Optimization method and system for simulating artificial cochlea nervous system

By simulating the cochlear implant nervous system model, combining piezoelectric ceramic units and nonlinear neuron circuits, the signal distortion of the cochlear implant system in complex sound signal processing is solved by using dynamic behavior analysis and time domain minimum residual value method. Responding to delay problems, improving auditory effects and wearing experience.

CN118903678APending Publication Date: 2024-11-08SUN YAT SEN UNIVERSITY SHENZHEN +1
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411028354.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-30
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The existing cochlear implant system has problems such as signal distortion, response delay and insufficient processing accuracy when processing complex sound signals, resulting in insufficient auditory perception and making it difficult for the wearer to distinguish different sounds.

Method used

The cochlear implant neural system model is used to combine piezoelectric ceramic units and nonlinear neuron circuits, and dynamic behavior analysis and fourth-order Longkuta method to solve dynamic equations, and optimize the processing with the time domain minimum residual value method to adjust the signal The amplitude and frequency of the signal transmission can be reduced, and the response residuals can be improved.

Benefits of technology

By optimizing the amplitude and frequency of the signal in real time, reducing distortion during signal conversion, improving the auditory effect of cochlear implant wearer, enhancing the ability to distinguish sound details, and improving the wearing experience.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118903678A_ABST
    Figure CN118903678A_ABST
Patent Text Reader

Abstract

The invention discloses an optimization method and system for simulating an artificial cochlea nervous system, and the method comprises the steps: constructing an artificial cochlea nervous system simulation model which comprises a piezoelectric ceramic unit and a nonlinear neuron circuit; performing dynamic behavior analysis on the nonlinear neuron circuit, and constructing a kinetic equation for simulating an artificial cochlea nervous system model; solving and calculating the kinetic equation of the simulated artificial cochlea nervous system model through a fourth-order Runge-Kutta method to obtain a displacement response of the simulated artificial cochlea nervous system model; and according to the displacement response of the simulated artificial cochlea nervous system model, performing optimization processing on the simulated artificial cochlea nervous system model through a time domain minimum residual value method. According to the embodiment of the invention, the amplitude and frequency of the input signal can be optimized in real time, and the authenticity and accuracy of signal transmission are improved. The optimization method and system for simulating the artificial cochlea nervous system can be widely applied to the technical field of nervous system simulation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of neural system simulation, and in particular to an optimization method and system for simulating an artificial cochlear neural system. Background Art

[0002] A cochlear implant is an electronic device used to help the hearing-impaired restore or improve their hearing. It directly stimulates the auditory nerve by converting sound signals into electrical signals. However, existing cochlear implant systems have certain limitations when processing complex sound signals, including signal distortion, response delays, and insufficient processing accuracy. First, signal distortion is a major problem. In the process of converting external sound waves into electrical signals, traditional cochlear implant systems are often interfered by environmental noise and complex sound signals, causing signal distortion during transmission and unclear auditory perception. Secondly, insufficient processing accuracy is also a major problem that plagues cochlear implant systems. Due to the limitations of signal processing algorithms and hardware design, the existing systems do not have sufficient processing accuracy for complex sound signals, resulting in loss of sound details and difficulty for wearers to clearly distinguish different sounds, such as speech and background noise, which in turn leads to a decline in the wearer's experience. Summary of the invention

[0003] In order to solve the above technical problems, the purpose of the present invention is to provide an optimization method and system for simulating the neural system of a cochlear implant, which can optimize the amplitude and frequency of the input signal in real time, improve the authenticity and accuracy of signal transmission, and enhance the auditory effect of cochlear implant wearers.

