Loudspeaker array sparse optimization method and system based on improved genetic algorithm

Abstract

The invention belongs to the field of array signal processing, and discloses a loudspeaker array sparse optimization method and system based on an improved genetic algorithm, and the method comprises the steps: constructing an optimization function which takes the number of array elements, the width of a beam main lobe and the amplitude of a side lobe as targets, carrying out the crossover mutation operation through an adaptive genetic algorithm, and generating a new population; and updating the original population by adopting the new population and the alternative population individuals, calculating a current temperature value based on a temperature attenuation mechanism to update iteration parameters, and outputting an optimal population until a termination condition is reached. According to the method, the dynamic adjustment capability and the temperature attenuation mechanism of the adaptive genetic algorithm are fused, the global search capability is enhanced, the number of crossover and mutation genes and the size of alternative populations are dynamically adjusted through the temperature value, and the iteration process is optimized. By adopting the method, the defects of a traditional algorithm are effectively overcome, the optimization efficiency and precision are improved, the method is excellent in multi-objective optimization and trade-off problems, and the high-precision and high-efficiency requirements of loudspeaker array sparse optimization are met.

Description

Technical Field

[0001] The present invention belongs to the technical field of array signal processing, and in particular relates to a loudspeaker array sparse optimization method and system based on an improved genetic algorithm. Background Art

[0002] In the field of acoustics, speaker arrays have been widely used in numerous scenarios due to their unique advantage in directivity. By arranging multiple speaker units in a specific structure, they achieve precise control of sound propagation in space. Sparse optimization technology for speaker arrays has garnered significant attention. This technology cleverly optimizes the position and number of array elements within a uniformly arranged array formation, achieving beamform optimization of the array's radiated sound waves and reducing the overall cost of the speaker array system, demonstrating its high practical value.

[0003] However, many challenges remain in the current sparse optimization technology for loudspeaker arrays. Intelligent optimization methods, such as simulated annealing, genetic algorithms, and particle swarm optimization, are widely used, but each has significant drawbacks. While the simulated annealing algorithm can effectively avoid local optimal solutions, its convergence rate is extremely slow, severely limiting its use in time-sensitive applications. While the particle swarm optimization algorithm converges quickly, it is prone to local optimal solutions and cannot guarantee a global optimal solution. While the genetic algorithm possesses good global search capabilities and a moderate convergence rate, it can also become trapped in local optimal solutions when faced with complex optimization problems. Even though the improved adaptive genetic algorithm can enhance global search capabilities by dynamically adjusting the crossover and mutation rates, its overall optimization effect is poor in multi-objective optimization, especially when there are trade-offs between objectives, making it difficult to meet the requirements for high-precision and high-efficiency optimization in practical applications. Summary of the Invention

[0004] The present invention provides a loudspeaker array sparse optimization method and system based on an improved genetic algorithm. This method introduces the temperature decay mechanism of a simulated annealing algorithm into an adaptive genetic algorithm. During the optimization process, the temperature decay is used to control the number of genes that cross and mutate in the adaptive genetic algorithm. The idea of ​​accepting differential solutions in the simulated annealing algorithm is applied to the selection phase of the genetic algorithm. A dynamic alternative population is used to ensure a certain population diversity, further expanding the search range. This method improves the performance of the adaptive genetic algorithm in complex multi-objective optimization problems while maintaining the global search capability of the algorithm without affecting the convergence of the algorithm.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A sparse optimization method for a loudspeaker array based on an improved genetic algorithm, comprising:

[0007] The optimization objective function is constructed by taking the number of array elements, the main lobe width and the side lobe amplitude of the array output beam as the optimization target parameters;

[0008] Crossover and mutation process: Adaptive genetic algorithm is used to perform crossover and mutation operations on the original population in the optimization objective function to update the new population;

[0009] Selection process: Use individuals from the new population to update the original population; or, starting from the second iteration, use individuals from the new population to update the original population, and then use individuals from the alternative population to update the original population a second time; all replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the alternative population to obtain the updated original population and alternative population;

[0010] Temperature calculation process: Based on the temperature decay mechanism, the current temperature value is calculated; the current temperature value is used to update the number of crossover and mutation gene points and the size of the candidate population for the next round of iteration;

[0011] The crossover mutation process, selection process and temperature calculation process are repeated until the iteration termination condition is reached, and the original population after the iteration update is output as the optimal population to obtain the sparse optimization result of the speaker array.

[0012] Furthermore, the optimization objective function is constructed based on the main lobe width change value, the first side lobe peak change value, the second side lobe peak change value, the third side lobe peak change value, the remaining side lobe peak change values, and the change value of the number of array elements before and after optimization of the original uniform rectangular array and the optimized array radiation beam pattern. The specific formula is as follows:

[0013] f=α·W b +β·S1+γ·S2+ε·S3+σ·S others +η·N array

[0014] Where f represents the optimization objective function; W b Indicates the change value of the main lobe width; S1, S2, S3, S others Respectively represent the first side lobe peak value change value, the second side lobe peak value change value, and the remaining side lobe peak value change values; N array It represents the change in the number of array elements before and after optimization; α, β, γ, ε, σ, and η are the adjustment coefficients of the corresponding target parameters, which are used to balance the proportion of the target parameters in the fitness function.

[0015] Furthermore, before the adaptive genetic algorithm is used to perform a crossover and mutation operation on the original population in the optimization objective function, the method further includes:

[0016] Initialize the original population in the optimization objective function, including:

[0017] Set the original population size, original number of array elements, number of iterations, initial annealing temperature, probability of array element removal, and alternative population size;

[0018] Based on the set probability of element removal, an element removal operation is performed on the initialized original population to obtain an initialized population.

