to perform the above-described sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization method is also provided. For example, the computer readable storage medium may be ROM, random Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program for instructing relevant hardware, where the program may be stored in a computer readable storage medium, and the storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.
The foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and scope of the invention are intended to be included within the scope of the invention.

Claims (10)

1. A sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization method, the method comprising:
s1, acquiring geometric parameters of a traditional array element group of a microphone array, and generating a primary array element group of the microphone array through improved Sine chaotic mapping;
s2, constructing a mixed fitness function;
and S3, obtaining the optimized three-dimensional microphone array for multi-objective coordination of sound field measurement according to the microphone array primary array element group, the mixed fitness function and the whale elite optimization algorithm.
2. The method according to claim 1, wherein the step of obtaining the geometric parameters of the traditional group of elements of the microphone array in S1, and generating the primary group of elements of the microphone array by improved fine chaotic mapping, includes:
s11, acquiring geometric parameters of a traditional group array element group of the microphone array; the geometric parameters comprise the number of array elements, the initial scale of the traditional array element group, the aperture of the microphone array, the lower limit of the constraint threshold of the array element spacing and the upper limit of the constraint threshold of the array element spacing;
s12, setting control parameters and initial parameters of improved Sine chaotic mapping;
s13, generating a sparse initial microphone array position according to the geometric parameters and the improved fine chaotic mapping, and obtaining a microphone array initial array element group.
3. The method of claim 1, wherein constructing the hybrid fitness function in S2 comprises:
s21, constructing a peak sidelobe level PSLL as an optimization target for evaluating the suitability of the microphone array for a beam forming algorithm in sound source positioning;
s22, constructing a peak signal-to-noise ratio PSNR as an optimization target for evaluating the suitability of the microphone array to the acoustic holographic algorithm in the sound field reconstruction;
s23, constructing a mixing function according to the sidelobe level and the peak signal-to-noise ratio, and further constructing a mixing fitness function.
4. A method according to claim 3, wherein the mixing function is represented by the following formula (1):
Figure FDA0004154991260000021
wherein alpha is 1 ,α 2 E (0, 1), w represents an iteration sequence value of the improved fine chaotic map, and d represents an initial parameter of the improved fine chaotic map.
5. The method of claim 1, wherein the hybrid fitness function is represented by the following formula (2):
Figure FDA0004154991260000022
wherein fitness represents a mixing function, BW -3dB Representing the main lobe width, BW thr Represents the main lobe width constraint threshold, d ij Represents the spacing of any two array elements, i, j=1, 2,3 … N P ，N P The traditional array element group size of the microphone array is set; d, d min Representing the lower limit of the array element spacing constraint threshold; d, d max Representing the upper limit of the array element spacing constraint threshold.
6. The method according to claim 1, wherein the obtaining the optimized sound field measurement-oriented multi-objective collaborative three-dimensional microphone array in S3 according to the microphone array primary array element group, the mixed fitness function and the whale elite optimization algorithm comprises:
s31, according to the primary array element group of the microphone array, an excellent array element file with the lowest value of the mixed fitness function is established, and a leading array element of the microphone array is selected to generate a deep search group array element group of the microphone array;
s32, updating the positions and the array element excitation weights of the traditional array element groups of the microphone array through a whale elite optimization algorithm to generate a new traditional array element group of the microphone array;
s33, updating an excellent array element file according to the new traditional array element group of the microphone array, selecting a new leading array element of the microphone array, and generating a new deep search array element group of the microphone array;
s34, judging whether the array element group size of the new microphone array deep search array element group reaches the maximum value of the preset array element group size, if so, adjusting the number of individuals of the new microphone array traditional array element group and the new microphone array deep search array element group, and executing the step S35; if not, executing step S35;
S35, updating the excellent array element files according to the traditional array element groups of the adjusted microphone array, and updating the deep search array element groups;
s36, judging whether the preset iteration times are reached, if so, obtaining the position of the optimal array element group and the excitation weight of the array element, and obtaining the optimized three-dimensional microphone array for the multi-objective coordination of sound field measurement; if not, go to execute step S32.
7. The method of claim 6, wherein the creating an excellent array element file with a lowest mixing fitness function value according to the microphone array primary array element group in S31, and selecting a microphone array leading array element, and generating a microphone array deep search group array element group includes:
s311, calculating a mixed fitness function value of each individual in the traditional array element group of the microphone array, and sequencing the individuals from small to large according to the mixed fitness function value;
s312, build scale N A Selecting the individuals with the lowest current generation fitness function values for archiving;
s313, selecting a microphone array leading array element according to a deep hunting group scheme, and generating a deep searching group array element group nearby the microphone array leading array element.
8. The method of claim 6, wherein updating the positions and the element activation weights of the traditional group of microphone arrays by a whale elite optimization algorithm in S32 generates a new traditional group of microphone arrays, comprising:
s321, carrying out iterative optimization on the positions and the array element excitation weights of the traditional group array element groups of the microphone array through surrounding and surrounding the prey, high-level spiral shrinkage and surrounding and high-level searching the prey to obtain updated traditional group array element groups of the microphone array;
s322, detecting whether an individual exceeding the aperture of the microphone array exists in the updated traditional array element group of the microphone array, and if so, randomly generating a new individual to replace the exceeding individual in the aperture of the microphone array to generate a new traditional array element group of the microphone array.
9. The method of claim 6, wherein the adjusting the individual numbers of the new microphone array legacy group element group and the new microphone array deep search group element group in S34 comprises:
increasing or decreasing the number of individuals of the new microphone array traditional group element group and the new microphone array deep search group element group by N according to a deep prey group scheme 0 And the total array element group is formed by the new traditional array element group of the microphone array and the new deep search array element group of the microphone array.
10. A sound field measurement-oriented multi-objective collaborative three-dimensional microphone array optimization apparatus, the apparatus comprising:
the acquisition module is used for acquiring geometric parameters of the traditional array element group of the microphone array and generating a primary array element group of the microphone array through improved Sine chaotic mapping;
the function construction module is used for constructing a mixed fitness function;
and the output module is used for obtaining the optimized three-dimensional microphone array for measuring multi-objective coordination in the sound field according to the microphone array primary array element group, the mixed fitness function and the whale elite optimization algorithm.
CN202310331057.9A 2023-03-30 2023-03-30 Sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization method and device Pending CN116381604A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310331057.9A CN116381604A (en) 2023-03-30 2023-03-30 Sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310331057.9A CN116381604A (en) 2023-03-30 2023-03-30 Sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization method and device

