A phased ultrasonic focusing optimization method based on an improved balanced optimization algorithm

Abstract

The present invention belongs to the field of ultrasonic technology, and is particularly a phased ultrasonic focusing optimization method based on an improved balanced optimization algorithm. The algorithm includes: hybrid particle initialization; establishing a delay and amplitude parameter library; dividing the target area; establishing an objective function; setting a perturbation coefficient and a generation probability; iteratively evaluating the fitness of parameters and selecting the best one; the specific process is: establishing a multi-physics field simulation model based on environmental information; setting simulation data, performing preliminary amplitude and waveform simulation on ultrasound, establishing an amplitude parameter library and a delay parameter library; running the improved balanced optimization algorithm to intelligently select parameters in the amplitude parameter library and the delay parameter library through hybrid particle initialization, iteratively simulating and outputting the best one; running the simulation according to the output delay and amplitude parameters to obtain the final waveform. The method of the present invention can coordinate the optimization of ultrasonic delay parameters and the allocation of amplitude parameters, and improve the focusing ability of phased ultrasonic in complex environments through visual ultrasonic beam simulation.

Description

Technical Field

[0001] The present invention relates to the field of ultrasonic technology, and in particular to a phase-controlled ultrasonic focusing optimization method based on an improved balanced optimization algorithm. Background Art

[0002] With the development of phased ultrasound technology, methods for focal area adjustment and optimization through swarm intelligence algorithms are becoming increasingly popular. Compared with traditional focusing algorithms, improved equalization optimization algorithms can significantly improve the focusing accuracy of ultrasound in complex environments and improve the sidelobe redundancy caused by delay misalignment and amplitude imbalance, showing good application prospects.

[0003] During phased ultrasonic focusing, the accuracy of ultrasound and the sidelobe suppression ratio mainly depend on the delay combination and amplitude distribution of the ultrasonic array element drive signal. However, these delay and amplitude parameters need to be iteratively optimized and the energy of the focal main lobe and sidelobe needs to be accurately calculated for preferential screening. Otherwise, it may cause focal offset or the sidelobe to submerge the mainlobe. Currently, most phased ultrasonic focusing algorithms lack personalized focusing. When faced with complex environments, it is difficult to accurately adjust the array element drive delay according to the sound speed of different media, and it is impossible to reasonably distribute the array element excitation amplitude to the sound field environment. Therefore, we propose a phased ultrasonic focusing optimization method based on an improved equalization optimization algorithm and COMSOL Multiphysics finite element acoustic field simulation to solve the above problems. Summary of the Invention

[0004] (1) Technical problems solved

[0005] To address the difficulty of properly setting delay and amplitude parameters in complex environments with existing phased-control ultrasonic focusing, which can lead to loss of focus or the main lobe being overwhelmed by side lobes, the present invention, based on environmental data, uses finite element simulation software, simulation coordination tools, and an improved equalization optimization algorithm to construct a method for iteratively simulating and selecting delay and amplitude parameters in complex environments. The proposed multi-physics acoustic field model and improved equalization optimization algorithm dynamically select delay and amplitude parameters based on the medium characteristics of the acoustic field, achieving precise focusing of phased-control ultrasound in complex environments, thus resolving the issues raised in the aforementioned background technology.

[0006] (2) Technical solution

[0007] In order to achieve the above-mentioned purpose, the present invention specifically adopts the following technical solutions:

[0008] A phased ultrasonic focusing optimization method based on an improved balanced optimization algorithm comprises the following steps:

[0009] S1: Use COMSOL Multiphysics finite element simulation software to establish a multi-physics simulation model based on the acoustic field information;

[0010] S2: Set the ultrasonic simulation frequency to 1MHz, perform waveform simulation on the elements in the phased array, establish a delay parameter library, and perform amplitude simulation on the complete phased array to establish an amplitude parameter library;

[0011] S3: Use Comsol LiveLink for Matlab to connect COMSOL Multiphysics finite element simulation software and MATLAB. Use MATLAB to run the improved equilibrium optimization algorithm to intelligently select parameters from the amplitude parameter library and delay parameter library through hybrid particle initialization, and then set the parameters to the COMSOL Multiphysics finite element simulation software for iterative optimization.