[0019] Furthermore, the adaptive genetic algorithm is used to perform a crossover and mutation operation on the original population in the optimization objective function, including:

[0020] Determine the number of gene points involved in crossover and mutation based on the current temperature value;

[0021] Traverse the individuals in the population and determine whether the traversal is completed. If it is completed, end the crossover mutation process; if not, proceed to the next step;

[0022] Determine whether the current individual is the best individual in the population. If so, select the next individual and return to the traversal step; if not, proceed to the next step;

[0023] Based on the individual fitness value and the average fitness of the original population, the measurement index is calculated;

[0024] Determine whether the measurement indicator is greater than or equal to the preset critical value; if the measurement indicator is greater than or equal to the critical value, perform a crossover operation first, then a mutation operation; otherwise, perform a mutation operation first, then a crossover operation; in the crossover operation, select a multi-point crossover mode or a single-point crossover mode according to the current temperature value;

[0025] After completing the crossover and mutation operations, a new population is obtained by updating.

[0026] Furthermore, the calculation formula of the measurement indicator is as follows:

[0027]

[0028] Where r is the population diversity index, which is used to represent the measurement index; N p represents the number of individuals in the original population, f i Represents the fitness of individual i, i ranges from 1 to n, and n is a natural number; represents the average fitness of the original population.

[0029] Furthermore, the original population is updated using individuals from the new population; or, starting from the second round of iteration, after the original population is updated using individuals from the new population, the original population is updated a second time using individuals from the alternative population; all replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the alternative population to obtain the updated original population and alternative population, including:

[0030] When the number of iteration rounds is the first round, traverse the individuals in the new population;

[0031] If the fitness of an individual in the new population is greater than the average fitness of the original population, then this individual will replace a random individual in the original population whose fitness value is lower than the average fitness of the original population; the individual replaced from the original population will be stored in the alternative population;

[0032] Obtain updated original population and alternative population;

[0033] When the number of iterations is greater than or equal to 2, traverse the individuals in the new population;

[0034] If the fitness of an individual in the new population is greater than the average fitness of the original population, this individual will be used to replace a random individual in the original population whose fitness value is lower than the average fitness of the original population; the replaced individual will be stored in the alternative population;

[0035] Recalculate the current average fitness of the original population;

[0036] Traverse the individuals in the candidate population;

[0037] If the fitness of an individual in the alternative population is greater than the current average fitness of the original population, this individual is used to replace a random individual in the original population whose fitness is lower than the average fitness; the remaining individuals in the new population whose fitness values ​​are lower than the current average fitness of the original population are stored in the alternative population;

[0038] Get the updated original population and alternative population.

[0039] Furthermore, the current temperature value is calculated based on the temperature attenuation mechanism. The specific formula is as follows:

[0040] T i+1 =τ·(T i -ln(T0 / i+1))

[0041] Where, T i represents the temperature value at the i-th time; T i+1 represents the i+1th temperature value; τ is the adjustment coefficient, which is used to control the overall speed of temperature decay; T0 represents the initial temperature; i is the current iteration number, and i≥1.

[0042] A speaker array sparse optimization system based on an improved genetic algorithm, comprising:

[0043] A construction module is used to construct an optimization objective function using the number of array elements of the speaker array, the main lobe width and the side lobe amplitude of the array output beam as optimization target parameters;

[0044] The crossover and mutation module is used to perform the crossover and mutation process: the adaptive genetic algorithm is used to perform crossover and mutation operations on the original population in the optimization objective function to update the new population;

[0045] The population update module is used to perform the selection process: the original population is updated with individuals from the new population; or, starting from the second iteration, the original population is updated with individuals from the new population, and then the original population is updated again with individuals from the alternative population; all replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the alternative population to obtain the updated original population and alternative population;

[0046] A calculation module is used to perform a temperature calculation process: based on the temperature decay mechanism, the current temperature value is calculated; the current temperature value is used to update the number of crossover and mutation gene points and the size of the candidate population for the next iteration;

[0047] The iteration module is used to repeatedly execute the cross-mutation process, the selection process and the temperature calculation process until the iteration termination condition is reached, and output the original population after the current iteration update as the optimal population to obtain the sparse optimization result of the speaker array.

[0048] An electronic device, comprising:

[0049] Memory for storing computer programs;

[0050] The processor is configured to implement the steps of the above-mentioned loudspeaker array sparse optimization method based on the improved genetic algorithm when executing the computer program.

[0051] A computer-readable storage medium stores a computer program, which, when executed by a processor, is used to implement the steps of the above-mentioned loudspeaker array sparse optimization method based on an improved genetic algorithm.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] The present invention provides a sparse optimization method for loudspeaker arrays based on an improved genetic algorithm. An optimization function is constructed with the number of array elements, beam main lobe width, and side lobe amplitude as targets. A new population is generated by performing crossover and mutation operations using an adaptive genetic algorithm, and the original population is updated using the new population and individuals from the alternative population. At the same time, the current temperature value is calculated based on a temperature attenuation mechanism to update the iteration parameters until the termination condition is reached and the optimal population is output. This method combines the dynamic adjustment capability of the adaptive genetic algorithm with the temperature attenuation mechanism, which not only enhances the global search capability, but also dynamically adjusts the number of crossover and mutation gene points and the size of the alternative population through the temperature value, thereby optimizing the iterative process. This method effectively overcomes the shortcomings of traditional algorithms, improves optimization efficiency and accuracy, and performs particularly well in multi-objective optimization and trade-off problems, meeting the high-precision and high-efficiency requirements of sparse optimization of loudspeaker arrays.

[0054] In the present invention, the optimization objective function is preferably constructed based on the mainlobe width, sidelobe peak, and change in the number of array elements of the original uniform rectangular array and the optimized array radiation beam patterns. Adjustment coefficients are then used to balance the weight of each objective parameter in the fitness function. This allows the optimization objective function to more comprehensively consider the key factors of loudspeaker array sparse optimization, more accurately reflect the optimization effect, facilitate reasonable trade-offs between objectives in multi-objective optimization, and improve the practicality and reliability of the optimization results.

[0055] In the present invention, the original population in the optimization objective function is preferably initialized, including setting the population size, the original number of elements, the number of iterations, the initial annealing temperature, the element removal probability, and the candidate population size, and then performing element removal. Reasonable initialization parameter settings and element removal lay a good foundation for the subsequent optimization process, helping to quickly find the optimal solution, improving the stability and efficiency of the optimization process, and reducing unnecessary waste of computing resources.