Publications (1)

Publication Number Publication Date
CN116381604A true CN116381604A (en) 2023-07-04

Family

ID=86962845

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310331057.9A Pending CN116381604A (en) 2023-03-30 2023-03-30 Sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization method and device

Country Status (1)

Country Link
CN (1) CN116381604A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114547817A (en) * 2022-01-24 2022-05-27 浙江大学 Sparse sensor array design method based on global enhanced whale optimization algorithm

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114547817A (en) * 2022-01-24 2022-05-27 浙江大学 Sparse sensor array design method based on global enhanced whale optimization algorithm

Similar Documents

Publication Publication Date Title
CN109085531B (en) Near-field source arrival angle estimation method based on neural network
Schmidt et al. Environmentally tolerant beamforming for high‐resolution matched field processing: Deterministic mismatch
US20190052957A1 (en) Microphone probe, method, system and computer program product for audio signals processing
CN114624689B (en) Near-field focusing sound source distance calculation method and system based on acoustic imaging instrument
CN111060872A (en) Sound source positioning method and device based on microphone array and computer equipment
CN116381604A (en) Sound field measurement-oriented multi-target cooperative three-dimensional microphone array optimization method and device
CN111432368A (en) Ranging and positioning method suitable for sparse anchor node WSN
CN109343003B (en) Method for identifying sound source formed by fast iterative shrinking wave beams
CN113311397B (en) Large array rapid self-adaptive anti-interference method based on convolutional neural network
CN114730000A (en) Method and device for determining array element layout of ultrasonic phased array and storage medium
CN109489796A (en) A kind of underwater complex structural radiation noise source fixation and recognition based on unit radiation method and acoustic radiation forecasting procedure
CN110736976B (en) Method for estimating performance of sonar beam former of any array
JP4293986B2 (en) Method and system for representing a sound field
CN110095755B (en) Sound source positioning method
CN111273231A (en) Indoor sound source positioning method based on different microphone array topological structure analysis
Frank et al. Constant-beamwidth kronecker product beamforming with nonuniform planar arrays
Rasumow et al. Robustness of virtual artificial head topologies with respect to microphone positioning
Li et al. Performance analysis for focused beamformers in passive underwater acoustic localization
CN115452141B (en) Non-uniform acoustic imaging method
CN114791590B (en) Roll compensation beam generation method, device, circuit and system
CN116309921A (en) Delay summation acoustic imaging parallel acceleration method based on CUDA technology
CN112305497B (en) DOA estimation direction finding fuzzy elimination method for near field microphone array
US4648057A (en) Robust iterative technique for high-resolution spatial processing and spectral estimation
Wang et al. A Survey of Target Orientation Detection Algorithms Based on GPU Parallel Computing
Zhu et al. A validation study for a bat-inspired sonar sensing simulator

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination