Partition control method and device of sound field and electronic equipment

Abstract

The invention relates to a partition control method and device for a sound field and electronic equipment, and the method is applied to a vehicle, and comprises the steps: determining a bright region and a dark region in the sound field of the vehicle, and obtaining a transfer function matrix corresponding to the bright region and a transfer function matrix corresponding to the dark region; based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area, establishing a partition control model of the sound field; determining an optimal solution set corresponding to the multi-objective optimization problem, and determining a target order of a filter based on the optimal solution set corresponding to the multi-objective optimization problem; and determining a driving signal based on the filter under the target order, and outputting the driving signal to a loudspeaker array of the vehicle so as to perform partition control on the sound field of the vehicle. The target order of the filter is related to the multi-target problem, so that the order of the filter is reduced, and computing resources and storage cost of partition control of the sound field are reduced.

Description

Technical Field

[0001] The present invention relates to the field of sound field technology, and in particular to a sound field zoning control method, device and electronic equipment. Background Art

[0002] In the related art, the sound field is usually controlled by combining the pressure matching (PM) algorithm and the acoustic contrast control (ACC) algorithm with a low-order finite impulse response (FIR) filter to achieve the partition control of the sound field. After the FIR filter is designed, the PM algorithm and the ACC algorithm are used to generate the driving signal of the speaker array through the FIR filter, so as to form the desired sound field in a specific area. However, in the above process, the order of the FIR filter is too high, and there are problems such as large computing resources and high storage costs. Summary of the invention

[0003] The present application provides a method, device and electronic device for controlling a sound field in a partitioned manner, so as to solve the problems of large amount of computing resources and high storage cost in the partitioned control of a vehicle sound field in the related art.

[0004] In order to achieve the above-mentioned object, the present application provides a method for controlling a sound field by partition, which is applied to a vehicle, wherein the vehicle is provided with a speaker array, and the speaker array is used to form a sound field of the vehicle, and the method comprises:

[0005] Determine a bright region and a dark region in the sound field of the vehicle, and obtain a transfer function matrix corresponding to the bright region and a transfer function matrix corresponding to the dark region;

[0006] Based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area, a partition control model of the sound field is established; wherein the partition control model of the sound field includes a multi-objective optimization problem;

[0007] Determine an optimal solution set corresponding to the multi-objective optimization problem, and determine a target order of the filter based on the optimal solution set corresponding to the multi-objective optimization problem;

[0008] Based on the filter at the target order, a driving signal is determined, and the driving signal is output to the speaker array of the vehicle to perform zone control on the sound field of the vehicle.

[0009] According to the above-mentioned technical means, on the one hand, a zoning control model of the sound field is established by acquiring the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area. Since the zoning control model of the sound field established in the present application takes into account multiple different objectives that need to be optimized, the accuracy of the zoning control of the sound field is improved; on the other hand, the target order of the filter is determined by the optimal solution set corresponding to the multi-objective optimization problem in the zoning control model of the sound field. Since the target order of the filter is related to the multi-objective optimization problem, the order of the filter is reduced while realizing the zoning control of the sound field. Furthermore, since the order of the filter is reduced, the computing resources of the zoning control of the sound field are reduced, and the storage cost is saved.

[0010] Furthermore, the multi-objective optimization problem includes a first objective function, a second objective function and a third objective function, and a zoning control model of the sound field is established based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area, including: determining the first objective function, the second objective function and the third objective function based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area; wherein the first objective function and the second objective function are used to realize the zoning control of the sound field, and the third objective function is used to reduce the order of the filter; and establishing the zoning control model of the sound field based on the first objective function, the second objective function and the third objective function.

[0011] According to the above technical means, a partition control model of the sound field is established through the first objective function, the second objective function and the third objective function, so that the partition control model of the sound field can not only realize the partition control of the sound field but also reduce the order of the filter, thereby improving the flexibility of the partition control method of the sound field.

[0012] Further, based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area, the first objective function, the second objective function and the third objective function are determined, including: obtaining the number of microphones in the bright area, the desired sound field in the bright area, the number of microphones in the dark area, the number of control frequency points and the weight vector of the control frequency points; determining the first objective function based on the diagonal matrix of the transfer function matrix corresponding to the bright area, the desired sound field in the bright area, the number of control frequency points and the number of microphones in the bright area; determining the second objective function based on the diagonal matrix of the transfer function matrix corresponding to the dark area, the number of control frequency points and the number of microphones in the dark area; determining the third objective function based on the number of control frequency points and the weight vector of the control frequency points.

[0013] According to the above technical means, by obtaining the number of microphones in the bright area, the expected sound field in the bright area, the number of microphones in the dark area, the number of control frequency points and the weight vector of the control frequency points, combined with the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area, the first objective function, the second objective function and the third objective function are determined respectively, thereby improving the accuracy of the first objective function, the second objective function and the third objective function.

[0014] Furthermore, the method also includes: determining a single-objective problem corresponding to the multi-objective optimization problem, and converting the single-objective problem into an equality-constrained optimization problem using an alternating direction multiplier method; decomposing the equality-constrained optimization problem into at least one sub-problem based on Lagrange multipliers and Lagrange variables, and iteratively solving at least one sub-problem to obtain a solution set of at least one sub-problem; determining the optimal solution set corresponding to the multi-objective optimization problem, including: determining the optimal solution set corresponding to the multi-objective optimization problem based on the solution set of at least one sub-problem.

[0015] According to the above technical means, by determining the equality constraint optimization problem corresponding to the multi-objective optimization problem, using the Lagrange multipliers and Lagrange variables of the alternating direction multiplier method, at least one sub-problem is iteratively solved to obtain a solution set of at least one sub-problem, and then the optimal solution set corresponding to the multi-objective optimization problem is determined, which simplifies the solution process of the multi-objective optimization problem and improves the accuracy of the optimal solution set corresponding to the multi-objective optimization problem.

[0016] Furthermore, based on the optimal solution set corresponding to the multi-objective optimization problem, the target order of the filter is determined, including: projecting the optimal solution set corresponding to the multi-objective optimization problem into relevant indicators to obtain a quasi-Pareto front corresponding to the optimal solution set; wherein the relevant indicators include at least one of the following: the acoustic energy contrast between the bright area and the dark area, the normalized error of the sound field reconstruction in the bright area, and the mean of at least one filter order; based on the quasi-Pareto front corresponding to the optimal solution set, the target order of the filter is determined.

[0017] According to the above technical means, by projecting the optimal solution set corresponding to the multi-objective optimization problem into the relevant indicators, the quasi-Pareto front corresponding to the optimal solution set is obtained, and then the target order of the filter is determined according to the quasi-Pareto front corresponding to the optimal solution set. Since different relevant indicators correspond to different measurement standards, the target order of the filter can be determined from multiple dimensions, thereby improving the accuracy of the target order of the filter.

[0018] Furthermore, based on the Pareto-like front corresponding to the optimal solution set, the target order of the filter is determined, including: obtaining actual demand, and determining an ideal solution from the optimal solution set based on the actual demand and the Pareto-like front corresponding to the optimal solution set; and determining the target order of the filter based on the ideal solution.

[0019] According to the above technical means, the ideal solution is determined from the optimal solution set through the Pareto-like front corresponding to the actual demand and the optimal solution set, and then the target order of the filter is determined according to the ideal solution, thereby improving the accuracy of the target order of the filter. At the same time, it allows users to flexibly set the partition control effect of the sound field according to actual needs, thereby improving the user experience.

[0020] Furthermore, when the relevant indicators include the mean of at least one filter order, the method also includes: obtaining at least one filter coefficient before truncation, and determining the total energy corresponding to at least one filter coefficient before truncation; obtaining a preset energy percentage, and truncating the total energy of at least one filter coefficient before truncation according to the preset energy percentage to obtain at least one truncated filter coefficient; and determining at least one filter order based on at least one truncated filter coefficient.

[0021] According to the above-mentioned technical means, the total energy corresponding to at least one filter coefficient before truncation is truncated according to a preset energy percentage to obtain at least one truncated filter coefficient, and then at least one filter order is determined by at least one truncated filter coefficient, thereby improving the accuracy of at least one filter order.

[0022] Furthermore, the vehicle is also provided with a microphone array to obtain a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area, including: obtaining a collection signal of the microphone array and a driving signal of the speaker array; based on the driving signal of the speaker array and the collection signal of the microphone array, obtaining the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area.

[0023] According to the above technical means, the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area are obtained through the driving signal of the speaker array and the collection signal of the microphone array, thereby improving the accuracy of the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area.

[0024] A sound field partition control device is applied to a vehicle. The vehicle is provided with a speaker array, and the speaker array is used to form a sound field of the vehicle. The device comprises:

[0025] A determination unit, used to determine a bright area and a dark area in the sound field of the vehicle, and obtain a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area;

[0026] An establishing unit is used to establish a partition control model of the sound field based on a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area; wherein the partition control model of the sound field includes a multi-objective optimization problem;

[0027] The determination unit is also used to determine the optimal solution set corresponding to the multi-objective optimization problem, and determine the target order of the filter based on the optimal solution set corresponding to the multi-objective optimization problem; based on the filter under the target order, determine the driving signal and output the driving signal to the speaker array of the vehicle to perform zone control on the sound field of the vehicle.

[0028] An electronic device comprises a processor and a memory, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, any one of the above methods is implemented.

