A method and system for solving quantum electromagnetic problems
By solving the Dahlamper equation and Schrödinger equation in the nanoelectromagnetic system, iterative formulas are derived and spatial filtering operations are performed, the problems of low calculation efficiency and large numerical dispersion error of the traditional FDTD method are solved, and the efficient numerical value in the nanoelectromagnetic system is achieved calculate.
Patent Information
- Application Number
- CN202310365719.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-04-07
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2043-04-07
AI Technical Summary
The traditional FDTD method has low computational efficiency and large numerical dispersion error in nano-electromagnetic systems, making it difficult to meet the CFL stability conditions. Especially in the high-order time domain finite difference method, implicit numerical solution to the A-φ equation is complicated and numerical implementation is difficult.
By solving the Dahlamper equation and the Schrödinger equation under the action of external electromagnetics, the iterative formulas of vector magnetic position, scalar potential, wave function and quantum current are derived, the CFL stability conditions and filter radius are determined, and spatial filtering operations are performed to meet the stability conditions. , and numerical dispersion calculation and analysis were performed.
It realizes that CFL stability conditions are met in nano electromagnetic systems, reduces calculation time, improves calculation efficiency, and reduces numerical dispersion errors.
Smart Images

Figure CN117272708B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of electromagnetic research, and in particular to a method and system for solving quantum electromagnetic problems by high-order time-domain finite difference based on spatial filtering. Background Art
[0002] Computational Electromagnetics (CEM) plays an important role in the modeling and simulation of electromagnetic effects (radiation, scattering or propagation) in various electronic devices today. Its applications include but are not limited to communications, bio-imaging and diagnosis, electromagnetic compatibility or interference, signal integrity analysis, etc. In recent years, the rapid development of modern electronic technology has reduced the characteristic size of integrated circuit-related components to the nanometer level, and the ability to store materials and process information per unit volume has increased by a million times, which has greatly promoted the rapid development of intelligent miniaturized electronic devices such as computers. However, since the spatial size of its nanoscale structure is comparable to the de Broglie wavelength of electrons, the quantum effect can no longer be ignored, and the quantum properties of surface plasmons are constantly being discovered in experiments. Therefore, the modeling scheme based on classical electromagnetic theory can no longer meet the design and simulation needs of nanoelectronic devices. The multi-physics field modeling and simulation of microscopic quantum systems and macroscopic electromagnetic systems has become a scientific problem that urgently needs to be solved in the development of high-frequency, high-speed, broadband and multi-functional integrated electromagnetic components and chip miniaturization, and it is also a core technical problem that needs to be overcome in nanoelectronics and information technology.
[0003] Since the electrons are trapped in the sub-wavelength range, the electronic state will be quantized, so the movement of the electrons should follow the Schrödinger equation. At the same time, the excited electrons generate quantum currents and radiate electromagnetic fields. Therefore, in order to accurately describe the interaction mechanism between electromagnetic fields and microscopic particles, it is necessary to combine the Maxwell equations with the Schrödinger equation. The solution of the MS coupling equation has attracted widespread attention in the past decade. Some researchers have introduced quantum effects into the electromagnetic simulation of nanocircuit systems by solving the MS equation, designed laser pulses to control the discrete quantum states of electrons, and compared and analyzed the influence of quantum radiation on the transport current in nanotubes. In the above studies, the steps for solving the coupled electromagnetic (EM) equations and quantum mechanics (QM) equations are basically the same. First, the Maxwell equations are solved according to the initial conditions to obtain the electric field and magnetic field (E and H); then, the vector magnetic potential (A) and scalar potential (φ) are obtained according to the relationship between the potential function and the electromagnetic field components; then, the potential function (A-φ) is substituted into the Schrödinger equation to calculate the wave function (ψ) and quantum current (J); finally, the quantum current is re-injected into the Maxwell equations as a self-generated source. The self-consistent coupling process of quantum and electromagnetic is repeated until the system reaches a steady state. Since the Schrödinger equation under the action of an external electromagnetic field is only affected by the vector magnetic potential and scalar potential, a set of variant Maxwell equations based on vector magnetic potential and scalar potential are proposed and coupled with the Schrödinger equation to solve multi-physics problems. This method avoids the numerical solution of the electromagnetic field components and the numerical implementation is more direct.