[0004] The first technical solution adopted by the present invention is: an optimization method for simulating a cochlear implant neural system, comprising the following steps:

[0005] Constructing a simulated cochlear implant neural system model, wherein the simulated cochlear implant neural system model includes a piezoelectric ceramic unit and a nonlinear neuron circuit;

[0006] Conduct dynamic behavior analysis on nonlinear neuronal circuits and construct dynamic equations to simulate the neural system model of cochlear implants;

[0007] The dynamic equations of the cochlear implant neural system model are solved and calculated by the fourth-order Runge-Kutta method to obtain the displacement response of the cochlear implant neural system model.

[0008] According to the displacement response of the simulated cochlear implant nerve system model, the simulated cochlear implant nerve system model is optimized by using the time domain minimum residual method to obtain the optimized simulated cochlear implant nerve system model.

[0009] Further, the simulated cochlear implant neural system model includes a piezoelectric ceramic unit, a first resistor, a second resistor, a third resistor, a first capacitor, a first inductor and a first power supply, wherein the first resistor, the second resistor, the third resistor, the first capacitor, the first inductor and the first power supply constitute a nonlinear neuron circuit, the first end of the first inductor, the first end of the first resistor, the first end of the first capacitor and the first end of the third resistor are connected, the second end of the first inductor is connected to the first end of the second resistor, the second end of the second resistor is connected to the negative end of the first power supply, the positive end of the first power supply, the second end of the first capacitor, and the second end of the third resistor are connected to the negative end of the piezoelectric ceramic unit and grounded, and the positive end of the piezoelectric ceramic unit is connected to the second end of the first resistor.

[0010] Furthermore, the step of performing dynamic behavior analysis on the nonlinear neuron circuit and constructing a dynamic equation for simulating the cochlear implant neural system model specifically includes:

[0011] The nonlinear neuron circuit simulating the cochlear implant neural system model is analyzed for dynamic behavior by Kirchhoff's law, and the circuit equation of the cochlear implant neural system model is obtained;

[0012] The circuit equations simulating the neural system model of the cochlear implant are scaled, and the original circuit equations containing dimensions are converted into equivalent dimensionless mathematical equations, so as to obtain the dynamic equations simulating the neural system model of the cochlear implant.

[0013] Furthermore, the expression of the dynamic equation of the simulated cochlear implant neural system model is specifically as follows:

[0014]

[0015] In the above formula, X represents the state variable, represents the derivative of the state variable with respect to dimensionless time, C represents the damping coefficient matrix, K represents the stiffness coefficient matrix, represents the nonlinear term, F(t) represents the external excitation, and t represents the dimensionless time.

[0016] Further, the step of optimizing the simulated cochlear implant nervous system model by using a time domain minimum residual method according to the displacement response of the simulated cochlear implant nervous system model to obtain an optimized simulated cochlear implant nervous system model specifically includes:

[0017] The displacement response of the cochlear implant neural system model is fitted by fast Fourier transform to obtain the truncated series;

[0018] Based on the enhanced response sensitivity algorithm, an optimization objective function for simulating the neural system model of the cochlear implant is constructed;

[0019] The truncated series is used as the initial value of the time domain minimum residual method, and the optimization objective function of the simulated cochlear implant neural system model is optimized by the time domain minimum residual method to obtain the optimal truncated series;

[0020] The optimal truncation series is embedded into the simulated cochlear implant neural system model to obtain the optimized simulated cochlear implant neural system model.

[0021] Furthermore, the expression of the optimization objective function of the simulated cochlear implant neural system model is specifically as follows:

[0022]

[0023] In the above formula, represents the nonlinear objective function, a represents the unknown coefficient, A represents the feasible domain of the unknown coefficient, t represents the discrete time point, R(a,t) represents the residual vector of the system, T represents the maximum value in the time domain, a * Represents the optimized coefficients.

[0024] Furthermore, the step of using the truncated series as the initial value of the time domain minimum residual method, optimizing the optimization objective function of the simulated cochlear implant neural system model by the time domain minimum residual method, and obtaining the optimal truncated series specifically includes:

[0025] The truncated series is used as the initial value of the time domain minimum residual method, and the optimization objective function of the cochlear implant neural system model is linearly transformed to obtain an approximate linear objective function.