[0012] S4: Using COMSOL Multiphysics finite element simulation software, run the simulation according to the delay and amplitude parameters of the optimal parameter output module to obtain the final waveform.

[0013] Furthermore, the S1 uses COMSOL Multiphysics finite element simulation software to establish a trapezoidal two-dimensional phased array with an array element diameter of 2 mm and an array element center spacing of 2.5 mm.

[0014] Furthermore, the S1 uses COMSOL Multiphysics finite element simulation software to establish a multi-physical field containing the skin layer, fat layer, muscle layer, cyst mass, and water base layer, and configures parameters such as density and sound speed for each physical field, configures the ultrasonic frequency to 1 MHz, the simulation time to 1.035e-5s, the time step to 1us, and the grid size to one-sixth of the ultrasonic wavelength.

[0015] Furthermore, S2 uses an integral form, a Gaussian envelope integrated sine wave signal with a position of 2*T0 and a standard deviation of T0 / 2, where T0 is a time constant and the value is the inverse of the ultrasonic frequency. The integrated signal is used to drive the ultrasonic array element to perform sound beam simulation, and the time when the upper and lower edges of the sound beam reach the target position is obtained to form a delay parameter library, and the ultrasonic amplitude that can meet the engineering sound intensity requirements is obtained to form an amplitude parameter library.

[0016] The formula of the driving signal is:

[0017]

[0018] Where, is the sinusoidal driving signal integrated by Gaussian envelope, is the Gaussian envelope, is the amplitude parameter of the driving signal, is the initial time of the driving signal, is the delay parameter of the driving signal.

[0019] Furthermore, in S3, MATLAB is used to run the algorithm to select and set the delay parameters and amplitude parameters, which is specifically a process of hybrid particle initialization, establishment of the objective function, iterative simulation of the delay and amplitude parameters, and output of the optimal delay and amplitude parameters, which is specifically manifested as follows:

[0020] The hybrid particle initialization module configures random initialization to place random particles between the upper and lower bounds of the interval. It also configures Cauchy initialization and sets the Cauchy size parameter to one-sixth of the interval difference. This module uses the long-tail characteristic to improve the diversity of initial particles. Random initialization accounts for 80% and Cauchy initialization accounts for 20%. This module uses a mixture of random initialization and Cauchy initialization to initialize particles in the amplitude parameter library and the delay parameter library.

[0021] The objective function module obtains ultrasonic simulation data through Comsol LiveLink for Matlab and sets the coordinates of the target position to the center of the main lobe. The elliptical area with a long radius of 0.5mm on the X axis and a short radius of 0.3mm on the Y axis is set as the main lobe position of the ultrasound focus, and the remaining positions are set as side lobes. The ultrasonic energy information of the main lobe and side lobe positions is indexed and the sound pressure is calculated. The energy of the main lobe and side lobes is integrated using an adaptive weight distribution method to obtain an objective function that can represent the particle fitness value.

[0022] The objective function and adaptive weight are:

[0023]

[0024]

[0025]

[0026] Where, is the particle fitness value, is the weight of the main lobe energy, is the weight of the sidelobe energy, is the main lobe sound pressure energy, is the sidelobe sound pressure energy, Take a very small positive number as the anti-zero factor.

[0027] The parameter iteration simulation module establishes an optimization candidate pool with a capacity of 5. In each round of iteration, the objective function is used to evaluate the particle fitness value and the better solution is stored in the candidate pool. In the next round of iteration, random perturbations are added based on the parameter particles of the previous round to recalculate the particle fitness value and update the candidate pool based on the best solution. The degree of random perturbation and generation probability are set to decrease with the increase of the number of iterations, so that the particle update amplitude is larger in the early stage and focuses on global search, while the particle update amplitude is smaller in the later stage and focuses on local search.