[0056] In the present invention, when performing crossover and mutation operations on the original population in the optimization objective function using an adaptive genetic algorithm, the number of gene points involved in crossover and mutation is determined based on the current temperature value, and the order and pattern of crossover and mutation are selected based on individual fitness and the average fitness of the original population. This method dynamically adjusts the number of gene points involved in crossover and mutation based on the temperature value and selects the order and pattern of crossover and mutation based on individual fitness differences. This enhances the algorithm's adaptability and search capabilities, helps avoid falling into local optima, improves global search capabilities, and accelerates convergence.

[0057] In the present invention, the metric is preferably calculated based on the number of individuals, individual fitness, and average fitness of the original population, reflecting population diversity. This metric provides a decision-making basis for the crossover and mutation operations, helping to maintain population diversity during the crossover and mutation process, preventing premature population convergence, and improving the algorithm's global search capabilities, thereby increasing the likelihood of finding a more optimal solution.

[0058] In the present invention, the selection process preferably uses new and candidate populations to update the original population, and individuals with below-average fitness in both the replaced and new populations are stored in the candidate population. This mechanism, through multiple updates and the candidate population mechanism, fully utilizes information from excellent individuals, improves population quality, helps accelerate convergence, and enhances the accuracy and stability of optimization results, making the optimization process more flexible and efficient.

[0059] In the present invention, a temperature decay mechanism is preferably used to calculate the current temperature value by adjusting the coefficient, the initial temperature, and the current iteration number. This temperature decay mechanism can dynamically adjust the temperature value, control changes in crossover and mutation parameters, and the size of the candidate population, and balance the algorithm's global and local search capabilities. This helps gradually focus on the optimal solution area during the optimization process, improves optimization efficiency and accuracy, and makes the optimization process more stable and controllable. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A schematic diagram of a sparse optimization model for a uniform rectangular array provided in an embodiment of the present invention;

[0061] Figure 2 The main steps of the AGA-SA algorithm in array sparse optimization provided by the embodiment of the present invention;

[0062] Figure 3 A flowchart of the crossover mutation process provided by an embodiment of the present invention;

[0063] Figure 4 A flowchart of a selection process provided by an embodiment of the present invention;

[0064] Figure 5 A comparison diagram of the array formations obtained before and after array sparse optimization provided by an embodiment of the present invention;

[0065] Figure 6 A two-dimensional beam comparison diagram of the array output beam after optimization using the adaptive genetic algorithm (IAGA), dynamic search strategy discrete particle swarm optimization (DSSDPSO), and improved adaptive genetic algorithm (AGA-SA) provided in an embodiment of the present invention;

[0066] Figure 7 A three-dimensional diagram of the array output beam before array sparse optimization provided in an embodiment of the present invention;

[0067] Figure 8 A three-dimensional diagram of the array output beam after array sparse optimization provided by an embodiment of the present invention;

[0068] Figure 9 A comparison chart of the relationship between the optimization fitness function of the three methods IAGA, DSSDPSO and AGA-SA provided in an embodiment of the present invention and the number of iterations;

[0069] Figure 10 A flowchart of a speaker array sparse optimization method based on an improved genetic algorithm provided by an embodiment of the present invention;

[0070] Figure 11 A schematic structural diagram of a loudspeaker array sparse optimization system based on an improved genetic algorithm provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0071] The embodiment provides a sparse optimization method for a loudspeaker array based on an improved genetic algorithm, which uses the number of array elements, the main lobe width and the side lobe amplitude of the array output beam as optimization target parameters to construct an optimization objective function.

[0072] Crossover and mutation process: Adaptive genetic algorithm is used to perform crossover and mutation operations on the original population in the optimization objective function to update the new population;

[0073] Selection process: Use individuals from the new population to update the original population; or, starting from the second iteration, use individuals from the new population to update the original population, and then use individuals from the alternative population to update the original population a second time; all replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the alternative population to obtain the updated original population and alternative population;

[0074] Temperature calculation process: Based on the temperature decay mechanism, the current temperature value is calculated; the current temperature value is used to update the number of crossover and mutation gene points and the size of the candidate population for the next round of iteration;

[0075] The crossover mutation process, selection process and temperature calculation process are repeated until the iteration termination condition is reached, and the original population after the iteration update is output as the optimal population to obtain the sparse optimization result of the speaker array.

[0076] The optimization method provided by this embodiment is further described below with reference to the accompanying drawings:

[0077] like Figure 2 As shown, this embodiment provides a speaker array sparse optimization method based on an improved genetic algorithm, and the specific steps include:

[0078] Step 1) Design the optimization objective function, which is the fitness function in the genetic algorithm.

[0079] The target parameters in the fitness function include: Figure 1 As shown, the main lobe width change value W of the original uniform rectangular array and the optimized array radiation beam pattern b =BW o / BW i , BW o and BW i Represents the main lobe width of the original array and the optimized array radiation beam pattern respectively; the first side lobe peak change value S1=SL1 o / SL1 i , of which SL1 o and SL1 i Represent the first sidelobe level of the original array and optimized array radiation beam pattern respectively. Similarly, define S1, S2, S3, S others are the change degrees of the second and third side lobes and the remaining side lobe peaks of the original array and the optimized array respectively. In addition, the change degree of the number of array elements before and after optimization, N, is also introduced. array , which is defined as N / n, where N represents the initial number of array elements and n represents the number of array elements after optimization, so as to fully evaluate the optimization effect.

[0080] The overall fitness function f is defined as:

[0081] f=α·W b +β·S1+γ·S2+ε·S3+σ·S others +η·N array (1)

[0082] Where α, β, γ, ε, σ, and η are adjustment coefficients of different target parameters, which are used to balance the proportion of parameter values ​​in the fitness function.

[0083] Step 2) The specific execution process of crossover and mutation.