[0026] By introducing Tikhonov regularization, the approximate linear objective function is regularized and solved, and the regularized linear objective function is obtained;

[0027] The regularization parameter of the regularized linear objective function is determined by the L-curve method, and the confidence region restriction is introduced to screen the regularized linear objective function and determine the optimal truncation series.

[0028] Furthermore, the regularized linear objective function is expressed as follows:

[0029]

[0030] In the above formula, λ represents the regularization parameter, I represents the unit matrix, and δa λ represents the iterative update amount, represents the residual vector, represents the first-order response sensitivity matrix.

[0031] Furthermore, the expression of the confidence region restriction is specifically as follows:

[0032]

[0033] In the above formula, represents the confidence region limit, represents the original nonlinear objective function, represents the objective function after linearization, represents known parameters, a represents unknown coefficients, represents the residual vector.

[0034] The second technical solution adopted by the present invention is: an optimization system for simulating a cochlear implant neural system, comprising:

[0035] A first building module is used to build a simulated cochlear implant neural system model, wherein the simulated cochlear implant neural system model includes a piezoelectric ceramic unit and a nonlinear neuron circuit;

[0036] The second building block is used to perform dynamic behavior analysis on nonlinear neuron circuits and construct dynamic equations to simulate the neural system model of the cochlear implant;

[0037] A solution module is used to solve the dynamic equations of the simulated cochlear implant neural system model by using the fourth-order Runge-Kutta method to obtain the displacement response of the simulated cochlear implant neural system model;

[0038] The optimization module is used to optimize the simulated cochlear implant nerve system model according to the displacement response of the simulated cochlear implant nerve system model through the time domain minimum residual method to obtain the optimized simulated cochlear implant nerve system model.

[0039] The beneficial effects of the method and system of the present invention are as follows: the present invention combines a piezoelectric ceramic unit with a nonlinear neuron circuit to construct a simulated cochlear implant neural system model, adopts high-performance piezoelectric ceramic materials, can more efficiently convert external sound wave signals into electrical signals, reduces the distortion problem in the signal conversion process, and then performs dynamic behavior analysis on the nonlinear neuron circuit, constructs the dynamic equation of the simulated cochlear implant neural system model, solves and calculates the dynamic equation of the simulated cochlear implant neural system model through the fourth-order Runge-Kutta method and optimizes the simulated cochlear implant neural system model in combination with the time domain minimum residual method, adjusts the amplitude and frequency of the system input signal, minimizes the system response residual, optimizes the amplitude and frequency of the input signal in real time, ensures the response accuracy of the system in different sound environments, improves the authenticity and accuracy of signal transmission, and improves the hearing effect of cochlear implant wearers. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a flowchart of the steps of an optimization method for simulating a cochlear implant neural system of the present invention;

[0041] Figure 2 It is a structural block diagram of an optimization system for simulating a cochlear implant neural system according to the present invention;

[0042] Figure 3 It is a schematic diagram of the structure of a neural system model for simulating cochlear implants according to a specific embodiment of the present invention. DETAILED DESCRIPTION

[0043] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The step numbers in the following embodiments are only provided for the convenience of explanation and description, and the order between the steps is not limited in any way. The execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0044] Reference Figure 1 The present invention provides an optimization method for simulating a cochlear implant neural system, the method comprising the following steps:

[0045] S100, constructing a simulated cochlear implant neural system model, wherein the simulated cochlear implant neural system model includes a piezoelectric ceramic unit and a nonlinear neuron circuit;

[0046] Specifically, Figure 3 As shown, the simulated cochlear implant neural system model includes a piezoelectric ceramic unit, a first resistor, a second resistor, a third resistor, a first capacitor, a first inductor and a first power supply, wherein the first resistor, the second resistor, the third resistor, the first capacitor, the first inductor and the first power supply constitute a nonlinear neuron circuit, the first end of the first inductor, the first end of the first resistor, the first end of the first capacitor and the first end of the third resistor are connected, the second end of the first inductor is connected to the first end of the second resistor, the second end of the second resistor is connected to the negative terminal of the first power supply, the positive terminal of the first power supply, the second end of the first capacitor, and the second end of the third resistor are connected to the negative terminal of the piezoelectric ceramic unit and grounded, and the positive terminal of the piezoelectric ceramic unit is connected to the second end of the first resistor.