[0028] The random perturbation is:

[0029]

[0030] Where, The force with which particles explore new areas, is the current iteration number, is the maximum number of iterations;

[0031]

[0032] Where, is the randomness of particles during update, is the current iteration number, is the maximum number of iterations;

[0033] The generation probability is:

[0034]

[0035] Where, is the generation probability, The initial global search weight is 0.8. The initial value of the global search weight at the end is 0.2. is the current iteration number, is the maximum number of iterations.

[0036] In order to avoid the incorrect selection of parameter particles with low fitness values ​​due to inaccurate ultrasonic data transmission or accidental simulation events, the optimal parameter output module outputs the particle with the fourth-ranked fitness value in the candidate pool as the optimal solution after the parameter iteration simulation is completed.

[0037] Furthermore, the improved balanced optimization algorithm of S3 has specific improvements in particle initialization, adaptive weight distribution, random perturbation and generation probability. Mixed initialization is used to enhance the diversity of parameter particles and make the initial distribution area of ​​parameters larger. Adaptive weight distribution is used to dynamically transform the energy proportions of the main lobe and side lobe to prevent the phenomenon that the main lobe optimization force suppresses the side lobe optimization force or the side lobe optimization force suppresses the main lobe optimization force during parameter iterative optimization. Gradual random perturbation and generation probability make the algorithm focus more on global search in the early stage and more on local search in the later stage, so as to achieve comprehensive search as much as possible while avoiding the algorithm from falling into local optimality.

[0038] Furthermore, the S4 uses COMSOL Multiphysics finite element simulation software to set multiple simulation grid sizes, runs simulations based on the delay and amplitude parameters of the output module, compares the beam shape and focal energy under different grid sizes of the delay and amplitude parameters, and verifies the rationality of the parameters.

[0039] (3) Beneficial effects

[0040] Compared with the prior art, the present invention provides a phased ultrasonic focusing optimization method based on an improved balanced optimization algorithm, which has the following beneficial effects:

[0041] The present invention uses COMSOL Multiphysics finite element simulation software to set up corresponding multi-physical fields for simulation based on complex environments. It can set the propagation speed of the sound beam in different areas according to different media in the sound field, and visualize the sound beam waveform when the ultrasonic beam passes through different media, reducing the risk of sound speed mismatch due to uneven media.

[0042] The present invention simulates the array elements in phased ultrasound through finite element simulation software, obtains the time interval for the sound beam to reach the target position to form a delay parameter library, obtains the sound intensity amplitude interval that can meet the engineering needs to form an amplitude parameter library, and uses an improved equalization optimization algorithm to obtain the optimal parameter combination, thereby improving the personalization and accuracy of ultrasonic focusing.

[0043] The present invention adopts hybrid initialization to improve the diversity of particles, so that the initialized particles are distributed over a wider area in the parameter library. The use of adaptive weight distribution reduces the occurrence of target offset during the simulation process and improves the robustness of the algorithm. At the same time, through gradual random perturbation and generation probability, global search is achieved while avoiding the algorithm from falling into local optimality. It has higher universality and is more conducive to the promotion of the method and related equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 Schematic diagram of the flow of a phased ultrasonic focusing optimization method based on an improved balanced optimization algorithm according to an embodiment of the present invention;

[0045] Figure 2 Schematic diagram of a simplified multi-physics sound field model according to an embodiment of the present invention;

[0046] Figure 3 Graph showing the Gaussian envelope and Gaussian-integrated sine wave signal effects according to an embodiment of the present invention;

[0047] Figure 4 This is a driving signal result diagram obtained based on parameter combination according to an embodiment of the present invention;

[0048] Figure 5 Schematic diagram of the distribution of phased ultrasonic array elements of a model according to an embodiment of the present invention;

[0049] Figure 6 This is a diagram showing the optimization effect of phase-controlled ultrasonic focusing and sound pressure distribution according to an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0051] Example

[0052] like Figure 1 As shown, a phased ultrasonic focusing optimization method based on an improved balanced optimization algorithm has the following specific steps:

[0053] S1: Use COMSOL Multiphysics finite element simulation software to establish a multi-physics simulation model based on the acoustic field information;