[0029] Beneficial effects of this application:

[0030] (1) determining the target order of the filter by using the optimal solution set corresponding to the multi-objective optimization problem in the determined partition control model of the sound field. Since the target order of the filter is related to the multi-objective optimization problem, the order of the filter is reduced while achieving the partition control of the sound field;

[0031] (2) The establishment of the sound field partition control model of the present application takes into account different objectives, improves the accuracy of the sound field partition control, and, because the order of the filter is reduced, reduces the computing resources and storage cost of the sound field partition control;

[0032] (3) A partition control model of the sound field is established through the first objective function, the second objective function and the third objective function, so that the partition control model of the sound field can realize the partition control of the sound field while reducing the order of the filter, thereby improving the flexibility of the partition control method of the sound field;

[0033] (4) using the Lagrange multipliers and Lagrange variables of the alternating direction multiplier method, iteratively solving at least one sub-problem to obtain a solution set of at least one sub-problem, and then determining the optimal solution set corresponding to the multi-objective optimization problem, thereby simplifying the solution process of the multi-objective optimization problem and improving the accuracy of the optimal solution set corresponding to the multi-objective optimization problem;

[0034] (5) Through the Pareto-like front corresponding to the actual demand and the optimal solution set, the ideal solution is determined from the optimal solution set, and then the target order of the filter is determined according to the ideal solution, thereby improving the accuracy of the target order of the filter. At the same time, it allows users to flexibly set the partition control effect of the sound field according to actual needs, thereby improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A schematic diagram of the implementation process of a sound field partition control method provided in an embodiment of the present application Figure 1 ;

[0036] Figure 2A schematic diagram of a speaker array provided in an embodiment of the present application;

[0037] Figure 3 A schematic diagram of a microphone array provided in an embodiment of the present application;

[0038] Figure 4 A schematic diagram of the implementation process of a sound field partition control method provided in an embodiment of the present application Figure 2 ;

[0039] Figure 5 A schematic diagram of the contrast between a bright area and a dark area provided in an embodiment of the present application;

[0040] Figure 6 A schematic diagram of a normalized error of reconstruction of a bright area sound field provided in an embodiment of the present application;

[0041] Figure 7 A schematic diagram of a filter order provided in an embodiment of the present application;

[0042] Figure 8 A schematic diagram of the structure of a sound field zoning control system provided in an embodiment of the present application;

[0043] Fig. 9 A schematic diagram of the structure of a sound field partition control device provided in an embodiment of the present application;

[0044] Fig.10 A hardware entity schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0045] The following will describe the implementation methods of the present application with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be understood that the preferred embodiments are only for illustrating the present application, not for limiting the scope of protection of the present application.

[0046] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application, and thus the drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.

[0047] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0048] In the following description, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0050] The application of zoning control technology of sound field in the intelligent cockpit audio system of automobiles has received widespread attention. Passengers have personalized needs for the acoustic environment in the cabin, and passengers in different seats expect to have different sound fields that do not interfere with each other. The zoning control technology of the sound field generates driving signals for the speaker array by designing filters to form the desired sound field in a specific area, thereby performing zoning control of the sound field. On the premise of meeting the zoning control requirements of the sound field, it is necessary to reduce the computing resources and storage costs of the zoning control system of the sound field in the cabin. The order represents the number of electronic component delays in the filter. The lower the order, the fewer the number of electronic components in the filter and the lower the manufacturing cost. The FIR filter can effectively reduce the complexity and cost of the zoning control system of the sound field.

[0051] In the related art, when performing zoning control of the sound field, the main algorithms used are the PM algorithm and the ACC algorithm, and the other algorithms are improved on the basis of the PM algorithm and the ACC algorithm. Among them, the PM algorithm is committed to accurately synthesizing the required sound field, setting the desired sound field in the dark area to 0, and reconstructing the desired sound field in the bright area and the dark area. The PM algorithm obtains the weight vector of the speaker array by minimizing the error between the desired sound field and the reconstructed sound field. The ACC algorithm is committed to maximizing the acoustic energy contrast between the bright area and the dark area, and optimizing the obtained weight vector of the speaker array so that the sound signal is directed to the bright area and the sound signal is prevented from leaking to the dark area as much as possible. It is worth noting that the driving signal of the speaker array is obtained by convolution operation of the FIR filter with the input signal to be played, and the application of a higher order FIR filter will lead to an increase in the calculation and storage cost of the control system, and will also cause a higher delay in the sound field zoning control system. However, in the process of using the PM algorithm and the ACC algorithm, the order of the FIR filter is too high, and there are problems such as a lot of computing resources and high storage cost.

[0052] The embodiment of the present application provides a method for zoning control of a sound field. On the one hand, a zoning control model of the sound field is established by acquiring a transfer function matrix corresponding to a bright area and a transfer function matrix corresponding to a dark area. Since the zoning control model of the sound field established in the present application takes into account multiple different objectives that need to be optimized, the accuracy of the zoning control of the sound field is improved. On the other hand, the target order of the filter is determined by the optimal solution set corresponding to the multi-objective optimization problem in the zoning control model of the sound field. Since the target order of the filter is related to the multi-objective optimization problem, while realizing the zoning control of the sound field, the order of the filter is also reduced. Furthermore, since the order of the filter is reduced, the computing resources of the zoning control of the sound field are reduced, and the storage cost is saved. The method provided in the embodiment of the present application can be executed by an electronic device, and the electronic device can be a vehicle computer in a vehicle.

[0053] Below, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the drawings in the embodiments of the present application.

[0054] Figure 1 A schematic diagram of the implementation process of a sound field partition control method provided in an embodiment of the present application Figure 1 ,like Figure 1 As shown, the method is applied in a vehicle, the vehicle is provided with a speaker array, the speaker array is used to form a sound field of the vehicle, the method includes steps S101 to S104, wherein:

[0055] Step S101 : determining a bright area and a dark area in the sound field of the vehicle, and obtaining a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area.

[0056] Here, the sound field refers to the area in the medium where sound waves exist. The bright area refers to the area in the vehicle where the sound field is desired to be generated. The bright area can be any suitable area, such as the front area, the main driving area, etc. The dark area refers to the area in the vehicle where the sound field is not desired to be generated. The dark area can be any suitable area, such as the rear area, the co-pilot area, etc.

[0057] The speaker array is an array composed of at least one speaker. The number of speakers in the speaker array can be any suitable size, for example, 10, 12, etc. In some embodiments, the speakers in the speaker array can be located at any suitable position, for example, at the door, at the roof, etc.

[0058] Figure 2 A schematic diagram of a speaker array provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the speaker array is divided into speakers at the door and speakers 22 at the roof, wherein the speakers at the door include 4 speakers, namely: a first speaker 211, a second speaker 212, a third speaker 213 and a fourth speaker 214, and the speakers 22 at the roof include 8 speakers.

[0059] In some embodiments, the speaker array generates a corresponding sound field after receiving a driving signal, and the range covered by the sound field is called the control area of ​​the vehicle. The control area can be divided into an area where a sound field is desired to be generated and an area where a sound field is not desired to be generated, thereby determining a bright area and a dark area in the sound field of the vehicle.

[0060] The transfer function matrix corresponding to the bright area is used to characterize the propagation characteristics of the audio signal from the sound source to the receiving point in the bright area. The transfer function matrix corresponding to the dark area is used to characterize the propagation characteristics of the audio signal from the sound source to the receiving point in the dark area. In some embodiments, the transfer function matrix includes at least one transfer function from a speaker to a microphone.

[0061] The method for obtaining the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area may include but is not limited to: reading the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area from the data stored in the memory, calculating the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area through the acoustic model, etc. For example, the memory may include but is not limited to: non-volatile memory (Read One Momory, ROM), volatile memory (Random Access Memory, RAM), etc. By reading the data in the memory, the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area can be obtained. For another example, the acoustic model is one of the most important parts in the speech recognition system. By establishing the geometric structure of the acoustic model, using finite element analysis (Finite Element Analysis, FEA), boundary element method (Boundary Element Method, BEM), statistical energy analysis (Statistical Energy Analysis, SEA) and other methods, the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area can be obtained. The geometric structure of the acoustic model may include but is not limited to: sound source, sound transmission medium, receiving point, etc.

[0062] In some embodiments, the vehicle is also provided with a microphone array, and the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area can be obtained by acquiring the collection signal of the microphone array and the driving signal of the speaker array, and then based on the driving signal of the speaker array and the collection signal of the microphone array.

[0063] Step S102, based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area, a partition control model of the sound field is established; wherein the partition control model of the sound field includes a multi-objective optimization problem.

[0064] Here, the partition control model of the sound field is a model for realizing the partition control of the sound field and evaluating the performance of the partition control of the sound field. A multi-objective optimization problem refers to a problem that requires minimizing or maximizing two or more objective functions at the same time. In a multi-objective optimization problem, there is usually no single solution that can optimally satisfy all objective functions at the same time, but rather a series of solutions that provide different trade-offs between objective functions.

[0065] In some embodiments, the multi-objective optimization problem may include at least one objective function. The objectives corresponding to different objective functions may be the same or different. The objectives may include, but are not limited to: achieving partition control of the sound field, achieving reduction of the filter order, etc.

[0066] In some embodiments, the partition control model of the sound field can be established by elements in the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area. For example, the diagonal matrix of the transfer function matrix corresponding to the bright area and the diagonal matrix of the transfer function matrix corresponding to the dark area can be read, and then the partition control model of the sound field can be established by the diagonal matrix of the transfer function matrix corresponding to the bright area and the diagonal matrix of the transfer function matrix corresponding to the dark area.

[0067] In some embodiments, the multi-objective optimization problem includes a first objective function, a second objective function, and a third objective function, which can be determined based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area, and then a partition control model of the sound field is established based on the first objective function, the second objective function, and the third objective function. The first objective function and the second objective function are used to achieve partition control of the sound field, and the third objective function is used to reduce the order of the filter.

[0068] Step S103, determining an optimal solution set corresponding to the multi-objective optimization problem, and determining a target order of the filter based on the optimal solution set corresponding to the multi-objective optimization problem.