[0047] In this embodiment, a nonlinear neuron circuit including capacitors, inductors, voltage sources and resistors is designed, and piezoelectric ceramics are combined with the nonlinear neuron circuit to construct a complete model simulating the neural system of a cochlear implant.

[0048] S200, dynamic behavior analysis of nonlinear neuron circuits, and construction of dynamic equations for simulating the neural system model of cochlear implants;

[0049] Specifically, the dynamic behavior of the nonlinear neuron circuit simulating the cochlear implant neural system model is analyzed by Kirchhoff's law to obtain the circuit equation of the cochlear implant neural system model; the circuit equation of the cochlear implant neural system model is scaled to convert the original dimensioned circuit equation into an equivalent dimensionless mathematical equation to obtain the dynamic equation of the cochlear implant neural system model.

[0050] In this embodiment, Kirchhoff's law (KCL and KVL) is used to analyze the dynamic behavior of the circuit to obtain the circuit equation of the system. Through scale transformation, the original circuit equation containing dimensions is converted into an equivalent dimensionless mathematical equation, wherein the expression of the dynamic equation simulating the cochlear implant neural system model is specifically as follows:

[0051]

[0052] In the above formula, X represents the state variable, represents the derivative of the state variable with respect to dimensionless time, C represents the damping coefficient matrix, K represents the stiffness coefficient matrix, represents the nonlinear term, F(t) represents the external excitation, and t represents the dimensionless time.

[0053] It should be noted that the expressions of state variables and their derivatives are as follows:

[0054]

[0055] The expressions of the damping coefficient matrix and the stiffness coefficient matrix are as follows:

[0056]

[0057] The expressions of nonlinear terms and external excitation are as follows:

[0058]

[0059] In the above formula, α, μ and δ represent normalized parameters, A represents the amplitude of the external excitation, and Ω represents the frequency of the external excitation.

[0060] S300, solving and calculating the dynamic equation of the simulated cochlear implant neural system model by a fourth-order Runge-Kutta method to obtain a displacement response of the simulated cochlear implant neural system model;

[0061] Specifically, the initial conditions and solution interval are determined according to the dynamic equations of the cochlear implant neural system model. The solution interval is divided into several small steps, and the approximate solution at each step is determined. The iterative formula of the Runge-Kutta method is used to calculate the approximate solution at each step in turn. The approximate solution at each step is weighted averaged to obtain the approximate solution for the entire solution interval.

[0062] S400: Optimizing the simulated cochlear implant nerve system model by using a time domain minimum residual method according to the displacement response of the simulated cochlear implant nerve system model to obtain an optimized simulated cochlear implant nerve system model.

[0063] Specifically, the displacement response of the simulated cochlear implant neural system model is fitted by fast Fourier transform to obtain a truncated series; based on the enhanced response sensitivity algorithm, the optimization objective function of the simulated cochlear implant neural system model is constructed; the truncated series is used as the initial value of the time domain minimum residual method, and the optimization objective function of the simulated cochlear implant neural system model is optimized by the time domain minimum residual method to obtain the optimal truncated series; the optimal truncated series is embedded into the simulated cochlear implant neural system model to obtain the optimized simulated cochlear implant neural system model.