[0054] In this embodiment, COMSOL Multiphysics finite element simulation software is used to establish a multi-physics field containing the skin layer, fat layer, muscle layer, cyst mass and water base layer, and the corresponding sound speed is configured for each physical field, which is 1610 m / s, 1450 m / s, 1540 m / s, 1700 m / s and 1500 m / s respectively. Figure 2 As shown, numbers 1-11 are perfect matching layers with a water base for absorbing sound waves, number 12 is a water-based layer for simulating ultrasonic coupling agent, number 13 is the skin layer, number 14 is a cyst mass for simulating the diseased area in the body, number 15 is the fat layer, and number 16 is the muscle layer. The configured ultrasonic frequency is 1MHz, the simulation time is 1.035e-5s, the time step is 1us, and the grid size is one-sixth of the wavelength.

[0055] S2: Set the ultrasonic simulation frequency to 1MHz, perform waveform simulation on the elements in the phased array, establish a delay parameter library, and perform amplitude simulation on the complete phased array to establish an amplitude parameter library;

[0056] The specific steps are:

[0057] A1: First, Figure 3 As shown in FIG, a sine wave signal is integrated using an integral form, a Gaussian envelope with a position of 2*T0 and a standard deviation of T0 / 2 to obtain a driving signal that is only sufficient to drive a limited number of ultrasound waves, thereby improving the concentration of the spectrum and the time resolution, while reducing the time domain ringing and the nonlinear distortion of the transducer.

[0058] The formula of the driving signal is:

[0059]

[0060] Where, is the sinusoidal driving signal integrated by Gaussian envelope, is the Gaussian envelope, is the amplitude parameter of the driving signal, is the initial time of the driving signal, is the delay parameter of the driving signal.

[0061] A2: Simulate the sound beam propagation time of each array element in the phased ultrasonic array, record the time it takes for the upper and lower edges of the sound beam to reach the target position, and obtain the time region when the sound beam contacts the target position to form a delay parameter library.

[0062] A3: Set the initial amplitude for the array according to the basic rule of decreasing from the center of the array element to the edge of the array element. Increase the sound intensity contribution of the central array element to the main lobe in the focal area, improve the focal area convergence, reduce the contribution of the edge array elements to the side lobes around the focal area, and improve the sidelobe suppression ratio. The amplitude of the central array element is 24V, the amplitude of the edge array element is 18V, and the amplitude floating range of each array element is 1.5V, forming an amplitude parameter library.

[0063] A4: Integrate the delay parameter library and the amplitude parameter library, use MATLAB to incorporate the parameter library into the improved equalization optimization algorithm, and perform iterative simulation on the delay parameter library and the amplitude parameter library twice.

[0064] In this example, the multiphysics acoustic field model will be run in the COMSOL Multiphysics 6.2 environment using an Intel(R) Core(TM) i7-10700K processor, an NVIDIA GeForce RTX 2080Ti GPU, and an ultrasonic parameter library. Iterative optimization is performed based on the ultrasonic parameter library to determine the ultrasonic parameter combination that achieves the best focusing effect.

[0065] S3: Use Comsol LiveLink for Matlab to connect COMSOL Multiphysics finite element simulation software and MATLAB. Use MATLAB to run the improved equilibrium optimization algorithm to intelligently select parameters in the amplitude parameter library and delay parameter library through hybrid particle initialization and set the parameters to the COMSOL Multiphysics finite element simulation software for iterative optimization.

[0066] A1: Based on the original equalization optimization algorithm, this algorithm is improved. Because the original algorithm's particle initialization method during encoding is too simple, the random perturbation capability and generation probability are too rigid, and the objective function is not suitable for calculating sound field energy information, this embodiment makes corresponding improvements based on the original algorithm to make the algorithm more adaptable when facing phased ultrasonic focusing optimization. The specific improvements are as follows:

[0067] (1) Optimize particle initialization: Add Cauchy initialization on the basis of random initialization of the original algorithm, use random initialization to make the particle distribution more uniform to ensure the global exploration ability of the algorithm, use the long-tail characteristics of Cauchy initialization to generate some values ​​in the center of the principle parameter library, improve particle diversity, and reduce the possibility of the algorithm falling into local optimality due to excessive particle averaging. Random initialization accounts for 80% and Cauchy initialization accounts for 20%. Mix random initialization and Cauchy initialization, and initialize 100 particles in each of the amplitude parameter library and delay parameter library.