[0004] Equally important, the numerical solution of the MS coupling equation is also the focus of multi-physics problem research. Since both Maxwell's equations and Schrödinger's equation are differential equations in the time domain and space domain, the time-domain finite-difference (FDTD(2,2)) method and its improved method with second-order accuracy in time and space have been proven to be one of the most effective and concise methods for such coupling equations. For example, the FDTD(2,2) method has been successfully used to solve the MS method and simulate the interaction process between artificial atoms (quantum dots) and electromagnetic fields. However, the explicit difference strategy of the traditional FDTD(2,2) method is limited by the grid size in the computational space and needs to meet the CFL stability condition. For nano-electromagnetic systems, the grid size is at the nanometer / subnanometer level, and a smaller time step will drastically reduce the simulation efficiency. To this end, the explicit FDTD(2,2) method and the implicit unconditionally stable FDTD method are combined to solve the MS coupling equations. Due to the unconditional stability characteristics of the implicit FDTD method, a larger time step can be used to improve the computational efficiency. However, it is not easy to numerically solve the A-φ equation using the implicit FDTD method. This is because the A-φ equation has second-order partial derivatives in the time domain and the space domain, making the numerical implementation based on the FDTD method difficult. In addition, the traditional FDTD (2,2) method has only second-order accuracy in time and space, and the numerical dispersion error is large, which will produce large error accumulation in long-term numerical simulations and even cause the numerical results to diverge. In view of the disadvantage of large numerical errors of FDTD (2,2), some scholars have proposed the use of high-order FDTD methods for simulating electromagnetic systems. The high-order FDTD method has the simplicity and easy implementation of the ordinary FDTD (2,2) method, while the numerical dispersion error is improved. However, the HO-FDTD method is an explicit algorithm and still needs to meet the CFL stability condition. Summary of the invention
[0005] The purpose of the present invention is to provide a method and system for solving quantum electromagnetic problems based on high-order time-domain finite difference spatial filtering, which can not only meet the CFL stability conditions, but also shorten the calculation time and improve the calculation efficiency.
[0006] To achieve the above object, the present invention provides the following solutions:
[0007] A method for solving quantum electromagnetic problems based on high-order finite-difference time-domain method with spatial filtering includes:
[0008] The electromagnetic part of the quantum electromagnetic coupling system is solved by solving the self-consistent coupling equations of the D'Alembert equation and the Schrödinger equation under external electromagnetic action to obtain the iterative formulas of the vector magnetic potential, scalar potential, wave function and quantum current of the FDTD method;
[0009] According to the iterative formulas, the stability condition and filter radius of the CFL under the external electromagnetic action are determined;
[0010] Performing a spatial filtering operation on the filtering radius to obtain a filtering radius that satisfies a CFL stability condition;
[0011] According to the filter radius satisfying the CFL stability condition, numerical dispersion calculation and analysis is performed on the quantum electromagnetic coupling system to obtain the numerical dispersion error.
[0012] Optionally, the electromagnetic part in the quantum electromagnetic coupling system is obtained by solving the self-consistent coupling equations of the D'Alembert equation and the Schrödinger equation under external electromagnetic action to obtain the iterative formula of the vector magnetic potential, scalar potential, wave function and quantum current of the FDTD method, specifically including:
[0013] In the case of uniform medium space, the iterative formula for determining the vector magnetic potential in the quantum electromagnetic coupling system is:
[0014]
[0015] Iterative formula for determining the scalar potential in a quantum electromagnetic coupling system:
[0016]
[0017] Where J is the current source term, is the D'Alembert equation, It is a simple Lorentz gauge, where A is the vector magnetic potential, φ is the scalar electric potential, and c0 is the speed of light;
[0018] For the quantum part, under the external electromagnetic effect, the wave function iteration formula of the particle is determined:
[0019]
[0020] Where ψ is the wave function that describes the probability amplitude of the quantum state of a particle at position r and time t in quantum mechanics; m * , and e are the electron mass, reduced Planck constant and electron charge, respectively, V is the confining potential, is the modified Schrödinger equation;
[0021] Iterative formula for determining the quantum current generated by particle motion:
[0022]
[0023] Optionally, the stability condition of the CFL under the electromagnetic action is:
[0024]
[0025]
[0026] Where Δt is the time step, Q max The wave function is restricted to a k-order spatial difference under q-order spatial difference. max is the maximum value of the trigonometric function part in a spherical region with a radius of , when the spatial step size is the same, q is the spatial differential accuracy, Δ is the discretization interval, Δt CFL is the maximum discrete time step of the traditional HO-FDTD(2,q) method to solve the Schrödinger equation. CE is the multiple by which the stability condition of the traditional HO-FDTD(2,q) algorithm is expanded. qm is the expansion multiple of the quantum system.