[0069] Here, the optimal solution set includes at least one optimal solution. In some embodiments, the optimal solution set corresponding to the multi-objective optimization problem may be an effective solution of the partition control model of the sound field. An effective solution refers to a solution that can meet the requirements of the partition control model of the sound field.

[0070] Methods for determining the optimal solution set corresponding to a multi-objective optimization problem may include, but are not limited to: determining the optimal solution set corresponding to a multi-objective optimization problem by using an Elitist Non-Dominated Sorting Genetic Algorithm (NSGA-II) with an elite strategy, determining the optimal solution set corresponding to a multi-objective optimization problem by using a Multi-Objective Particle Swarm Optimization (MOPSO), etc. For example, NSGA-II is a genetic algorithm based on the concept of Pareto (also known as Pareto) optimality. Through improvements such as non-dominated sorting, crowding, and elite strategies, the algorithm's computing speed and robustness are improved, and it can be used to determine the optimal solution set corresponding to a multi-objective optimization problem. For another example, MOPSO is an optimization method that extends the Particle Swarm Optimization (PSO) algorithm and is specifically used to solve multi-objective optimization problems. MOPSO determines the optimal solution set corresponding to a multi-objective optimization problem by simulating the foraging behavior of a flock of birds or a school of fish, combined with the special needs of multi-objective optimization.

[0071] In some embodiments, the single objective problem corresponding to the multi-objective optimization problem can be first determined, and the single objective problem corresponding to the multi-objective optimization problem can be converted into an equality constrained optimization problem using the alternating direction method of multipliers (ADMM). Then, based on Lagrange multipliers and Lagrange variables, the equality constrained optimization problem can be decomposed into at least one sub-problem, and at least one sub-problem can be iteratively solved to obtain a solution set of at least one sub-problem. Finally, based on the solution set of at least one sub-problem, the optimal solution set corresponding to the multi-objective optimization problem can be determined.

[0072] A filter is a frequency selection device that allows specific frequency components in a signal to pass through while greatly attenuating other frequency components. Filters may include, but are not limited to: FIR filters, infinite impulse response (IIR) filters, etc. FIR filters, also known as non-recursive filters, are the most basic components in digital signal processing systems. They can have strict linear phase-frequency characteristics while ensuring arbitrary amplitude-frequency characteristics. IIR filters, also known as recursive filters, have a feedback loop in their structure, and historical outputs participate in feedback, which can achieve better filtering effects. The target order of the filter can be any suitable size, for example, 2022, 2053, etc.

[0073] In some implementations, different optimal solutions have corresponding filter orders, and the filter orders corresponding to different optimal solutions may be the same or different. After determining the optimal solution set corresponding to the multi-objective optimization problem, the filter order corresponding to at least one optimal solution in the optimal solution set may be determined to obtain at least one filter order. A filter order with the smallest filter order is selected from the at least one filter order, and the filter order is used as the target filter order.

[0074] In some embodiments, the optimal solution set corresponding to the multi-objective optimization problem can be projected into relevant indicators to obtain the quasi-Pareto front corresponding to the optimal solution set, and then the target order of the filter is determined based on the quasi-Pareto front corresponding to the optimal solution set. The relevant indicators include at least one of the following: the acoustic energy contrast between the bright area and the dark area, the normalized error of the sound field reconstruction in the bright area, and the mean of at least one filter order.

[0075] Step S104 , determining a driving signal based on the filter at the target order, and outputting the driving signal to the speaker array of the vehicle to perform partition control on the sound field of the vehicle.

[0076] Here, the driving signal is a signal for driving the speakers in the speaker array to output sound. The driving signal can be used to control the vibration of the speaker, the output volume of the speaker, etc. In some embodiments, each speaker in the speaker array can be controlled by a separate driving signal. By controlling the phase and amplitude of the driving signals of different speakers, the partition control of the sound field can be achieved.

[0077] In some embodiments, the signal type of the signal output by the filter is a digital signal, and the speaker can only process signals whose signal type is an analog signal. After determining the target order of the filter, the signal can be filtered by the filter under the target order to remove unnecessary frequency components or adjust the phase and amplitude of the signal according to the partition control requirements of the sound field in the bright area and the dark area, and a processed digital signal is obtained. Then, the processed digital signal is converted into an analog signal through a digital to analog converter (also known as a D / A converter), and the analog signal is used as a driving signal.

[0078] In some embodiments, the filter can be connected to the input end of the speaker through an audio cable, and then the driving signal can be output to the speaker array of the vehicle through the audio cable to perform partition control on the sound field of the vehicle. The audio cable may include, but is not limited to: RCA (Radio Corporation of America) connector, Cannon (Cannon X Series, Latch Rubber, XLR) connector, etc. The RCA connector uses a coaxial signal transmission method, the middle axis is used to transmit the signal, and the contact layer around the outer edge is used for grounding. The XLR connector is used to transmit various audio signals in the audio system. Generally, the balanced input and output terminals are connected using XLR connectors.

[0079] In an embodiment of the present application, on the one hand, a zoning control model of the sound field is established by acquiring a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area. Since the zoning control model of the sound field established in the present application takes into account multiple different objectives that need to be optimized, the accuracy of the zoning control of the sound field is improved. On the other hand, the target order of the filter is determined by determining the optimal solution set corresponding to the multi-objective optimization problem in the zoning control model of the sound field. Since the target order of the filter is related to the multi-objective optimization problem, the order of the filter is reduced while realizing the zoning control of the sound field. Furthermore, since the order of the filter is reduced, the computing resources of the zoning control of the sound field are reduced, and the storage cost is saved.

[0080] In some embodiments, the vehicle is further provided with a microphone array, and the step of "obtaining a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area" in step S101 includes steps S113 and S114, wherein:

[0081] Step S113, acquiring a collection signal of the microphone array and a driving signal of the speaker array.

[0082] Here, the microphone array is an array composed of at least one microphone. The number of microphones in the microphone array can be any suitable size, for example, 12, 16, etc. In some embodiments, at least one microphone array can be provided in a vehicle. The microphones in the microphone array can be located at any suitable position, for example, at a vehicle door, at a vehicle roof, etc.

[0083] Figure 3 A schematic diagram of a microphone array provided in an embodiment of the present application, such as Figure 3 As shown, two microphone arrays are arranged on the roof of the vehicle, wherein a first microphone array 31 is located in the front area of ​​the vehicle and includes 12 microphones, and a second microphone array 32 is located in the rear area of ​​the vehicle and includes 12 microphones.

[0084] The collected signal of the microphone array refers to the sound signal captured by the microphone array. In some embodiments, after the speaker array emits sound according to the input driving signal, the sound emitted by the speaker array can be collected by the microphone array to obtain the collected signal of the microphone array.

[0085] In some implementations, the collection signal of the microphone array and the driving signal of the speaker array may be obtained by reading parameter information of the microphone array and the speaker array.

[0086] Step S114: obtaining a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area based on the driving signal of the speaker array and the collection signal of the microphone array.

[0087] Here, the cross power spectral density between the driving signal of the speaker array and the collected signal of the microphone array can be determined, and then the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area can be obtained according to the cross power spectral density. The cross power spectral density is a method for describing the degree of statistical correlation between two different signals in the frequency domain. In some embodiments, the cross power spectral density can be expressed by the cross correlation function of the two signals. In some embodiments, the cross power spectral density of the two signals can be calculated by the Fast Fourier Transformation (FFT) method.

[0088] In some embodiments, a white noise input signal can be given to the speaker array to obtain the auto-power spectrum density corresponding to the white noise input signal, and then based on the cross-power spectrum density between the driving signal of the speaker array and the collection signal of the microphone array and the auto-power spectrum density corresponding to the white noise input signal, the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area are obtained.

[0089] In some embodiments, the transfer function matrix G can be determined by the following formula (1-1): i (f), that is, refer to formula (1-1):

[0090]

[0091] In formula (1-1), is the transfer function from the lth loudspeaker to the mth microphone in the ith control area, The sound signal collected by the mth microphone in the ith control area and the lth speaker excitation signal x l The cross power spectral density of (t), x l (f)(l=1,...,L) is a white noise input signal x given to the speaker array l (t)(l=1,...,L) corresponding to the auto-power spectral density, G 1 (f), G 2 (f) are the transfer function matrices of the front row area and the back row area, respectively. M is the number of microphones and L is the number of speakers.

[0092] In some implementations, the transfer function matrix G obtained according to formula (1-1) is i (f) Satisfaction in, refers to the complex field, M×L represents the matrix dimension; Represents a set of complex matrices of dimension M×L.

[0093] In some embodiments, the front row area can be used as a bright area, and the back row area can be used as a dark area. Therefore, the transfer function matrix of the front row area is the transfer function matrix corresponding to the bright area, and the transfer function matrix of the back row area is the transfer function matrix corresponding to the dark area. Correspondingly, G B (f) = G 1 (f), G D (f) = G 2 (f). Where G B (f) is the transfer function matrix corresponding to the bright area, G D (f) is the transfer function matrix corresponding to the dark area.

[0094] In an embodiment of the present application, a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area are obtained through the driving signal of the speaker array and the collection signal of the microphone array, thereby improving the accuracy of the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area.

[0095] In some implementations, the multi-objective optimization problem includes a first objective function, a second objective function, and a third objective function, and step S102 includes step S121 and step S122, wherein:

[0096] Step S121, based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area, determine the first objective function, the second objective function and the third objective function; wherein the first objective function and the second objective function are used to achieve partition control of the sound field, and the third objective function is used to reduce the order of the filter.