[0064] Among them, it should be noted that the optimization process of the time-domain minimum residual method is: using the truncated series as the initial value of the time-domain minimum residual method, linearly transforming the optimization objective function of the simulated cochlear implant neural system model to obtain an approximate linear objective function; regularizing the approximate linear objective function by introducing Tikhonov regularization to obtain a regularized linear objective function; determining the regularization parameter of the regularized linear objective function by the L-curve method, introducing a trust region restriction, screening the regularized linear objective function, and determining the optimal truncated series.

[0065] In this embodiment, a fast Fourier transform (FFT) is introduced to fit the displacement response to obtain a set of truncation coefficients with a truncation order of N.

[0066] As a further preferred embodiment of the method, the truncated series obtained above is used as the initial value of the time domain minimum residual method.

[0067] As a further preferred embodiment of the method, the expression of the objective function is as follows:

[0068]

[0069] In the above formula, represents the nonlinear objective function, a represents the unknown coefficient, A represents the feasible domain of the unknown coefficient, t represents the discrete time point, R(a,t) represents the residual vector of the system, T represents the maximum value in the time domain, a* Represents the optimized coefficients.

[0070] According to the known parameters Find a reasonable iterative update amount δa so that As small as possible. Give an initial guess value a (0) After that, the iterative update amount δa is calculated (j) ,j=1,2,...until the target function Until convergence.

[0071] Based on the enhanced response sensitivity algorithm, the objective function is optimized and the optimal truncation series is further determined. The high-precision response of the system is determined according to the optimal truncation series. The specific steps are as follows:

[0072] The objective function At discrete time nodes {t k ∈[0,T]|k=1,2…l} to linearize the original nonlinear objective function is transformed into an approximate linear objective function Its expression is:

[0073]

[0074] In the above formula The first-order response sensitivity matrix is ​​as follows, and its expression is:

[0075]

[0076] By introducing Tikhonov regularization to solve the approximate linear objective function This may be an ill-posed problem, so the objective function equation becomes as follows:

[0077]

[0078] In the above formula, I is the unit matrix, λ represents the regularization parameter and λ≥0.

[0079] The regularization parameter λ is determined by the L-curve method, which corresponds to the maximum curvature of the L-curve. Since the L-curve method is only for the linearized objective function Instead of the original nonlinear objective function Therefore, it is necessary to introduce a confidence region restriction, namely the "consistency index", which is expressed as:

[0080]

[0081] In the above formula, represents the confidence region limit, represents the original nonlinear objective function, represents the objective function after linearization, represents known parameters, a represents unknown coefficients, represents the residual vector.

[0082] When the consistency index meets , it means that the objective function after linearization is close enough to the original nonlinear objective function.

[0083] When the iteration conditions are met, the calculation is completed and the optimal truncation coefficient is obtained.

[0084] It should be noted that the truncated series is a set of coefficients that need to be determined when solving the system. The larger the truncated series, the higher the accuracy of the system solution. Through the optimal truncated series, the system's response to the input sound signal can be accurately controlled to reduce distortion and noise interference. The displacement response of the system is fitted using FFT to obtain the truncated series. The truncated series reflects the main frequency components of the system response. Then the time domain minimum residual method (TMRM) is used to further optimize the initial truncated series, minimize the residual between the system response and the expected response, and finally determine the optimal truncated series.

[0085] In summary, the embodiment of the present invention receives external sound wave signals and converts them into electrical signals. At the same time, the signal conditioner performs preliminary processing and amplification on the received electrical signals to ensure that the signal strength is suitable for subsequent nonlinear neuron circuit processing, and further simulates the nonlinear dynamic response of biological neurons. The operational amplifier is used to amplify and filter the signal. The resistor, inductor and capacitor elements jointly construct a circuit model that simulates neurons, reflecting the spatiotemporal response characteristics of neurons. By adjusting the adjustable resistor and capacitor, the parameters of the circuit can be optimized to make it closer to the actual behavior of biological neurons and convert it into a signal suitable for auditory nerve stimulation. The unit includes functional units such as signal filtering, amplification and modulation, and finally the embedded processor runs the TMRM optimization algorithm to minimize the residual between the system response and the expected response through iterative calculation. The optimization algorithm software analyzes and adjusts the system parameters in real time to improve the response accuracy and signal processing capability of the system.