[0084] The temperature decay mechanism in the simulated annealing algorithm is introduced. A logarithmic function is used to achieve a smoother temperature drop process. The specific temperature decay formula is as follows

[0085] T i+1 =τ·(T i -ln(T0 / i+1)) (2)

[0086] Where τ is the adjustment coefficient, which is used to control the overall speed of temperature decay, T0 is the initial temperature value, i is the current iteration number, and i ≥ 1.

[0087] The number of genes involved in crossover and mutation is controlled by temperature decay during the iteration process. In the early stages of the algorithm, multi-point crossover and mutation are used to expand the search range. As the temperature decreases, single-point crossover and mutation are gradually used to improve search accuracy. The crossover and mutation probabilities during the iteration process are shown in Equations (3) and (4).

[0088]

[0089]

[0090] In the formula, f' represents the fitness value of the individual with the largest fitness among the individuals on both sides of the crossover, f avg is the mean fitness of the population, P c is the crossover probability, P m is the mutation probability, P cmax , P cmin , P mmax , P mmin They are the maximum crossover probability, minimum crossover probability, maximum mutation probability, and minimum mutation probability.

[0091] At the same time, the population diversity index is introduced to control the order of crossover and mutation. Specifically, the standard deviation function is used as a measure of population diversity, as shown in formula (5).

[0092]

[0093] Where N p represents the number of individuals in the population. A critical value is set. When the population standard deviation is greater than the critical value, it indicates that the individual differences in the current population are large. In this case, the algorithm prioritizes crossover and then mutation to increase the diversity of the population and avoid local optimal solutions. This is because a larger population standard deviation indicates that there are more different solutions (i.e., array element arrangements) in the population, which helps explore different regions of the search space, thereby increasing the possibility of finding the global optimal solution.

[0094] When the population standard deviation is less than the critical value, it means that the differences between individuals in the population are small. At this time, the algorithm first performs the mutation operation and then the crossover operation, which ensures a certain local search capability in the later stage of the algorithm to maintain the diversity of the population. This is because a smaller population standard deviation may mean that the population is converging to a local optimal solution. At this time, performing the mutation operation can introduce new changes, help jump out of the local optimal solution, and continue to look for a better solution. The complete crossover and mutation process is as follows Figure 3 shown.

[0095] Step 3) involves the specific implementation process of the selection.

[0096] Utilizing a dynamic candidate population to improve the selection process. In the speaker array optimization problem, simply combining a simulated annealing algorithm with an adaptive genetic algorithm presents challenges in handling "difference solutions" (i.e., solutions with poorer fitness than the current solution). Specifically, when a new solution with poorer fitness is accepted, it randomly replaces an even poorer solution in the original population. Due to the random nature of the genetic algorithm, the performance of the two difference solutions in the next generation after subsequent crossover and mutation operations is difficult to predict. This approach can easily lead to a loss of population diversity, as some potential difference solutions cannot be fully retained and utilized after being replaced from the population.

[0097] To solve the above problems, this embodiment introduces an improved selection strategy that maintains population diversity and improves algorithm performance by dynamically adjusting the size of the candidate population and utilizing historical information.

[0098] By calculating the current annealing temperature T, the size of the candidate population is dynamically adjusted to The minimum value is 10 to ensure the diversity of the population and the coverage of the search space. In the first iteration, the average fitness value f of the original population is calculated. avg In the selection process, if the fitness value of the new population individuals generated by crossover and mutation is higher than f avg , then randomly replace any fitness value in the original population that is lower than f avg Individuals, and the replaced individuals are stored in the candidate population. Then the individuals in the new population whose fitness is lower than f avg The individuals are stored in the candidate population, and the individuals in the candidate population are sorted in descending order according to the fitness value to obtain the initial candidate population.

[0099] In the subsequent iterations, when the selection process updates the original population, in addition to using the information in the new population, it also uses the information in the candidate population that has also undergone crossover and mutation. Specifically, after the new population updates the original population, the f of the original population is calculated again. avg , if the individual fitness value in the alternative population is higher than f at this time avg , then randomly replace any fitness value in the original population that is lower than f avg Similarly, the replaced individuals are temporarily stored and used to update the candidate population, so as to update the original population twice, so as to enhance the global search ability of the algorithm by utilizing the potential difference solution of the previous generation. The overall selection operation flow chart is as follows: Figure 4 shown.

[0100] Step 4) mainly involves the overall process of executing the AGA-SA optimization algorithm.

[0101] The simulated annealing mechanism is introduced into the crossover, mutation and selection steps of the adaptive genetic algorithm to improve the algorithm. An adaptive genetic algorithm combined with the simulated annealing algorithm is proposed. The algorithm execution pseudo code is shown in Table 1:

[0102] Table 1 shows the pseudo code execution steps of the improved adaptive genetic algorithm

[0103]

[0104]

[0105] It can be seen that the present invention can optimize the output beam of the loudspeaker array by performing sparse optimization design on the array. The present invention introduces an adaptive genetic algorithm combined with simulated annealing to perform array optimization, specifically realizing sidelobe suppression while limiting the mainlobe width, thereby achieving a beam output effect similar to or better than the original array with a smaller number of array elements.

[0106] The basic principles and implementation scheme of the present invention have been verified by computer numerical simulation. The results show that compared with existing commonly used optimization methods, the method proposed in the present invention has certain advantages in computational efficiency and the optimization effect of array output waveform after array sparse optimization, and is feasible in the sparse optimization design of loudspeaker arrays.

[0107] The specific implementation process of the speaker array sparse optimization method based on the improved genetic algorithm provided in this embodiment is as follows in combination with the application scenario:

[0108] Through the sparse optimization simulation of uniform rectangular planar array, the existing adaptive genetic algorithm (IAGA), dynamic search strategy discrete particle swarm optimization (DSSDPSO) and improved adaptive genetic algorithm (AGA-SA) are compared.