[0097] Here, the first objective function is used to measure the difference between the reconstructed sound field in the bright area and the expected sound field in the bright area, and the second objective function is used to measure the quality of the reconstructed sound field in the dark area. Therefore, the first objective function and the second objective function can be used to achieve zoning control of the sound field. The third objective function is used to measure the flatness of the loudspeaker output signal, that is, the uniformity of the distribution of the signal in the spectrum. In some embodiments, the smaller the value of the third objective function, the more uniform the distribution of the output signal in the spectrum, and the better the flatness. At the same time, the flatness of the signal is positively correlated with the energy concentration of the corresponding time domain signal. The more concentrated the energy, the smaller the order of the filter. Therefore, the third objective function is used to reduce the order of the filter.

[0098] In some embodiments, the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area include at least one transfer function, and the first objective function, the second objective function and the third objective function can be determined by extracting at least one transfer function.

[0099] In some embodiments, a first correspondence between a transfer function matrix and an objective function may be established, and then the first objective function, the second objective function, and the third objective function corresponding to the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area may be determined through the first correspondence.

[0100] In some embodiments, the target parameters are obtained, and then the diagonal matrix of the transfer function matrix corresponding to the bright area and the diagonal matrix of the transfer function matrix corresponding to the dark area are read, and the first target function, the second target function, and the third target function are determined according to the target parameters and the diagonal matrix of the transfer function matrix corresponding to the bright area and the diagonal matrix of the transfer function matrix corresponding to the dark area. The target parameters may include, but are not limited to: the number of microphones in the bright area, the desired sound field in the bright area, the number of microphones in the dark area, the number of control frequency points, and the weight vectors of the control frequency points.

[0101] Step S122: establishing a partition control model of the sound field based on the first objective function, the second objective function and the third objective function.

[0102] Here, the first objective function, the second objective function and the third objective function may be used as objective functions in a multi-objective optimization problem, and a partition control model of the sound field may be further established according to the multi-objective optimization problem including the first objective function, the second objective function and the third objective function.

[0103] In some implementations, a partition control model of the sound field may be established based on the first objective function, the second objective function, and the third objective function through deep learning or machine learning methods.

[0104] In the implementation mode of the present application, a partition control model of the sound field is established through the first objective function, the second objective function and the third objective function, so that the partition control model of the sound field can realize the partition control of the sound field while reducing the order of the filter, thereby improving the flexibility of the partition control method of the sound field.

[0105] In some implementations, step S121 includes steps S1211 to S1214, wherein:

[0106] Step S1211, obtaining the number of microphones in the bright area, the desired sound field in the bright area, the number of microphones in the dark area, the number of control frequency points, and the weight vectors of the control frequency points.

[0107] Here, the number of microphones in the bright area can be any suitable size, for example, 12, 10, etc. The number of microphones in the dark area can be any suitable size, for example, 12, 10, etc. In some embodiments, the number of microphones in the bright area can be the same as or different from the number of microphones in the dark area. In some embodiments, after determining the bright area and the bright area, the number of microphones in the bright area and the number of microphones in the dark area can be obtained by reading the parameter information of the microphones.

[0108] The expected sound field of the bright area refers to the sound field expected to be formed in the bright area. The expected sound field of the bright area may include but is not limited to: navigation sound, music sound, broadcast sound, etc. In some embodiments, after the bright area is determined, the parameters of the sound field can be set by selecting an operation to obtain the expected sound field of the bright area.

[0109] The control frequency point refers to the value selected in the control frequency range. The number of control frequency points can be any suitable size, for example, 1000, 1200, etc. The control frequency range is a pre-set frequency range, and the control frequency range can be any suitable range, for example, 200Hz (Hertz) ~ 1200Hz, 500Hz ~ 1800Hz, etc. In some embodiments, after obtaining the control frequency range, the control frequency interval can be obtained, and then the control frequency point can be determined according to the control frequency range and the control frequency interval. The control frequency interval refers to the interval between frequency control points. The control frequency interval can be any suitable size, for example, 1Hz, 2Hz, etc. Exemplarily, the control frequency range is 200Hz ~ 1200Hz, and the control frequency interval is 1Hz, then the control frequency points can include but are not limited to: 200Hz, 201Hz, 1200Hz, etc. Correspondingly, the number of control frequency points is 1001. In some embodiments, the control frequency range can be obtained by reading the parameter information of the speaker, and then the number of control frequency points can be obtained.

[0110] The weight vector of the control frequency point refers to the importance of the control frequency point in the multi-objective optimization problem. In some embodiments, the weight vector is positively correlated with the importance of the control frequency point in the multi-objective optimization problem. The larger the modulus of the weight vector, the more important the control frequency point. In some embodiments, the weight vector of the control frequency point may refer to the weight vector of the speaker at all control frequency points. In some embodiments, the weight vector of the control frequency point may be obtained by reading the parameter information of the speaker.

[0111] Step S1212: determining a first objective function based on the diagonal matrix of the transfer function matrix corresponding to the bright area, the desired sound field of the bright area, the number of control frequency points, and the number of microphones in the bright area.

[0112] Here, the diagonal matrix of the transfer function matrix corresponding to the bright area is a matrix formed by the transfer functions on the diagonal of the transfer function matrix corresponding to the bright area. In some embodiments, the first objective function may be the square sum of the errors between the reconstructed sound field of the bright area and the expected sound field of the bright area.

[0113] In some implementations, the first objective function A may be determined by the following formula (1-2), that is, see formula (1-2):

[0114] A=||p B -p BT || 2 (1-2);

[0115] In formula (1-2), p B is the reconstructed sound field in the bright area, where M B is the number of bright area microphones (also known as the total number of bright area control points), J is the number of control frequency points, is the field of complex numbers, M B J is a constant (positive integer), Represents an M B · J-dimensional complex vector, is the loudspeaker array weight vector for all frequency points, p BT is the expected sound field in the bright area, {f j |j=1,…,J} is the control frequency bin, H B is the diagonal matrix of the transfer function matrix of all control frequency points in the bright area. More specifically, H B =diag{G B (f1),...,G B (f J )}.

[0116] In some implementations, the desired sound field in the bright area region is generally set to a unit pulse signal.

[0117] In some implementations, the control frequency bin is a set of all control frequency points. For example, the control frequency range is 200 Hz to 1200 Hz, and the control frequency interval is 1 Hz; then, the control frequency bin may be {200 Hz, 201 Hz, ..., 1200 Hz}.

[0118] Step S1213: determining a second objective function based on the diagonal matrix of the transfer function matrix corresponding to the dark area, the number of control frequency points, and the number of microphones in the dark area.

[0119] Here, the diagonal matrix of the transfer function matrix corresponding to the dark region is a matrix formed by the transfer functions on the diagonal line of the transfer function matrix corresponding to the dark region. In some embodiments, the second objective function may be the square sum of the reconstructed sound field of the dark region.

[0120] In some implementations, the second objective function B may be determined by the following formula (1-3), that is, see formula (1-3):

[0121] B=||p D || 2 (1-3);

[0122] In formula (1-3), p D is the reconstructed sound field in the dark area, where M D is the number of dark area microphones (also known as the total number of dark area control points), H D is the diagonal matrix of the transfer function matrix of all frequency control points in the dark region. More specifically, H D =diag{G D (f1),...,G D (f J )}.

[0123] Step S1214: determining a third objective function based on the number of control frequency points and the weight vectors of the control frequency points.

[0124] Here, the third objective function may be a sum term of absolute values ​​of differences between adjacent frequency points of the speaker array weight vector.

[0125] In some implementations, the third objective function C may be determined by the following formula (1-4), that is, see formula (1-4):

[0126] C = ||Fq||1 (1-4);

[0127] In formula (1-4), Among them, q l =(q l (f1),...,q l (f J )) is the weight vector of the lth speaker at all control frequency points, I L is the L-dimensional identity matrix, is the Kronecker product (also known as Kronecker product), is the difference matrix, is the real number field, D is a (J-1)×J dimensional matrix in the real number field, D ij is the element in the i-th row and j-th column of matrix D,

[0128] In some embodiments, after determining the first objective function, the second objective function, and the third objective function, the multi-objective optimization problem D may be determined by the following formula (1-5), that is, see formula (1-5):

[0129] D=min q {A, B, C} (1-5);

[0130] In formula (1-5), is the speaker array weight vector of all frequency points, A is the first objective function, B is the second objective function, and C is the third objective function.

[0131] In an embodiment of the present application, by acquiring the number of microphones in the bright area, the expected sound field in the bright area, the number of microphones in the dark area, the number of control frequency points and the weight vector of the control frequency points, combined with the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area, the first objective function, the second objective function and the third objective function are determined respectively, thereby improving the accuracy of the first objective function, the second objective function and the third objective function.

[0132] In some embodiments, the method further includes steps S105 and S106, and the "determining the optimal solution set corresponding to the multi-objective optimization problem" in step S103 includes step S103a, wherein:

[0133] Step S105, determining the single objective problem corresponding to the multi-objective optimization problem, and converting the single objective problem into an equality constrained optimization problem by using an alternating direction multiplier method.

[0134] Here, the weight method can be used to assign weights to the objective functions in the multi-objective optimization problem, and the objective functions with assigned weights can be summed up to determine the single objective problem corresponding to the multi-objective optimization problem.

[0135] In some implementations, the single objective problem E may be determined by the following formula (1-6), that is, see formula (1-6):

[0136] E=min q (δ1A+δ2B+δ3C) (1-6);

[0137] In formula (1-6), for any given δ=(δ1,δ2,δ3)∈Λ, where the simplex Λ is, Λ={δ|δ k >0(k=1,2,3),δ1+δ2+δ3=1}, A is the first objective function, B is the second objective function, C is the third objective function, is the speaker array weight vector for all frequency points.