[0027] The filtering radius is:
[0028]
[0029] Among them, R is the filter radius, k max is the radius of the spherical region to which the wave function is confined.
[0030] Optionally, performing a spatial filtering operation on the filtering radius to obtain a filtering radius that satisfies a CFL stability condition specifically includes:
[0031] According to the filter radius, a spatial filtering operation is added in each numerical iteration process to filter out spatial high-frequency components generated by adopting a large time step that does not meet the CFL stability condition, so as to obtain a filter radius that meets the CFL stability condition.
[0032] A system for solving quantum electromagnetic problems based on high-order finite-difference time-domain methods using spatial filtering includes:
[0033] The module for determining the iterative formula of the quantum electromagnetic coupling system is used to obtain the iterative formula of the vector magnetic potential, scalar potential, wave function and quantum current of the FDTD method by solving the self-consistent coupling equations of the D'Alembert equation and the Schrödinger equation under the external electromagnetic action for the electromagnetic part of the quantum electromagnetic coupling system.
[0034] A stability condition and filter radius determination module, used to determine the stability condition and filter radius of the CFL under external electromagnetic action according to each of the iterative formulas;
[0035] A filter radius determination module that satisfies the CFL stability condition, used for performing a spatial filtering operation on the filter radius to obtain a filter radius that satisfies the CFL stability condition;
[0036] The numerical dispersion error determination module is used to perform numerical dispersion calculation and analysis on the quantum electromagnetic coupling system according to the filter radius that meets the CFL stability condition to obtain the numerical dispersion error.
[0037] Optionally, the quantum electromagnetic coupling system iterative formula determination module specifically includes:
[0038] The iterative formula determination unit of the vector magnetic potential is used to determine the iterative formula of the vector magnetic potential in the quantum electromagnetic coupling system under the condition of uniform medium space:
[0039]
[0040] The iterative formula determination unit of the scalar potential is used to determine the iterative formula of the scalar potential in the quantum electromagnetic coupling system:
[0041]
[0042] Where J is the current source term, is the D'Alembert equation, It is a simple Lorentz gauge, where A is the vector magnetic potential, φ is the scalar electric potential, and c0 is the speed of light;
[0043] The wave function iteration formula determination unit is used to determine the wave function iteration formula of the particle under the external electromagnetic action:
[0044]
[0045] Where ψ is the wave function that describes the probability amplitude of the quantum state of a particle at position r and time t in quantum mechanics; m * , and e are the electron mass, reduced Planck constant and electron charge, respectively, V is the confining potential, is the modified Schrödinger equation;
[0046] Iterative formula for quantum current, used to determine the iterative formula for quantum current generated by particle motion:
[0047]
[0048] Optionally, the stability condition of the CFL under the electromagnetic action is:
[0049]
[0050]
[0051] Where Δt is the time step, Q max The wave function is restricted to a k-order spatial difference under q-order spatial difference. maxis the maximum value of the trigonometric function part in a spherical region with a radius of , when the spatial step size is the same, q is the spatial differential accuracy, Δ is the discretization interval, Δt CFL is the maximum discrete time step of the traditional HO-FDTD(2,q) method to solve the Schrödinger equation. CE is the multiple by which the stability condition of the traditional HO-FDTD(2,q) algorithm is expanded. qm is the expansion multiple of the quantum system.
[0052] The filtering radius is:
[0053]
[0054] Among them, R is the filter radius, k max is the radius of the spherical region to which the wave function is confined.