[0068] (2) Optimize random perturbations and generation probabilities: Modify the fixed random perturbation intensity to decrease with the increase of the number of iterations, so that the particle update amplitude is larger in the early stage and focuses on global search, and the particle update amplitude is smaller in the later stage and focuses on local search. Modify the fixed generation probability to a mode that gradually decays from 0.8 to 0.2 with the number of iterations, so that the degree to which the generation probability participates in the optimization process gradually decreases, reducing its exploration ability and increasing the convergence efficiency of the algorithm.

[0069] The random perturbation is:

[0070]

[0071] Where, The force with which particles explore new areas, is the current iteration number, is the maximum number of iterations;

[0072]

[0073] Where, is the randomness of particles during update, is the current iteration number, is the maximum number of iterations;

[0074] The generation probability is:

[0075]

[0076] Where, is the generation probability, is the initial global search weight, the initial value is 0.8, is the global search weight at the end, with an initial value of 0.2. is the current iteration number, is the maximum number of iterations.

[0077] (3) Modify the objective function: Replace the original objective function so that it can obtain ultrasonic simulation data and perform calculations through Comsol LiveLink for Matlab. Set the target position coordinates to the center of the main lobe, set the elliptical area with a long radius of 0.5 mm on the X axis and a short radius of 0.3 mm on the Y axis as the main lobe position for ultrasonic focusing, and the remaining positions as side lobes. Index the ultrasonic energy information of the main lobe and side lobe positions, calculate the sound pressure, and use adaptive weight distribution to fuse the energy of the main lobe and side lobe to obtain an objective function that can characterize the particle fitness value.

[0078] The objective function and adaptive weight are:

[0079]

[0080]

[0081]

[0082] Where, is the particle fitness value, is the weight of the main lobe energy, is the weight of the sidelobe energy, is the main lobe sound pressure energy, is the sidelobe sound pressure energy, Take a very small positive number as the anti-zero factor.

[0083] A2: Use COMSOL LiveLink for Matlab to invoke MATLAB R2023b. Set the user information, server, and port number in the finite element simulation software. Based on the above information, link the finite element simulation software and algorithm environment to achieve the link between COMSOL Multiphysics 6.2 and MATLAB R2023b.

[0084] A3: Encode the delay parameter library and amplitude parameter library separately, use the improved equalization optimization algorithm twice to intelligently select particles initialized in the delay parameter library and amplitude parameter library for simulation calculation, and transmit the sound field information to MATLAB through ComsolLiveLink for Matlab for fitness value calculation. There are two calculation methods, as follows:

[0085] (1) After each simulation is completed, MATLAB is used to directly read the sound field information of the finite element simulation software through Comsol LiveLink for Matlab. The main lobe and side lobe positions divided by the objective function are indexed, and the grid blocks loaded with sound pressure information are divided according to the position. The sound pressure information loaded by the grid is squared to obtain the absolute sound pressure, and the energy attribution is divided according to the grid position.

[0086] (2) After each simulation is completed, the data containing coordinate information and sound pressure information is exported to MATLAB in the form of a document using finite element simulation software. The coordinates of the document are divided using the position information of the objective function, and the main lobe calculation formula, side lobe calculation formula, and grating lobe calculation formula are used to obtain the distribution of sound pressure energy.

[0087] The main lobe calculation formula is:

[0088]

[0089] Where, is the main lobe energy, is the ultrasonic information in the time domain, is the total simulation time;

[0090] The sidelobe calculation formula is:

[0091]

[0092] Where, is the ultrasound sidelobe energy, is the ultrasonic information in the time domain, For time delay;

[0093] The grating lobe calculation formula is:

[0094]

[0095] Where, is the ultrasonic grating lobe energy, is the ultrasonic signal under the spectrum, is the ultrasonic frequency fluctuation range during simulation;

[0096] The sidelobe sound pressure energy of the objective function is composed of the sidelobe and the grating lobe, specifically:

[0097]

[0098] Where, is the ultrasonic sidelobe sound pressure energy, is the ultrasound sidelobe energy, is the grating lobe energy of ultrasound.