[0109] Take the original uniform rectangular array with the number of rows M = 10, the number of columns N = 10, and the array element spacing of half a wavelength d x =d y =λ / 2, the pitch angle θ and the azimuth angle are The range is -90° to 90°, with a step size of 0.5°. The probability of element removal is P e =0.3, the number of individuals in the population N p =150, the number of individual genes D=M×N=100, the number of iterations K=150, the annealing temperature T0=300, and all optimization algorithms in the simulation use the same fitness function. The parameters α, β, γ, ε, σ, and η in formula (1) are 1.5, 1.6, 1.2, 0.8, 0.8, and 0.6, respectively.

[0110] For the IAGA algorithm, the number of genes in the crossover and mutation operations was fixed at 4 and 2, respectively. For the DSSDPSO algorithm, the number of particles in the particle swarm was set to 500, the number of iterations was 150, the learning factor was 1.5, and the inertia weight was 0.8. Slice beam data with an azimuth angle step of 5° was selected for comprehensive calculations.

[0111] In the AGA-SA algorithm, the number of genes in the crossover and mutation operations is dynamically adjusted according to the current temperature T. Specifically, the number of genes selected in the crossover operation is 7, 4, and 1 in the high, medium, and low temperature stages, respectively, and the number of genes selected in the mutation operation is 5, 3, and 1 in the corresponding temperature stages. This design aims to promote the diversity of the population by increasing the number of crossover and mutation genes at higher temperatures, and to increase the convergence speed by reducing the number of genes when the temperature decreases. At the same time, the size of the dynamic alternative population matrix is ​​determined by the current temperature T / 6 rounded up. When the calculated result is less than 10, the matrix size is fixed to 10. This strategy ensures that the population can maintain a certain scale of alternative solutions under different temperature conditions, thereby improving the global search capability of the algorithm.

[0112] Figure 5 Schematic diagrams of the directional pattern of a uniform planar rectangular array before and after optimization using different algorithms are presented. Compared with the traditional adaptive genetic algorithm and the particle swarm algorithm with a dynamic search strategy, the proposed method (i.e., AGA-SA) can achieve multi-objective beam optimization with limited mainlobe width and suppressed sidelobe levels, and obtain relatively better beam output results. Figure 6 is a schematic diagram of the initial array three-dimensional beam pattern in the xz plane and yz plane, Figure 7 This is a schematic diagram of the three-dimensional beam pattern on the corresponding plane after optimization using the method proposed in this invention (AGA-SA). The corresponding coordinate system is set as Where Beam is the normalized beam data, Figure 6 and Figure 7 It can be seen from the comparison that the AGA-SA method can maximize the sidelobe suppression effect while meeting the mainlobe width limitation. Figure 8 The variation of fitness values ​​of this method and the other two comparison algorithms with the number of iterations in a total of 150 iterations is given. When the fitness function is the same, the AGA-SA method can achieve a higher fitness value and end the iteration earlier without falling into a local optimal solution. Figure 9 A comparison diagram of the array arrangement before and after optimization using the method proposed in the present invention (i.e., the AGA-SA algorithm) is shown. The hollow circles represent the positions of the array elements in the initial uniform rectangular array, and the solid circles represent the positions of the array elements after optimization using the method proposed in the present invention, indicating that the AGA-SA algorithm effectively achieves the sparsification of the original array.

[0113] It can be seen that the method uses an improved genetic algorithm to solve the maximum value of the fitness function to achieve the goal of optimizing the performance parameters of the output sound field beam diagram of the optimized loudspeaker array. The fitness function designed by the present invention includes the optimization parameters of the number of array elements, the main lobe width and the sidelobe amplitude of the array output beam. Then, an improved adaptive genetic algorithm (AGA-SA) algorithm combined with simulated annealing is proposed to solve the multi-objective optimization problem; in the crossover and mutation steps of the adaptive genetic algorithm (AGA), the temperature decay mechanism of the simulated annealing algorithm is introduced, and the temperature decay in the iterative process is used to control the number of genes involved in crossover and mutation; at the same time, the standard deviation is used to represent the population diversity index, which is used to control the order of crossover and mutation; in the selection step of the AGA algorithm, an alternative population mechanism is introduced to maintain population diversity by dynamically adjusting the size of the alternative population and using historical information.

[0114] The present invention can optimize the output beam of the loudspeaker array by performing sparse optimization design on the array. The present invention introduces an adaptive genetic algorithm combined with simulated annealing to perform array optimization, specifically implementing sidelobe suppression while limiting the mainlobe width, thereby achieving a beam output effect similar to or better than the original array with a smaller number of array elements.

[0115] The basic principles and implementation scheme of the present invention have been verified by computer numerical simulation. The results show that compared with existing commonly used optimization methods, the method proposed in the present invention has certain advantages in computational efficiency and the optimization effect of array output waveform after array sparse optimization, and is feasible in the sparse optimization design of loudspeaker arrays.

[0116] For example, Figure 10 As shown, this embodiment provides a speaker array sparse optimization method based on an improved genetic algorithm, comprising the following steps:

[0117] The optimization objective function is constructed by taking the number of array elements, the main lobe width and the side lobe amplitude of the array output beam as the optimization target parameters;

[0118] Crossover and mutation process: Adaptive genetic algorithm is used to perform crossover and mutation operations on the original population in the optimization objective function to update the new population;

[0119] Selection process: Use individuals from the new population to update the original population; or, starting from the second iteration, use individuals from the new population to update the original population, and then use individuals from the alternative population to update the original population a second time; all replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the alternative population to obtain the updated original population and alternative population;

[0120] Temperature calculation process: Based on the temperature decay mechanism, the current temperature value is calculated; the current temperature value is used to update the number of crossover and mutation gene points and the size of the candidate population for the next round of iteration;

[0121] The crossover mutation process, selection process and temperature calculation process are repeated until the iteration termination condition is reached, and the original population after the iteration update is output as the optimal population to obtain the sparse optimization result of the speaker array.