[0138] In some embodiments, δ should take all values ​​on Λ, but δ is an infinite set and it is actually impossible to obtain a complete optimal solution set. Therefore, we use average sampling to select a finite subset of Λ + , δ can uniformly select points on the simplex Λ according to the preset interval, and then obtain a finite subset Λ of Λ + The preset spacing can be any suitable size, for example, 0.02, 0.04, etc. For example, δ uniformly selects points on the simplex Λ at a spacing of 0.02, and the selected set Λ + is: + ={δ|δk =0.02t (k=1, 2, 3; t=1, .., 50), δ1+δ2+δ3=1}.

[0139] ADMM is an algorithm for solving optimization problems, especially those with separable structure. The core of the ADMM algorithm is the augmented Lagrangian method of the primal-dual algorithm.

[0140] Step S106, based on Lagrange multipliers and Lagrange variables, decompose the equality constraint optimization problem into at least one sub-problem, and iteratively solve the at least one sub-problem to obtain a solution set of the at least one sub-problem;

[0141] Here, ADMM can quickly calculate the single-objective problem in the above formula (1-6) and obtain the optimal solution set corresponding to the multi-objective optimization problem.

[0142] In some embodiments, ADMM decomposes the single-objective problem into simpler sub-problems for solution by introducing Lagrange multipliers and Lagrange variables. This decomposition method helps to reduce the complexity of the single-objective problem and makes each sub-problem easier to solve. At the same time, the introduction of Lagrange multipliers helps to deal with constraints and ensure that the optimal solution of the constraints can be gradually achieved each time. At the same time, since the sub-problems obtained by ADMM have analytical solutions, they can be summarized and solved, thereby accelerating the solution process of the single-objective problem.

[0143] In some embodiments, auxiliary variables may be introduced to transform the single objective problem in the above formula (1-6) into a constrained optimization problem. The main purpose of introducing auxiliary variables is to decompose the single objective problem into two simpler parts so that its variables can be separated. This decomposition method helps to reduce the complexity of the single objective problem and makes each sub-problem easier to solve.

[0144] In some implementations, the constrained optimization problem H may be determined by the following formula (1-7), that is, see formula (1-7):

[0145] H=min q (δ1A+δ2B+δ3||z||1)

[0146] st z=Fq (1-7);

[0147] In formula (1-7), z=Fq is an auxiliary variable, is the loudspeaker array weight vector of all frequency points, and st represents the constraint condition of the constrained optimization problem H as z=Fq.

[0148] In some embodiments, for the constrained optimization problem in formula (1-7), an augmented Lagrangian function of the equality constrained optimization problem may be introduced, and the augmented Lagrangian function may be determined by the following formula (1-8): Refer to formula (1-8):

[0149]

[0150] In formula (1-8), z=Fq is an auxiliary variable, is the loudspeaker array weight vector for all frequency points, μ is the Lagrange multiplier, and ρ is the penalty coefficient, where ρ>0.

[0151] In some embodiments, the augmented Lagrangian function can be determined according to the above formula (1-8): By iteratively solving the constrained optimization problem in formula (1-7) through ADMM, the iterative update formula of the constrained optimization problem in formula (1-7) can be determined by the following formula (1-9), that is, see formula (1-9):

[0152]

[0153] In formula (1-9), τ is taken as 1.5 as the iteration step, ρ is taken as 1 as the penalty coefficient, and the initial value of the iteration is (q 0 , z 0 , μ 0 ), the iteration termination condition is: err<0.01, where err=||q k+1 -q k || 2 +||z k+1 -z k || 2 +||μ k+1 -μ k || 2 .

[0154] Step S103a, based on the solution set of at least one sub-problem, determine the optimal solution set corresponding to the multi-objective optimization problem.

[0155] Here, since the sub-problems obtained by ADMM have analytical solutions, the solution set of at least one sub-problem can be spliced ​​together to determine the optimal solution set corresponding to the multi-objective optimization problem, thereby accelerating the solution process of the entire problem.

[0156] In an implementation manner of the present application, by determining the equality constrained optimization problem corresponding to the multi-objective optimization problem, using the Lagrange multipliers and Lagrange variables of the alternating direction multiplier method, at least one sub-problem is iteratively solved to obtain a solution set of at least one sub-problem, and then the optimal solution set corresponding to the multi-objective optimization problem is determined, thereby simplifying the solution process of the multi-objective optimization problem and improving the accuracy of the optimal solution set corresponding to the multi-objective optimization problem.

[0157] In some implementations, the step of “determining the target order of the filter based on the optimal solution set corresponding to the multi-objective optimization problem” in step S103 includes step S103b and step S103c, wherein:

[0158] Step S103b, projecting the optimal solution set corresponding to the multi-objective optimization problem into relevant indicators to obtain a quasi-Pareto front corresponding to the optimal solution set; wherein the relevant indicators include at least one of the following: acoustic energy contrast between the bright area and the dark area, normalized error of sound field reconstruction in the bright area, and mean value of at least one filter order.

[0159] Here, projecting the optimal solution set corresponding to the multi-objective optimization problem into the relevant indicators refers to the process of substituting the optimal solution set corresponding to the multi-objective optimization problem into the definition formula of each relevant indicator. In some embodiments, the optimal solution set corresponding to the multi-objective optimization problem is Among them, q # (δ)=(q(f1),...,q(f J )). When q # (δ) is obtained, then q(f i )(i=1,...,J), and then q(f i ) is substituted into the definition formula of each relevant indicator to complete the projection.

[0160] The quasi-Pareto frontier (also known as the quasi-Pareto frontier) refers to a set of optimal solutions that achieve the best trade-off between multiple objectives. The quasi-Pareto frontier can more intuitively evaluate the partition control performance of the sound field and the effect of reducing the filter order.

[0161] When sound waves propagate in a medium, on the one hand, they cause the particles of the medium to reciprocate near the equilibrium position to generate kinetic energy, and on the other hand, they cause the medium to undergo a process of compression and expansion, which gives the medium potential energy of deformation. The sum of these two parts of energy is the sound energy obtained by the medium due to the sound vibration. The sound energy can be of any appropriate size, for example, 100J (joules), 500J, etc.

[0162] The acoustic energy contrast between the bright area and the dark area refers to the ratio between the acoustic energy of the bright area and the acoustic energy of the dark area. The acoustic energy contrast between the bright area and the dark area can be any suitable value, for example, 30, 21, etc. In some embodiments, the greater the acoustic energy contrast between the bright area and the dark area, the better the partitioning effect of characterizing the sound field.

[0163] In some embodiments, the acoustic energy contrast AC(f) may be determined by the following formula (1-10), that is, see formula (1-10):

[0164]

[0165] In formula (1-10), M B is the number of microphones in the bright area, M D is the number of microphones in the dark area, and q(f) is the optimal solution corresponding to any frequency.

[0166] In some embodiments, when calculating the acoustic energy contrast between the bright area and the dark area, only the points in the control frequency bin, i.e., the control frequency points, may be calculated. After calculating the acoustic energy contrast between the bright area and the dark area at multiple control frequency points, the average of the acoustic energy contrast between the bright area and the dark area within the control frequency range is calculated, so as to better measure the partitioning effect of the sound field. For example, q(f i )(i=1,...,J) is substituted into the contrast definition formula AC(f) to obtain the frequency f i (i=1,...,J) under the corresponding contrast index AC(f i ), and then {AC(f i By taking the arithmetic average of )(i=1,...,J)}, the mean value of the sound energy contrast between the bright and dark areas within the control frequency range can be obtained.

[0167] The normalized error of the sound field reconstruction in the bright area refers to the difference between the reconstructed sound field in the bright area and the sound field in the bright area before reconstruction. The normalized error of the sound field reconstruction in the bright area can be any suitable size, for example, 0.1, 0.03, etc. In some embodiments, the larger the normalized error of the sound field reconstruction in the bright area, the smaller the difference between the reconstructed sound field in the bright area and the sound field in the bright area before reconstruction, indicating that the sound quality of the reconstructed sound field in the bright area is higher.

[0168] In some embodiments, the normalized error E of the sound field reconstruction in the bright area can be determined by the following formula (1-11): r (f), that is, see formula (1-11):

[0169]

[0170] In formula (1-11), p BT (f) is the expected sound field in the bright area at frequency f, where p BT =(p BT (f1),...,p BT (f J )), p BT (f i ) represents the control frequency point f i The desired sound field in the lower bright area.

[0171] In some implementations, when calculating the normalized error of the sound field reconstruction in the bright area, only the points in the control frequency bin, i.e., the control frequency points, may be calculated. After calculating the normalized error of the sound field reconstruction in the bright area at multiple control frequency points, the average of the normalized error of the sound field reconstruction in the bright area within the control frequency range is calculated, so as to better measure the sound quality of the reconstructed sound field in the bright area.

[0172] The filter order may be any suitable size, for example, 1450, 2560, etc. In some implementations, the smaller the filter order, the faster the partition control of the sound field can be achieved, and correspondingly, the smaller the computing resources required to achieve the partition control of the sound field.

[0173] In some implementations, different optimal solutions in the optimal solution set corresponding to the multi-objective optimization problem may correspond to different filter orders. Therefore, after determining the optimal solution set corresponding to the multi-objective optimization problem, at least one filter order may be obtained. The arithmetic mean of at least one filter order may be taken to obtain the mean of at least one filter order.

[0174] In some embodiments, at least one filter coefficient before truncation can be obtained first, and the total energy corresponding to the at least one filter coefficient before truncation can be determined. Further, a preset energy percentage can be obtained, and the total energy of at least one filter coefficient before truncation can be truncated according to the preset energy percentage to obtain at least one truncated filter coefficient. Finally, based on the at least one truncated filter coefficient, at least one filter order can be determined.