[0099] A4: Use the objective function to evaluate the fitness value of the parameter particles, and iteratively calculate the newly generated parameter particles with the help of random perturbations and generation probabilities. Use the algorithm logic to save the information of particles with high fitness values ​​in the optimization candidate pool as output options.

[0100] A5: Use the same method to iteratively optimize the remaining parameter pool and export the calculation results from the optimization candidate pool, such as Figure 4 As shown, the optimal delay and amplitude parameter combination is formed.

[0101] S4: Using COMSOL Multiphysics finite element simulation software, run the simulation according to the delay and amplitude parameters of the optimal parameter output module to obtain the final waveform.

[0102] In this embodiment, in order to save simulation time and hardware resources, this model uses 7 aluminum nitride transducers with a center frequency of 1 MHz and a diameter of 2 mm, and establishes a two-dimensional simulation model according to the rule of 0.5 mm array element spacing. Figure 5 The simulation of the annular axisymmetric phased ultrasonic array with 37 array elements and protruding edges is shown.

[0103] like Figure 6 As shown in the figure, the phased ultrasound focusing effect diagram and sound field sound pressure distribution diagram are obtained through iterative simulation based on the improved equalization optimization algorithm of the present invention. Under the control of the improved equalization optimization algorithm, the focal area obtained by the phased ultrasound shows good focusing ability compared with the traditional spatial geometry algorithm, and can more accurately lock the ultrasound focusing position at the target position coordinate (0,12), maximize the sound pressure at the target position, and its sound pressure peak is improved by about 1Mpa compared with the traditional focusing algorithm.

[0104] In response to the problem that it is difficult to accurately set the delay parameters and amplitude parameters of existing ultrasound in complex environments, resulting in focal offset or side lobes submerging the main lobe, the present invention uses COMSOL Multiphysics finite element simulation software to establish a multi-physics field acoustic model based on environmental information, and cooperates with an improved equalization optimization algorithm to intelligently select delay and amplitude parameters and perform acoustic simulation, and iteratively optimize the delay and amplitude parameter combination. The proposed multi-physics field acoustic model and the improved equalization optimization algorithm can, through iterative optimization, derive a parameter combination that enables phased ultrasound to be accurately focused on the target position and meets engineering requirements, thereby alleviating the defocusing problem caused by the difficulty in achieving sound speed matching during propagation due to different media. Therefore, the method proposed in this article can be used as a possible way to achieve efficient focusing of phased ultrasound through finite element simulation combined with an improved equalization optimization algorithm.