[0086] Reference Figure 2 , an optimization system for simulating a cochlear implant neural system, comprising:

[0087] A first construction module 201 is used to construct a simulated cochlear implant neural system model, wherein the simulated cochlear implant neural system model includes a piezoelectric ceramic unit and a nonlinear neuron circuit;

[0088] The second construction module 202 is used to perform dynamic behavior analysis on the nonlinear neuron circuit and construct a dynamic equation for simulating the neural system model of the cochlear implant;

[0089] A solution module 203 is used to solve the dynamic equation of the simulated cochlear implant neural system model by using the fourth-order Runge-Kutta method to obtain the displacement response of the simulated cochlear implant neural system model;

[0090] The optimization module 204 is used to optimize the simulated cochlear implant nerve system model by using the time domain minimum residual method according to the displacement response of the simulated cochlear implant nerve system model to obtain an optimized simulated cochlear implant nerve system model.

[0091] The contents of the above method embodiments are all applicable to the present system embodiments. The functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0092] The above is a specific description of the preferred implementation of the present invention, but the invention is not limited to the embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. An optimization method for simulating the neural system of a cochlear implant, It is characterized in that The following steps are involved: Constructing a simulated cochlear implant neural system model, wherein the simulated cochlear implant neural system model includes a piezoelectric ceramic unit and a nonlinear neuron circuit; Conduct dynamic behavior analysis on nonlinear neuronal circuits and construct dynamic equations to simulate the neural system model of cochlear implants; The dynamic equations of the cochlear implant neural system model are solved and calculated by the fourth-order Runge-Kutta method to obtain the displacement response of the cochlear implant neural system model. According to the displacement response of the simulated cochlear implant nerve system model, the simulated cochlear implant nerve system model is optimized by using the time domain minimum residual method to obtain the optimized simulated cochlear implant nerve system model.

2. The optimization method for simulating a cochlear implant neural system according to claim 1, It is characterized in that The simulated cochlear implant neural system model includes a piezoelectric ceramic unit, a first resistor, a second resistor, a third resistor, a first capacitor, a first inductor and a first power supply, wherein the first resistor, the second resistor, the third resistor, the first capacitor, the first inductor and the first power supply constitute a nonlinear neuron circuit, the first end of the first inductor, the first end of the first resistor, the first end of the first capacitor and the first end of the third resistor are connected, the second end of the first inductor is connected to the first end of the second resistor, the second end of the second resistor is connected to the negative end of the first power supply, the positive end of the first power supply, the second end of the first capacitor and the second end of the third resistor are connected to the negative end of the piezoelectric ceramic unit and grounded, and the positive end of the piezoelectric ceramic unit is connected to the second end of the first resistor.

3. The optimization method for simulating a cochlear implant neural system according to claim 1, It is characterized in that The step of performing dynamic behavior analysis on the nonlinear neuron circuit and constructing a dynamic equation for simulating the cochlear implant neural system model specifically includes: The nonlinear neuron circuit simulating the cochlear implant neural system model is analyzed for dynamic behavior by Kirchhoff's law, and the circuit equation of the cochlear implant neural system model is obtained; The circuit equations simulating the neural system model of the cochlear implant are scaled, and the original circuit equations containing dimensions are converted into equivalent dimensionless mathematical equations, so as to obtain the dynamic equations simulating the neural system model of the cochlear implant.