[0175] Step S103c, determining the target order of the filter based on the Pareto-like front corresponding to the optimal solution set.

[0176] Here, after obtaining the Pareto-like front corresponding to the optimal solution set, the optimal solutions in the optimal solution set corresponding to the multi-objective optimization problem can be sorted under different related indicators to obtain an optimal solution set with a certain order under multiple different related indicators, and then the filter order corresponding to the optimal solution with the highest priority can be selected as the target order of the filter.

[0177] In some embodiments, a second correspondence between a quasi-Pareto front and a filter order can be established. After determining the quasi-Pareto front corresponding to the optimal solution set, the filter order corresponding to the quasi-Pareto front can be determined according to the second correspondence, and the filter order can be used as the target order of the filter.

[0178] In an implementation manner of the present application, the optimal solution set corresponding to the multi-objective optimization problem is projected into relevant indicators to obtain the quasi-Pareto front corresponding to the optimal solution set, and then the target order of the filter is determined according to the quasi-Pareto front corresponding to the optimal solution set. Since different relevant indicators correspond to different measurement standards, the target order of the filter can be determined from multiple dimensions, thereby improving the accuracy of the target order of the filter.

[0179] In some implementations, step S103c includes step S103c1 and step S103c2, wherein:

[0180] Step S103c1, obtaining actual demand, and determining an ideal solution from the optimal solution set based on the Pareto-like front corresponding to the actual demand and the optimal solution set.

[0181] Here, actual demand refers to the goal that the zoning control of the sound field focuses on achieving. The actual demand may include but is not limited to: pursuing the zoning effect of the sound field, pursuing the sound quality of the reconstructed sound field, pursuing faster zoning control of the sound field, etc. For example, when passengers listen to music, they often pursue the listening experience more. At this time, the actual demand may be the pursuit of the sound quality of the reconstructed sound field.

[0182] In some embodiments, different actual needs have corresponding related indicators. After determining the actual needs, the corresponding class Pareto front corresponding to the optimal solution set under the related indicators can be obtained according to the related indicators corresponding to the actual needs, and then the priority solution with the highest order in the class Pareto front can be selected as the ideal solution.

[0183] In some implementations, the ideal solutions corresponding to different actual requirements may be the same or different.

[0184] Step S103c2, determining the target order of the filter based on the ideal solution.

[0185] Here, after determining the ideal solution, the filter order corresponding to the ideal solution can be obtained, and then the target order of the filter can be determined according to the filter order corresponding to the ideal solution. In some embodiments, the target order of the filter can be determined by the filter order corresponding to the ideal solution, the weighting / logarithm / index of the filter order corresponding to the ideal solution, etc. For example, the filter order corresponding to the ideal solution is used as the target order of the filter.

[0186] In some embodiments, the ideal solution in the frequency domain can be * The filter coefficients are obtained by inverse Fourier transform to the time domain, thereby obtaining a lower-order filter. * After determination, the corresponding filter order K l The filter coefficients are also determined accordingly. The filter coefficients are:

[0187] In the implementation mode of the present application, an ideal solution is determined from the optimal solution set through a Pareto-like front corresponding to actual needs and an optimal solution set, and then the target order of the filter is determined according to the ideal solution, thereby improving the accuracy of the target order of the filter. At the same time, the user can flexibly set the zoning control effect of the sound field according to actual needs, thereby improving the user experience.

[0188] In some embodiments, when the relevant indicator includes a mean value of at least one filter order, the method further includes steps S107 to S109, wherein:

[0189] Step S107, obtaining at least one filter coefficient before truncation, and determining the total energy corresponding to the at least one filter coefficient before truncation.

[0190] Here, the order of the filter is usually equal to the number of filter coefficients minus 1. After the loudspeaker driving signal in the frequency domain is obtained by the sound field partition control algorithm, it needs to be converted to the time domain to obtain the filter coefficient vector before truncation.

[0191] In some embodiments, q l (l = 1, ..., L) is the weight vector of the lth loudspeaker in the frequency domain obtained by the sound field partition control algorithm, and its control frequency range is [f L , f H ], the control frequency interval is Δf, and the sampling frequency of the control system is f s To quickly calculate the filter coefficients before truncation, q l Perform zero padding to obtain in, The zero padding operation is specifically performed in the frequency range The remaining control frequency points in the range are filled with 0 to form q l A symbiotic vector

[0192] In some embodiments, the symbiosis vector can be determined by the following formula (1-12): Refer to formula (1-12):

[0193]

[0194] In formula (1-12), [f L , f H ] is the control frequency range, f s is the sampling frequency of the control system.

[0195] In some embodiments, the symbiotic vector in the above formula (1-12) can be Perform fast Fourier transform to obtain the filter coefficient vector before truncation is the field of real numbers.

[0196] Step S108, obtaining a preset energy percentage, and truncating the total energy of at least one filter coefficient before truncation according to the preset energy percentage to obtain at least one filter coefficient after truncation.

[0197] Here, since the filter coefficients before truncation may contain some low-amplitude coefficients, the filter coefficients with larger values ​​in the middle part can be truncated, thereby reducing the computation and storage overhead while retaining the main characteristics of the filter. The preset energy percentage can be any suitable size, for example, 99%, 98%, etc.

[0198] In some embodiments, the untruncated filter coefficient vector F can be determined by the following formula (1-13): l The total energy of ||F l || 2 , that is, see formula (1-13):

[0199]

[0200] In formula (1-13), is the field of real numbers.

[0201] In some embodiments, when the total energy of at least one filter coefficient before truncation is truncated, the interval Medium energy accounts for F l Total energy||F l || 2 The ratio C l (K), determine the preset energy percentage, where, Refers to the set of natural numbers.

[0202] In some embodiments, the interval can be determined by the following formula (1-14): Medium energy accounts for F l Total energy||F l || 2 The ratio C l (K), that is, refer to formula (1-14):

[0203]

[0204] In formula (1-14), ||F l || 2 is the filter coefficient vector F before truncation l The total energy, is the field of real numbers.

[0205] Step S109: determining at least one filter order based on at least one truncated filter coefficient.

[0206] Here, since at least one truncated filter coefficient can retain the main characteristics of the filter, the final filter can still accurately achieve the partition control of the sound field under the premise of reducing the order.

[0207] In some implementations, the filter order K can be determined by the following formula (1-15): l , that is, see formula (1-15):

[0208]

[0209] In formula (1-15), C l (K) is the interval segment Medium energy accounts for F l Total energy||F l || 2 At the same time, the interval contains 2K+1 positive integers, that is, the number of filter coefficients is 2K+1. Since the filter order is usually equal to the number of filter coefficients minus one, the filter order K here is l is 2K.

[0210] In an embodiment of the present application, the total energy corresponding to at least one filter coefficient before truncation is truncated according to a preset energy percentage to obtain at least one truncated filter coefficient, and then at least one filter order is determined by the at least one truncated filter coefficient, thereby improving the accuracy of at least one filter order.

[0211] In some embodiments, the step of “determining a bright area and a dark area in the sound field of the vehicle” in step S101 includes steps S111 and S112, wherein:

[0212] Step S111 : determining an area where a sound field is desired to be generated and an area where a sound field is not desired to be generated based on a sound field division requirement.

[0213] Here, the sound field division requirement represents the division of the sound field. The sound field division requirement may include but is not limited to: the front row area is expected to have sound, the front passenger area is expected to be quiet, etc.

[0214] In some embodiments, based on the need to divide the sound field, areas in the vehicle where sound is desired to appear and areas in the vehicle where sound is not desired to appear can be determined, and then the areas in the vehicle where sound is desired to appear are used as areas where a sound field is desired to be generated, and the areas in the vehicle where sound is not desired to appear are used as areas where a sound field is not desired to be generated.

[0215] In some embodiments, the control area that can be covered by the sound field of the vehicle can be presented on the central control screen of the vehicle, and the division requirements of the sound field can be determined through preset operations. The preset operations may include but are not limited to: single click, double click, circle selection, etc.

[0216] Step S112: The area where the sound field is desired to be generated is regarded as a bright area, and the area where the sound field is not desired to be generated is regarded as a dark area.

[0217] Here, the user can select the control area as a bright area or a dark area according to the light and dark area custom display screen, and use the area where the sound field is desired to be generated as the bright area, and the area where the sound field is not desired to be generated as the dark area.

[0218] In the implementation manner of the present application, based on the need to divide the sound field, the area where the sound field is desired to be generated is regarded as a bright area, and the area where the sound field is not desired to be generated is regarded as a dark area. The bright area and the dark area can be accurately divided according to user needs, thereby improving the accuracy of the bright area and the dark area.

[0219] The following describes the application of the sound field zoning control method provided in an embodiment of the present application in an actual scenario, taking a vehicle provided with a microphone array and a speaker array as an example.

[0220] The application of zoning control technology of sound field in the intelligent cockpit audio system of automobiles has received widespread attention. Passengers have personalized needs for the acoustic environment in the cabin, and passengers in different seats expect to have different sound fields that do not interfere with each other. The zoning control technology of the sound field generates driving signals for the speaker array by designing filters to form the desired sound field in a specific area, thereby performing zoning control of the sound field. On the premise of meeting the zoning control requirements of the sound field, it is necessary to reduce the computing resources and storage costs of the zoning control system of the sound field in the cabin. The order represents the number of electronic component delays in the filter. The lower the order, the fewer the number of electronic components in the filter and the lower the manufacturing cost. The FIR filter can effectively reduce the complexity and cost of the zoning control system of the sound field.