[0122] In this embodiment, the optimization objective function is constructed based on the main lobe width change value, the first side lobe peak change value, the second side lobe peak change value, the third side lobe peak change value, the remaining side lobe peak change values, and the change value of the number of array elements before and after optimization of the original uniform rectangular array and the optimized array radiation beam pattern. The specific formula is as follows:

[0123] f=α·W b +β·S1+γ·S2+ε·S3+σ·S others +η·N array

[0124] Where f represents the optimization objective function; W b Indicates the change value of the main lobe width; S1, S2, S3, S others Respectively represent the first side lobe peak value change value, the second side lobe peak value change value, and the remaining side lobe peak value change values; N array It represents the change in the number of array elements before and after optimization; α, β, γ, ε, σ, and η are the adjustment coefficients of the corresponding target parameters, which are used to balance the proportion of the target parameters in the fitness function.

[0125] In this embodiment, before the adaptive genetic algorithm is used to perform the crossover and mutation operation on the original population in the optimization objective function, the method further includes:

[0126] Initialize the original population in the optimization objective function, including:

[0127] Set the original population size, original number of array elements, number of iterations, initial annealing temperature, probability of array element removal, and alternative population size;

[0128] Based on the set probability of element removal, an element removal operation is performed on the initialized original population to obtain an initialized population.

[0129] In this embodiment, the adaptive genetic algorithm is used to perform a crossover and mutation operation on the original population in the optimization objective function, including:

[0130] Determine the number of gene points involved in crossover and mutation based on the current temperature value;

[0131] Traverse the individuals in the population and determine whether the traversal is completed. If it is completed, end the crossover mutation process; if not, proceed to the next step;

[0132] Determine whether the current individual is the best individual in the population. If so, select the next individual and return to the traversal step; if not, proceed to the next step;

[0133] Based on the individual fitness value and the average fitness of the original population, the measurement index is calculated;

[0134] Determine whether the measurement indicator is greater than or equal to the preset critical value; if the measurement indicator is greater than or equal to the critical value, perform a crossover operation first, then a mutation operation; otherwise, perform a mutation operation first, then a crossover operation; in the crossover operation, select a multi-point crossover mode or a single-point crossover mode according to the current temperature value;

[0135] After completing the crossover and mutation operations, a new population is obtained by updating.

[0136] In this embodiment, the calculation formula of the measurement index is as follows:

[0137]

[0138] Where r is the population diversity index, which is used to represent the measurement index; N p represents the number of individuals in the original population, f i Represents the fitness of individual i, i ranges from 1 to n, and n is a natural number; represents the average fitness of the original population.

[0139] In this embodiment, the original population is updated using individuals from the new population; or, starting from the second round of iteration, after the original population is updated using individuals from the new population, the original population is updated a second time using individuals from the alternative population; all replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the alternative population to obtain the updated original population and alternative population, including:

[0140] When the number of iteration rounds is the first round, traverse the individuals in the new population;

[0141] If the fitness of an individual in the new population is greater than the average fitness of the original population, then this individual will replace a random individual in the original population whose fitness value is lower than the average fitness of the original population; the individual replaced from the original population will be stored in the alternative population;

[0142] Obtain updated original population and alternative population;

[0143] When the number of iterations is greater than or equal to 2, traverse the individuals in the new population;

[0144] If the fitness of an individual in the new population is greater than the average fitness of the original population, this individual will be used to replace a random individual in the original population whose fitness value is lower than the average fitness of the original population; the replaced individual will be stored in the alternative population;

[0145] Recalculate the current average fitness of the original population;

[0146] Traverse the individuals in the candidate population;

[0147] If the fitness of an individual in the alternative population is greater than the current average fitness of the original population, this individual is used to replace a random individual in the original population whose fitness is lower than the average fitness; the remaining individuals in the new population whose fitness values ​​are lower than the current average fitness of the original population are stored in the alternative population;

[0148] Get the updated original population and alternative population.

[0149] In this embodiment, the current temperature value is calculated based on the temperature attenuation mechanism, and the specific formula is as follows:

[0150] T i+1 =τ·(T i -ln(T0 / i+1))

[0151] Where, T i represents the temperature value at the i-th time; T i+1 represents the i+1th temperature value; τ is the adjustment coefficient, which is used to control the overall speed of temperature decay; T0 represents the initial temperature; i is the current iteration number, and i≥1.

[0152] like Figure 11As shown, this embodiment also provides a loudspeaker array sparse optimization system based on an improved genetic algorithm, including: a construction module, used to construct an optimization objective function using the number of array elements of the loudspeaker array, the main lobe width and the side lobe amplitude of the array output beam as optimization target parameters; a crossover mutation module, used to perform a crossover mutation process: using an adaptive genetic algorithm to perform a crossover mutation operation on the original population in the optimization objective function to update a new population; a population update module, used to perform a selection process: using individuals in the new population to update the original population; or, starting from the second round of iteration, using individuals in the new population to update the original population, and then using individuals in the alternative population The individuals perform a secondary update on the original population; all replaced individuals in the original population and individuals in the new population with lower fitness than the average fitness of the original population are stored in the alternative population to obtain an updated original population and alternative population; a calculation module is used to perform a temperature calculation process: based on the temperature attenuation mechanism, the current temperature value is calculated; the current temperature value is used to update the number of crossover and mutation gene points and the size of the alternative population in the next round of iteration; an iteration module is used to repeatedly perform the crossover and mutation process, the selection process and the temperature calculation process until the iteration termination condition is reached, and the original population after the current round of iteration update is output as the optimal population to obtain the sparse optimization result of the speaker array.

[0153] The present invention also provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the steps of the speaker array sparse optimization method based on the improved genetic algorithm when executing the computer program.

[0154] When the processor executes the computer program, the steps of the above-mentioned sparse optimization of the loudspeaker array based on the improved genetic algorithm are implemented, for example: the number of array elements of the loudspeaker array, the main lobe width and the side lobe amplitude of the array output beam are used as optimization target parameters to construct an optimization objective function; a crossover mutation process: using an adaptive genetic algorithm to perform a crossover mutation operation on the original population in the optimization objective function to update a new population; a selection process: using individuals in the new population to update the original population; or, starting from the second round of iteration, using individuals in the new population to update the original population, and then using individuals in the alternative population to select the new population. The original population is updated twice; all replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the candidate population to obtain the updated original population and candidate population; temperature calculation process: based on the temperature attenuation mechanism, the current temperature value is calculated; the current temperature value is used to update the number of crossover and mutation gene points and the size of the candidate population for the next round of iteration; the crossover and mutation process, the selection process and the temperature calculation process are repeated until the iteration termination condition is reached, and the original population after the current round of iteration is output as the optimal population to obtain the sparse optimization result of the speaker array.