4. The optimization method for simulating a cochlear implant neural system according to claim 3, It is characterized in that The expression of the dynamic equation of the simulated cochlear implant neural system model is specifically as follows: In the above formula, X represents the state variable, represents the derivative of the state variable with respect to dimensionless time, C represents the damping coefficient matrix, K represents the stiffness coefficient matrix, represents the nonlinear term, F(t) represents the external excitation, and t represents the dimensionless time.

5. The optimization method for simulating a cochlear implant neural system according to claim 1, It is characterized in that The step of optimizing the simulated cochlear implant nervous system model by using a time domain minimum residual method according to the displacement response of the simulated cochlear implant nervous system model to obtain an optimized simulated cochlear implant nervous system model specifically includes: The displacement response of the cochlear implant neural system model is fitted by fast Fourier transform to obtain the truncated series; Based on the enhanced response sensitivity algorithm, an optimization objective function for simulating the neural system model of the cochlear implant is constructed; The truncated series is used as the initial value of the time domain minimum residual method, and the optimization objective function of the simulated cochlear implant neural system model is optimized by the time domain minimum residual method to obtain the optimal truncated series; The optimal truncation series is embedded into the simulated cochlear implant neural system model to obtain the optimized simulated cochlear implant neural system model.

6. The optimization method for simulating a cochlear implant neural system according to claim 5, It is characterized in that The expression of the optimization objective function of the simulated cochlear implant neural system model is specifically as follows: In the above formula, represents the nonlinear objective function, α represents the unknown coefficient, A represents the feasible domain of the unknown coefficient, t represents the discrete time point, R(a,t) represents the residual vector of the system, T represents the maximum value in the time domain, a * Represents the optimized coefficients.

7. The optimization method for simulating a cochlear implant neural system according to claim 5, It is characterized in that The step of using the truncated series as the initial value of the time domain minimum residual method, optimizing the optimization objective function of the simulated cochlear implant neural system model by the time domain minimum residual method, and obtaining the optimal truncated series specifically includes: The truncated series is used as the initial value of the time domain minimum residual method, and the optimization objective function of the cochlear implant neural system model is linearly transformed to obtain an approximate linear objective function. By introducing Tikhonov regularization, the approximate linear objective function is regularized and solved, and the regularized linear objective function is obtained; The regularization parameter of the regularized linear objective function is determined by the L-curve method, and the confidence region restriction is introduced to screen the regularized linear objective function and determine the optimal truncation series.

8. The optimization method for simulating a cochlear implant neural system according to claim 7, It is characterized in that The regularized linear objective function is expressed as follows: In the above formula, λ represents the regularization parameter, I represents the unit matrix, and δa λ represents the iterative update amount, represents the residual vector, represents the first-order response sensitivity matrix.

9. The optimization method for simulating a cochlear implant neural system according to claim 7, It is characterized in that The expression of the confidence region restriction is specifically as follows: In the above formula, represents the confidence region limit, represents the original nonlinear objective function, represents the objective function after linearization, represents known parameters, a represents unknown coefficients, represents the residual vector.

10. An optimization system for simulating the neural system of a cochlear implant, It is characterized in that Includes the following modules: A first building module is used to build a simulated cochlear implant neural system model, wherein the simulated cochlear implant neural system model includes a piezoelectric ceramic unit and a nonlinear neuron circuit; The second building block is used to perform dynamic behavior analysis on nonlinear neuron circuits and construct dynamic equations to simulate the neural system model of the cochlear implant; A solution module is used to solve the dynamic equations of the simulated cochlear implant neural system model by using the fourth-order Runge-Kutta method to obtain the displacement response of the simulated cochlear implant neural system model; The optimization module is used to optimize the simulated cochlear implant nerve system model according to the displacement response of the simulated cochlear implant nerve system model through the time domain minimum residual method to obtain the optimized simulated cochlear implant nerve system model.

Citation Information

Cited By

  • Time delay optimization method of nonlinear time delay control system for aircraft

    CN120871606A

  • A time delay optimization method for nonlinear time-delay control system of aircraft

    CN120871606B