[0221] In the related art, when performing zoning control of the sound field, the main algorithms used are the PM algorithm and the ACC algorithm, and the other algorithms are improved on the basis of the PM algorithm and the ACC algorithm. Among them, the PM algorithm is committed to accurately synthesizing the required sound field, setting the desired sound field in the dark area to 0, and reconstructing the desired sound field in the bright area and the dark area. The PM algorithm obtains the weight vector of the speaker array by minimizing the error between the desired sound field and the reconstructed sound field. The ACC algorithm is committed to maximizing the acoustic energy contrast between the bright area and the dark area, and optimizing the obtained weight vector of the speaker array so that the sound signal is directed to the bright area and the sound signal is prevented from leaking to the dark area as much as possible. It is worth noting that the driving signal of the speaker array is obtained by convolution operation of the FIR filter with the input signal to be played, and the application of a higher order FIR filter will lead to an increase in the calculation and storage cost of the control system, and will also cause a higher delay in the sound field zoning control system. However, in the process of using the PM algorithm and the ACC algorithm, the order of the FIR filter is too high, and there are problems such as a lot of computing resources and high storage cost.

[0222] An embodiment of the present application provides a method for zoning control of a sound field. On the one hand, a zoning control model of the sound field is established by acquiring a transfer function matrix corresponding to a bright area and a transfer function matrix corresponding to a dark area. Since the zoning control model of the sound field established in the present application takes into account multiple different objectives that need to be optimized, the accuracy of the zoning control of the sound field is improved. On the other hand, a target order of a filter is determined by determining an optimal solution set corresponding to a multi-objective optimization problem in the zoning control model of the sound field. Since the target order of the filter is related to the multi-objective optimization problem, while realizing the zoning control of the sound field, the order of the filter is also reduced. Furthermore, since the order of the filter is reduced, the computing resources for the zoning control of the sound field are reduced, and the storage cost is saved.

[0223] Figure 4 A schematic diagram of the implementation process of a sound field partition control method provided in an embodiment of the present application Figure 2 ,like Figure 4 As shown, the method includes steps S201 to S208, wherein:

[0224] Step S201, obtaining a collection signal of the microphone array and a driving signal of the speaker array, and measuring a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area based on the driving signal of the speaker array and the collection signal of the microphone array;

[0225] Step S202, determining a first objective function, a second objective function, and a third objective function based on a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area;

[0226] Step S203, establishing a partition control model of the sound field including a multi-objective optimization problem based on the first objective function, the second objective function and the third objective function;

[0227] Step S204, determining a single objective problem corresponding to the multi-objective optimization problem, and converting the single objective problem into an equality constrained optimization problem using an alternating direction multiplier method;

[0228] Step S205, based on Lagrange multipliers and Lagrange variables, decompose the equality constraint optimization problem into at least one sub-problem, and iteratively solve the at least one sub-problem to determine the optimal solution set corresponding to the multi-objective optimization problem;

[0229] Step S206, projecting the optimal solution set corresponding to the multi-objective optimization problem into the relevant indicators to obtain the Pareto-like frontier corresponding to the optimal solution set;

[0230] Step S207, obtaining actual demand, and determining an ideal solution from the optimal solution set based on the Pareto-like frontier corresponding to the actual demand and the optimal solution set;

[0231] Step S208, based on the ideal solution, determine the target order of the filter through inverse Fourier transform, and based on the filter at the target order, determine the driving signal and output the driving signal to the speaker array of the vehicle to perform zone control on the sound field of the vehicle.

[0232] In some embodiments, in order to verify the partition control performance and filter order reduction effect of the partition control method of the sound field provided by the present application, the partition control method of the sound field of the present application is compared with the PM algorithm. The specific arrangement of the speaker array and the microphone array is as follows: There are two microphone arrays, which are respectively arranged in the front area and the rear area of ​​the vehicle, and each microphone array consists of 12 microphones (refer to the aforementioned Figure 3 ), there is a microphone array, the speaker array consists of 12 speakers (refer to the above Figure 2 ). The specific situation of the frequency control point is as follows: the control frequency range is 200Hz~1200Hz, the control frequency interval is 1Hz, and the sampling frequency is 48kHz. At the same time, it is assumed that the front area is set as the bright area and the back area is set as the dark area. In this specific case, the comparison between the partition control method of the sound field provided by this application and the PM algorithm is as follows:

[0233] Figure 5 A schematic diagram of the contrast between a bright area and a dark area provided in an embodiment of the present application is shown in FIG. Figure 5As shown, 51 is the contrast curve of the bright area and the dark area of ​​the sound field partition control method provided by the present application, 52 is the contrast curve of the bright area and the dark area using the PM algorithm, and 53 is the expected value of the contrast between the bright area and the dark area. In the range of 200Hz to 1200Hz, the average value of the contrast between the bright area and the dark area of ​​the PM algorithm is 21.8dB (decibel), and the average value of the contrast between the bright area and the dark area of ​​the sound field partition control method provided by the present application is 23.9dB. The contrast between the bright area and the dark area of ​​the sound field partition control method provided by the present application is better than that of the PM algorithm.

[0234] Figure 6 A schematic diagram of a normalized error of reconstruction of a bright area sound field provided in an embodiment of the present application is shown as follows: Figure 6 As shown, 61 is the reconstruction normalized error curve of the bright area sound field of the partition control method of the sound field provided by the present application, 62 is the reconstruction normalized error curve of the bright area sound field using the PM algorithm, and 63 is the expected value of the reconstruction normalized error of the bright area sound field. In the range of 200Hz to 1200Hz, the average value of the reconstruction normalized error of the bright area sound field of the PM algorithm is 0.023, and the average value of the reconstruction normalized error of the bright area sound field of the partition control method of the sound field provided by the present application is 0.061. The reconstruction normalized error result of the bright area sound field of the partition control method of the sound field provided by the present application is slightly worse than that of the PM algorithm.

[0235] Figure 7 A schematic diagram of a filter order provided in an embodiment of the present application, such as Figure 7 As shown, 71 is at least one filter order of the partitioned control method of the sound field provided in the present application, and 72 is at least one filter order using the PM algorithm. The average value of at least one filter order of the partitioned control method of the sound field provided in the present application is 3313, and the average value of at least one filter order of the PM algorithm is 2021. Compared with the PM algorithm, the average value of at least one filter order of the partitioned control method of the sound field provided in the present application is reduced by 1292.

[0236] In summary, the sound field zoning control method provided by the present application has good sound field zoning control performance, and at the same time, it greatly reduces the order of the filter; compared with the PM algorithm, the sound field zoning control method provided by the present application sacrifices the lower reconstruction normalization error of the bright area sound field, improves the contrast between the bright area and the dark area and reduces the filter order.

[0237] Figure 8 A schematic diagram of the structure of a sound field partition control system provided in an embodiment of the present application is shown in FIG. Figure 8As shown, the zoning control system 80 of the sound field includes a measuring unit 81, a user interface 82, a control integration unit 83 and a signal processing unit 84, wherein:

[0238] The measuring unit 81 is composed of a speaker array and a microphone array, and is used to receive the driving signal of the speaker array and the collection signal of the microphone array as input, obtain the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area after processing, and input the obtained transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area to the control integration unit 83;

[0239] The user interface 82 is used to provide the sound field division requirements to the control integration unit 83, for example, setting the front row area as the area for generating a specific sound field (i.e., the bright area) and setting the back row area as the quiet area (i.e., the dark area);

[0240] The control integration unit 83 is used to simultaneously receive the sound field division requirement provided by the user interface 82, and the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area provided by the measurement unit 81, and then output the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area that meet the sound field division requirement to the signal processing unit 84;

[0241] The signal processing unit 84 is used to output the driving signal of the speaker array divided into the bright area and the dark area to the speaker array according to the information provided by the control integration unit 83.

[0242] Based on the above embodiments, the embodiments of the present application further provide a sound field partition control device, which is applied to a vehicle. The vehicle is provided with a speaker array, and the speaker array is used to form a sound field of the vehicle. Fig. 9 A schematic diagram of the structure of a sound field partition control device provided in an embodiment of the present application is shown in FIG. Fig. 9 As shown, the sound field partition control device 900 includes a determination unit 901 and an establishment unit 902, wherein:

[0243] A determination unit 901 is used to determine a bright area and a dark area in the sound field of the vehicle, and obtain a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area;

[0244] An establishing unit 902 is used to establish a partition control model of the sound field based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area; wherein the partition control model of the sound field includes a multi-objective optimization problem;

[0245] The determination unit 901 is also used to determine the optimal solution set corresponding to the multi-objective optimization problem, and determine the target order of the filter based on the optimal solution set corresponding to the multi-objective optimization problem; based on the filter under the target order, determine the driving signal and output the driving signal to the speaker array of the vehicle to perform zone control on the sound field of the vehicle.

[0246] In some embodiments, the multi-objective optimization problem includes a first objective function, a second objective function, and a third objective function. The establishment unit 902 is also used to determine the first objective function, the second objective function, and the third objective function based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area; wherein the first objective function and the second objective function are used to achieve zoning control of the sound field, and the third objective function is used to reduce the order of the filter; based on the first objective function, the second objective function, and the third objective function, a zoning control model of the sound field is established.

[0247] In some embodiments, establishing unit 902 is also used to obtain the number of microphones in the bright area, the desired sound field in the bright area, the number of microphones in the dark area, the number of control frequency points, and the weight vector of the control frequency points; determine the first objective function based on the diagonal matrix of the transfer function matrix corresponding to the bright area, the desired sound field in the bright area, the number of control frequency points, and the number of microphones in the bright area; determine the second objective function based on the diagonal matrix of the transfer function matrix corresponding to the dark area, the number of control frequency points, and the number of microphones in the dark area; determine the third objective function based on the number of control frequency points and the weight vector of the control frequency points.