[0155] Alternatively, when the processor executes the computer program, the functions of the modules in the above system are realized, for example: a construction module for constructing an optimization objective function using the number of array elements of the speaker array, the main lobe width and the side lobe amplitude of the array output beam as optimization target parameters; a crossover mutation module for executing a crossover mutation process: using an adaptive genetic algorithm to perform a crossover mutation operation on the original population in the optimization objective function to update a new population; a population update module for executing a selection process: using individuals in the new population to update the original population; or, starting from the second round of iteration, using individuals in the new population to update the original population, and then using individuals in the alternative population to select the new population. The original population is updated twice; all replaced individuals in the original population and individuals in the new population with lower fitness than the original population are stored in the candidate population to obtain updated original and candidate populations; a calculation module is used to perform a temperature calculation process: based on the temperature attenuation mechanism, the current temperature value is calculated; the current temperature value is used to update the number of crossover and mutation gene points and the size of the candidate population in the next round of iteration; an iteration module is used to repeatedly perform the crossover and mutation process, the selection process and the temperature calculation process until the iteration termination condition is reached, and the original population after the current round of iteration update is output as the optimal population to obtain the sparse optimization result of the speaker array.

[0156] Exemplarily, the computer program can be divided into one or more modules / units, and the one or more modules / units are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments that can complete preset functions, and the instruction segments are used to describe the execution process of the computer program in the speaker array sparse optimization device based on the improved genetic algorithm. For example, the computer program can be divided into a construction module, a crossover mutation module, a population update module, a calculation module and an iteration module; the construction module is used to construct an optimization objective function with the number of array elements of the speaker array, the main lobe width and the sidelobe amplitude of the array output beam as optimization target parameters; the crossover mutation module is used to perform a crossover mutation process: using an adaptive genetic algorithm to perform a crossover mutation operation on the original population in the optimization objective function to update a new population; the population update module is used to perform a selection process: using individuals in the new population to update the original population; or, starting from the second round of iteration, using individuals in the new population to update the original population, and then using an alternative population. Individuals in the population perform a secondary update on the original population; all replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the candidate population to obtain an updated original population and candidate population; a calculation module is used to perform a temperature calculation process: based on the temperature attenuation mechanism, the current temperature value is calculated; the current temperature value is used to update the number of crossover and mutation gene points and the size of the candidate population for the next round of iteration; an iteration module is used to repeatedly perform the crossover and mutation process, the selection process, and the temperature calculation process until the iteration termination condition is reached, and the original population after the current round of iteration update is output as the optimal population to obtain the sparse optimization result of the speaker array.

[0157] The speaker array sparse optimization device based on the improved genetic algorithm can be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The speaker array sparse optimization device based on the improved genetic algorithm can include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the above examples of speaker array sparse optimization devices based on the improved genetic algorithm do not constitute a limitation on speaker array sparse optimization devices based on the improved genetic algorithm. The speaker array sparse optimization device based on the improved genetic algorithm can include more components than those described above, or a combination of certain components, or different components. For example, the speaker array sparse optimization device based on the improved genetic algorithm can also include input and output devices, network access devices, buses, etc.

[0158] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor. The processor serves as the control center of the improved genetic algorithm-based speaker array sparse optimization system, connecting various components of the improved genetic algorithm-based speaker array sparse optimization system using various interfaces and circuits.

[0159] The memory may be used to store the computer program and / or module. The processor implements various functions of the speaker array sparse optimization device based on the improved genetic algorithm by running or executing the computer program and / or module stored in the memory and calling the data stored in the memory.

[0160] The memory may mainly include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function (such as a sound playback function, an image playback function, etc.); the data storage area may store data created based on the use of the mobile phone (such as audio data, a phone book, etc.). In addition, the memory may include a high-speed random access memory and may also include a non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0161] The present invention also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the computer program implements the steps of the speaker array sparse optimization method based on an improved genetic algorithm.

[0162] If the module / unit integrated in the loudspeaker array sparse optimization system based on the improved genetic algorithm is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0163] Based on this understanding, the present invention implements all or part of the processes in the above-mentioned improved genetic algorithm-based speaker array sparse optimization method, and can also be accomplished by using a computer program to instruct related hardware. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of the above-mentioned improved genetic algorithm-based speaker array sparse optimization method. The computer program includes computer program code, which can be in source code form, object code form, executable file, or a pre-set intermediate form.

[0164] The computer-readable storage medium may include: any entity or device that can carry the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0165] It should be noted that the content contained in the computer-readable storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practices in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practices, computer-readable storage media do not include electrical carrier signals and telecommunication signals.

[0166] The above embodiment is only one of the implementation methods that can realize the technical solution of the present invention. The scope of protection claimed by the present invention is not limited only to this embodiment, but also includes changes, replacements and other implementation methods that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed by the present invention.

[0167] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A sparse optimization method for loudspeaker array based on improved genetic algorithm, characterized in that: include: The optimization objective function is constructed by taking the number of array elements, the main lobe width and the side lobe amplitude of the array output beam as the optimization target parameters; Crossover and mutation process: Adaptive genetic algorithm is used to perform crossover and mutation operations on the original population in the optimization objective function to update the new population; Selection process: Use individuals from the new population to update the original population; or, starting from the second round of iteration, use individuals from the new population to update the original population, and then use individuals from the alternative population to update the original population a second time; All replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the alternative population to obtain the updated original population and alternative population; Temperature calculation process: Calculate the current temperature value based on the temperature attenuation mechanism; The current temperature value is used to update the number of crossover and mutation gene points and the size of the alternative population for the next iteration; The crossover mutation process, selection process and temperature calculation process are repeated until the iteration termination condition is reached, and the original population after the iteration update is output as the optimal population to obtain the sparse optimization result of the speaker array.