[0105] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A phased ultrasonic focusing optimization method based on an improved balanced optimization algorithm, characterized by: Specifically include: S1: Use COMSOL Multiphysics finite element simulation software to establish a multi-physics simulation model based on the environmental information of the acoustic field; S2: Set the ultrasonic simulation frequency to 1MHz, perform waveform simulation on the elements in the phased array, establish a delay parameter library, and perform amplitude simulation on the complete phased array to establish an amplitude parameter library; S3: Use Comsol LiveLink for Matlab to connect COMSOL Multiphysics finite element simulation software and MATLAB, and use MATLAB to run the improved balanced optimization algorithm to initialize, iteratively optimize, and output the amplitude and delay parameters; In S3, MATLAB is used to run an improved balanced optimization algorithm to select and set delay parameters and amplitude parameters, specifically a process of hybrid particle initialization, establishment of an objective function, iterative simulation of delay and amplitude parameters, and output of optimal delay and amplitude parameters; The improvements of the improved equilibrium optimization algorithm include: Hybrid particle initialization module uses a mixture of random initialization and Cauchy initialization, with random initialization accounting for 80% and Cauchy initialization accounting for 20%. It uses a mixture of random initialization and Cauchy initialization to initialize parameter particles in the delay parameter library and amplitude parameter library. The objective function module acquires ultrasound simulation data through Comsol LiveLink for Matlab and sets the coordinates of the target position to the center of the main lobe. The elliptical area with a long radius of 0.5 mm on the X axis and a short radius of 0.3 mm on the Y axis is set as the main lobe position of the ultrasound focus, and the remaining positions are set as side lobes. The energy of the main lobe and side lobe areas is indexed and calculated. The parameter iteration simulation module sets an optimization candidate pool with a capacity of 5, uses the objective function to calculate the fitness of the initialized particles and stores the calculation results in the optimization candidate pool. In the next round of iterative calculation, the improved perturbation is added to the particles in the previous round, and the particle fitness values ​​are calculated and compared again, and the optimization candidate pool is updated based on the best results. The optimal parameter output module outputs the parameters of the particle with the fourth-highest fitness value in the optimization candidate pool as the optimal solution after the parameter iteration simulation is completed; The objective function module of S3 is specifically to obtain ultrasonic simulation data through Comsol LiveLink for Matlab, set the target position coordinate to the center of the main lobe and divide the main lobe and side lobe areas, obtain the energy of the main lobe position and the side lobe position to calculate the sound pressure, and use adaptive weight distribution to fuse the energy of the main lobe and side lobe to obtain the objective function that can characterize the particle fitness value. The objective function and adaptive weight are: o=-(w1*mainLobe)+(w2*sideLobe) Where o is the particle fitness value, w1 is the weight of the main lobe energy, w2 is the weight of the side lobe energy, mainLoba is the main lobe sound pressure energy, sideLobe is the side lobe sound pressure energy, and ε is the anti-zero factor; S4: Using COMSOL Multiphysics finite element simulation software, run the simulation according to the delay and amplitude parameters of the optimal parameter output module to obtain the final waveform.

2. The phased ultrasonic focusing optimization method based on the improved balanced optimization algorithm according to claim 1, characterized in that: Specifically, S2 is to perform waveform simulation on the elements in the phased array to obtain the time when the upper edge and the lower edge of the waveform reach the target area, calculate the difference to obtain the time interval and establish a delay parameter library; Perform amplitude simulation on the complete phased array to obtain the amplitude parameter range that enables phased ultrasound to meet engineering requirements and establish an amplitude parameter library.

3. The phased ultrasonic focusing optimization method based on the improved balanced optimization algorithm according to claim 1, characterized in that: The hybrid particle initialization module of S3 specifically configures random initialization to place random particles between the upper and lower bounds of the interval, configures Cauchy initialization, sets the Cauchy size parameter to one-sixth of the interval difference, and utilizes its long-tail characteristic to improve the diversity of initial particle parameters. It mixes random initialization and Cauchy initialization to perform particle initialization in the amplitude parameter library and the delay parameter library.

4. The phased ultrasonic focusing optimization method based on the improved balanced optimization algorithm according to claim 1, characterized in that: The parameter iteration simulation module of S3 is specifically to establish an optimization candidate pool with a capacity of 5. In each round of iteration, the objective function is used to evaluate the particle fitness value and the better solution is stored in the candidate pool. In the next round of iteration, random perturbation is added based on the parameter particles of the previous round to recalculate the particle fitness value and update the candidate pool based on the best solution. The degree of random perturbation and the generation probability are set to decrease with the increase of the number of iterations, so that the particle update amplitude in the early stage is large and focuses on global search, and the particle update amplitude in the later stage is small and focuses on local search. The random perturbation is: a1=2-Iter*(2 / Max_iter) Where a1 is the force of the particle exploring the new area, Iter is the current number of iterations, and Max_iter is the maximum number of iterations; a2=1-Iter*(1 / Max_iter) Where a2 is the randomness of the particle during update, Iter is the current number of iterations, and Max_iter is the maximum number of iterations; The generation probability is: Where GP is the generation probability, GP0 is the initial global search weight with an initial value of 0.8, gP1 is the final global search weight with an initial value of 0.2, Iter is the current number of iterations, and Max_iter is the maximum number of iterations.

Images