[0248] In some embodiments, the determination unit 901 is also used to determine the single objective problem corresponding to the multi-objective optimization problem, and use the alternating direction multiplier method to convert the single objective problem into an equality constrained optimization problem; based on Lagrange multipliers and Lagrange variables, the equality constrained optimization problem is decomposed into at least one sub-problem, and at least one sub-problem is iteratively solved to obtain a solution set of at least one sub-problem; based on the solution set of at least one sub-problem, the optimal solution set corresponding to the multi-objective optimization problem is determined.

[0249] In some embodiments, the determination unit 901 is also used to project the optimal solution set corresponding to the multi-objective optimization problem into relevant indicators to obtain a quasi-Pareto front corresponding to the optimal solution set; wherein the relevant indicators include at least one of the following: the acoustic energy contrast between the bright area and the dark area, the normalized error of the sound field reconstruction in the bright area, and the mean of at least one filter order; based on the quasi-Pareto front corresponding to the optimal solution set, the target order of the filter is determined.

[0250] In some implementations, the determination unit 901 is further used to obtain actual demand, and determine an ideal solution from the optimal solution set based on the Pareto-like front corresponding to the actual demand and the optimal solution set; and determine the target order of the filter based on the ideal solution.

[0251] In some embodiments, the determination unit 901 is further used to obtain at least one filter coefficient before truncation, determine the total energy corresponding to the at least one filter coefficient before truncation; determine at least one filter order based on the at least one truncated filter coefficient; the sound field partition control device 900 also includes a processing unit, which is used to obtain a preset energy percentage, and truncate the total energy of the at least one filter coefficient before truncation according to the preset energy percentage to obtain at least one truncated filter coefficient;

[0252] In some embodiments, the vehicle is also provided with a microphone array, and the determination unit 901 is also used to obtain the collection signal of the microphone array and the driving signal of the speaker array; based on the driving signal of the speaker array and the collection signal of the microphone array, the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area are obtained.

[0253] The description of the above device embodiment is similar to the description of the above method embodiment, and has similar beneficial effects as the method embodiment. For technical details not disclosed in the device embodiment of the present application, please refer to the description of the method embodiment of the present application for understanding.

[0254] It should be noted that in the embodiment of the present application, if the above method is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, which is stored in a storage medium, including a number of instructions to enable an electronic device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods of each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory, a magnetic disk or an optical disk. In this way, the embodiment of the present application is not limited to any specific combination of hardware and software.

[0255] The present application also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the above method is implemented when the processor executes the computer program.

[0256] The present application also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. The computer-readable storage medium can be transient or non-transient.

[0257] The present application also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements some or all of the steps in the above method. The computer program product can be implemented specifically by hardware, software, or a combination thereof. The computer program product can be implemented specifically by hardware, software, or a combination thereof. In an optional embodiment, the computer program product is specifically embodied as a computer storage medium, and in another optional embodiment, the computer program product is specifically embodied as a software product, such as a software development kit (SDK), etc.

[0258] It should be noted that Fig.10 A hardware entity diagram of an electronic device provided in an embodiment of the present application, such as Fig.10 As shown, the hardware entity of the electronic device 1000 includes: a processor 1001, a communication interface 1002 and a memory 1003, wherein:

[0259] The processor 1001 generally controls the overall operation of the electronic device 1000 .

[0260] The communication interface 1002 may enable the electronic device 1000 to communicate with other terminals or servers via a network.

[0261] The memory 1003 is configured to store instructions and applications executable by the processor 1001, and can also cache data to be processed or processed by the processor 1001 and each module in the electronic device 1000 (for example, image data, audio data, voice communication data, and video communication data), which can be implemented by flash memory (FLASH) or RAM. Data can be transmitted between the processor 1001, the communication interface 1002, and the memory 1003 through the bus 1004.

[0262] Here, the electronic device may be a vehicle computer in a vehicle.

[0263] It should be noted here that the description of the above storage medium and device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects as the method embodiments. For technical details not disclosed in the storage medium and device embodiments of this application, please refer to the description of the method embodiments of this application for understanding.

[0264] It should be understood that "one embodiment" or "an embodiment" mentioned throughout the specification means that specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present application, the size of the serial number of each step / process mentioned above does not mean the order of execution, and the execution order of each step / process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application. The serial numbers of the embodiments of the present application mentioned above are for description only and do not represent the advantages and disadvantages of the embodiments.

[0265] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0266] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the above-mentioned units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0267] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0268] In addition, all functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may be a separate unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0269] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, read-only memories, magnetic disks or optical disks.

[0270] Alternatively, if the above-mentioned integrated unit of the present application is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can essentially or in other words, the part that contributes to the relevant technology can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods of each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0271] The above is only an implementation method of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for controlling a sound field by partitioning, characterized in that: Applied in a vehicle, the vehicle is provided with a speaker array, the speaker array is used to form a sound field of the vehicle, the method comprises: Determine a bright region and a dark region in the sound field of the vehicle, and obtain a transfer function matrix corresponding to the bright region and a transfer function matrix corresponding to the dark region; Based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area, a partition control model of the sound field is established; wherein the partition control model of the sound field includes a multi-objective optimization problem; Determine an optimal solution set corresponding to the multi-objective optimization problem, and determine a target order of the filter based on the optimal solution set corresponding to the multi-objective optimization problem; Based on the filter at the target order, a driving signal is determined, and the driving signal is output to a speaker array of the vehicle to perform zone control on the sound field of the vehicle.

2. The partition control method according to claim 1, characterized in that: The multi-objective optimization problem includes a first objective function, a second objective function and a third objective function. The establishment of a partition control model of the sound field based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area includes: Determine the first objective function, the second objective function and the third objective function based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area; wherein the first objective function and the second objective function are used to achieve partition control of the sound field, and the third objective function is used to reduce the order of the filter; A partition control model of the sound field is established based on the first objective function, the second objective function and the third objective function.

3. The partition control method according to claim 2, characterized in that: The determining the first objective function, the second objective function, and the third objective function based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area includes: Acquire the number of microphones in the bright area, the expected sound field in the bright area, the number of microphones in the dark area, the number of control frequency points, and the weight vectors of the control frequency points; Determining the first objective function based on a diagonal matrix of a transfer function matrix corresponding to the bright area, a desired sound field of the bright area, the number of control frequency points, and the number of microphones in the bright area; Determining the second objective function based on the diagonal matrix of the transfer function matrix corresponding to the dark area, the number of the control frequency points, and the number of microphones in the dark area; The third objective function is determined based on the number of the control frequency points and the weight vectors of the control frequency points.

4. The partition control method according to claim 1, characterized in that: The method further comprises: Determine a single objective problem corresponding to the multi-objective optimization problem, and convert the single objective problem into an equality constrained optimization problem using an alternating direction multiplier method; Based on Lagrange multipliers and Lagrange variables, decomposing the equality constraint optimization problem into at least one sub-problem, and iteratively solving the at least one sub-problem to obtain a solution set of the at least one sub-problem; Determining the optimal solution set corresponding to the multi-objective optimization problem includes: Based on the solution set of the at least one sub-problem, an optimal solution set corresponding to the multi-objective optimization problem is determined.

5. The partition control method according to claim 1, characterized in that: The step of determining the target order of the filter based on the optimal solution set corresponding to the multi-objective optimization problem comprises: The optimal solution set corresponding to the multi-objective optimization problem is projected into relevant indicators to obtain a quasi-Pareto front corresponding to the optimal solution set; wherein the relevant indicators include at least one of the following: the acoustic energy contrast between the bright area and the dark area, the normalized error of the sound field reconstruction in the bright area, and the mean value of at least one filter order; Based on the Pareto-like front corresponding to the optimal solution set, a target order of the filter is determined.

6. The partition control method according to claim 5, characterized in that: The step of determining the target order of the filter based on the Pareto-like front corresponding to the optimal solution set includes: Acquire actual demand, and determine an ideal solution from the optimal solution set based on a quasi-Pareto front corresponding to the actual demand and the optimal solution set; Based on the ideal solution, a target order of the filter is determined.

7. The partition control method according to claim 5, characterized in that: In the case where the relevant indicator comprises a mean value of the at least one filter order, the method further comprises: Obtain at least one filter coefficient before truncation, and determine the total energy corresponding to the at least one filter coefficient before truncation; Acquire a preset energy percentage, and truncate the total energy of the at least one filter coefficient before truncation according to the preset energy percentage to obtain at least one truncated filter coefficient; At least one filter order is determined based on the at least one truncated filter coefficient.

8. The partition control method according to any one of claims 1 to 7, wherein the vehicle is further provided with a microphone array, and the step of obtaining a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area comprises: Acquiring a collection signal of a microphone array and a driving signal of the speaker array; Based on the driving signal of the speaker array and the collection signal of the microphone array, a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area are obtained.

9. A sound field partition control device, characterized in that: Applied in a vehicle, the vehicle is provided with a speaker array, the speaker array is used to form a sound field of the vehicle, and the device comprises: A determination unit, used to determine a bright area and a dark area in the sound field of the vehicle, and obtain a transfer function matrix corresponding to the bright area and a transfer function matrix corresponding to the dark area; An establishing unit is used to establish a partition control model of the sound field based on the transfer function matrix corresponding to the bright area and the transfer function matrix corresponding to the dark area; wherein the partition control model of the sound field includes a multi-objective optimization problem; The determination unit is further used to determine the optimal solution set corresponding to the multi-objective optimization problem, and determine the target order of the filter based on the optimal solution set corresponding to the multi-objective optimization problem; based on the filter at the target order, determine the driving signal and output the driving signal to the speaker array of the vehicle to perform zone control on the sound field of the vehicle.

10. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program executable on the processor, and the processor implements the method according to any one of claims 1 to 8 when executing the computer program.

Images