2. The speaker array sparse optimization method based on improved genetic algorithm according to claim 1, characterized in that: The optimization objective function is constructed based on the main lobe width change value, the first side lobe peak change value, the second side lobe peak change value, the third side lobe peak change value, the remaining side lobe peak change values, and the change value of the number of array elements before and after optimization of the original uniform rectangular array and the optimized array radiation beam pattern. The specific formula is as follows: f=α·W b +β·S1+γ·S2+ε·S3+σ·S others +η·N array Where f represents the optimization objective function; W b Indicates the change value of the main lobe width; S1, S2, S3, S others Respectively represent the first side lobe peak value change value, the second side lobe peak value change value, and the remaining side lobe peak value change values; N array It represents the change in the number of array elements before and after optimization; α, β, γ, ε, σ, and η are the adjustment coefficients of the corresponding target parameters, which are used to balance the proportion of the target parameters in the fitness function.

3. The speaker array sparse optimization method based on improved genetic algorithm according to claim 1, characterized in that: Before the adaptive genetic algorithm is used to perform a crossover and mutation operation on the original population in the optimization objective function, the method further includes: Initialize the original population in the optimization objective function, including: Set the original population size, original number of array elements, number of iterations, initial annealing temperature, probability of array element removal, and alternative population size; Based on the set probability of element removal, an element removal operation is performed on the initialized original population to obtain an initialized population.

4. The speaker array sparse optimization method based on improved genetic algorithm according to claim 1, characterized in that: The method of using the adaptive genetic algorithm to perform a crossover and mutation operation on the original population in the optimization objective function includes: Determine the number of gene points involved in crossover and mutation based on the current temperature value; Traverse the individuals in the population and determine whether the traversal is completed. If it is completed, end the crossover mutation process; if not, proceed to the next step; Determine whether the current individual is the best individual in the population. If so, select the next individual and return to the traversal step; if not, proceed to the next step; Based on the individual fitness value and the average fitness of the original population, the measurement index is calculated; Determine whether the measurement indicator is greater than or equal to the preset critical value; if the measurement indicator is greater than or equal to the critical value, perform a crossover operation first, then a mutation operation; otherwise, perform a mutation operation first, then a crossover operation; in the crossover operation, select a multi-point crossover mode or a single-point crossover mode according to the current temperature value; After completing the crossover and mutation operations, a new population is obtained by updating.

5. The speaker array sparse optimization method based on improved genetic algorithm according to claim 4, characterized in that: The calculation formula of the measurement indicator is as follows: Where r is the population diversity index, which is used to represent the measurement index; N p represents the number of individuals in the original population, f i Represents the fitness of individual i, i ranges from 1 to n, and n is a natural number; represents the average fitness of the original population.

6. The speaker array sparse optimization method based on improved genetic algorithm according to claim 1, characterized in that: The original population is updated using individuals from the new population; or, starting from the second round of iteration, the original population is updated using individuals from the new population, and then the original population is updated a second time using individuals from the alternative population; All replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the candidate population to obtain the updated original population and candidate population, including: When the number of iteration rounds is the first round, traverse the individuals in the new population; If the fitness of an individual in the new population is greater than the average fitness of the original population, then this individual will replace a random individual in the original population whose fitness value is lower than the average fitness of the original population; the individual replaced from the original population will be stored in the alternative population; Obtain updated original population and alternative population; When the number of iterations is greater than or equal to 2, traverse the individuals in the new population; If the fitness of an individual in the new population is greater than the average fitness of the original population, this individual will be used to replace a random individual in the original population whose fitness value is lower than the average fitness of the original population; the replaced individual will be stored in the alternative population; Recalculate the current average fitness of the original population; Traverse the individuals in the candidate population; If the fitness of an individual in the alternative population is greater than the current average fitness of the original population, this individual is used to replace a random individual in the original population whose fitness is lower than the average fitness; the remaining individuals in the new population whose fitness values ​​are lower than the current average fitness of the original population are stored in the alternative population; Get the updated original population and alternative population.

7. The speaker array sparse optimization method based on improved genetic algorithm according to claim 1, characterized in that: The current temperature value is calculated based on the temperature attenuation mechanism. The specific formula is as follows: T i+1 =τ·(T i -ln(T0 / i+1)) Where, T i represents the temperature value at the i-th time; T i+1 represents the i+1th temperature value; τ is the adjustment coefficient, which is used to control the speed of temperature decay as a whole; T0 represents the initial temperature; i is the current iteration number, and i≥1.

8. A speaker array sparse optimization system based on an improved genetic algorithm, characterized in that: include: A construction module is used to construct an optimization objective function using the number of array elements of the speaker array, the main lobe width and the side lobe amplitude of the array output beam as optimization target parameters; The crossover and mutation module is used to perform the crossover and mutation process: the adaptive genetic algorithm is used to perform crossover and mutation operations on the original population in the optimization objective function to update the new population; The population update module is used to perform the selection process: the original population is updated with individuals from the new population; or, starting from the second iteration, the original population is updated with individuals from the new population and then updated again with individuals from the alternative population; All replaced individuals in the original population and individuals in the new population with fitness lower than the average fitness of the original population are stored in the alternative population to obtain the updated original population and alternative population; The calculation module is used to perform the temperature calculation process: based on the temperature attenuation mechanism, the current temperature value is calculated; The current temperature value is used to update the number of crossover and mutation gene points and the size of the alternative population for the next iteration; The iteration module is used to repeatedly execute the cross-mutation process, the selection process and the temperature calculation process until the iteration termination condition is reached, and output the original population after the iteration update in this round as the optimal population to obtain the sparse optimization result of the speaker array.

9. An electronic device, characterized in that: include: Memory for storing computer programs; A processor, configured to implement the steps of the loudspeaker array sparse optimization method based on an improved genetic algorithm according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it is used to implement the steps of the speaker array sparse optimization method based on the improved genetic algorithm according to any one of claims 1 to 7.

Images