[0055] Optionally, the filter radius determination module satisfying the CFL stability condition specifically includes:
[0056] The spatial filtering operation unit is used to add a spatial filtering operation in each numerical iteration process according to the filtering radius, filter out the spatial high-frequency components generated by adopting a large time step that does not meet the CFL stability condition, and obtain a filtering radius that meets the CFL stability condition.
[0057] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:
[0058] The present invention provides a method for solving quantum electromagnetic problems based on high-order time-domain finite difference spatial filtering, wherein the electromagnetic part in the quantum electromagnetic coupling system is obtained by solving the self-consistent coupling equations of the D'Alembert equation and the Schrödinger equation under external electromagnetic action to obtain the iterative formulas of the vector magnetic potential, scalar potential, wave function and quantum current of the FDTD method; according to each iterative formula, the CFL stability condition and the filter radius under the external electromagnetic action are determined; the filter radius is subjected to spatial filtering operation to obtain the filter radius that meets the CFL stability condition; according to the filter radius that meets the CFL stability condition, the quantum electromagnetic coupling system is subjected to numerical dispersion calculation and analysis to obtain the numerical dispersion error. The present invention can not only meet the CFL stability condition, but also reduce the calculation time and improve the calculation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0060] Figure 1 It is a flow chart of the method for solving quantum electromagnetic problems based on high-order finite-difference time-domain of spatial filtering of the present invention;
[0061] Figure 2 k max When Δ=π / 10, the q-order spatial difference Follow The curve of change;
[0062] Figure 3 is the q-order spatial difference under CE with k max Variation curve of Δ;
[0063] Figure 4 For different CE em The filter radius is determined by the value of ;
[0064] Figure 5 Numerical dispersion error of SF-HO-FDTD(2,q) method when B=0.2,CE=3;
[0065] Figure 6 When B = 0.2, the numerical dispersion error of the SF-HO-FDTD (2,4) (q = 4) method changes with CE;
[0066] Figure 7 The numerical dispersion error under q-order spatial difference when B=0.3 and CE=3;
[0067] Figure 8 When B = 0.05, the variation curve of numerical dispersion error with CE under fourth-order spatial difference;
[0068] Fig. 9 Comparison diagram of vector magnetic potential of observation points at different time steps;
[0069] Fig.10 is the eigenstate of the Schrödinger equation at different time steps;
[0070] Fig.11 is the geometric model of the nanotube;
[0071] Fig.12 Schematic diagram of the eigenwave functions of the ground state and the first excited state;
[0072] Fig.13 is the probability distribution of electrons in the nanotube at different time steps;
[0073] Fig.14 It is the conversion diagram between the ground state and the first excited state at different time steps;
[0074] Fig.15 is the probability wave packet of the second excited state and the third excited state of the electron in the nanotube;
[0075] Fig.16 is the probability distribution of electrons in the nanotube at different time steps;
[0076] Fig.17 is the change of the proportion of multi-state quantum states over time at different time steps;
[0077] Fig.18 This is a system structure diagram of the present invention for solving quantum electromagnetic problems based on high-order time-domain finite difference spatial filtering. DETAILED DESCRIPTION
[0078] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.
[0079] The purpose of the present invention is to provide a method and system for solving quantum electromagnetic problems based on high-order time-domain finite difference spatial filtering, which can not only meet the CFL stability conditions, but also shorten the calculation time and improve the calculation efficiency.
[0080] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0081] Figure 1 The flowchart of the method for solving quantum electromagnetic problems based on high-order finite-difference time-domain spatial filtering of the present invention is shown in FIG. Figure 1 As shown, a method for solving quantum electromagnetic problems based on high-order finite-difference time-domain spatial filtering includes:
[0082] Step 101: The electromagnetic part in the quantum electromagnetic coupling system is solved by solving the self-consistent coupling equations of the D'Alembert equation and the Schrödinger equation under the external electromagnetic action to obtain the iterative formulas of the vector magnetic potential, scalar potential, wave function and quantum current of the FDTD method, which specifically include:
[0083] In the case of uniform medium space, the iterative formula for determining the vector magnetic potential in the quantum electromagnetic coupling system is:
[0084]
[0085] The time evolution of the scalar potential (φ) can be described by the following equation:
[0086]
[0087] Where J represents the current source term, formula (1) is the D'Alembert equation, and formula (2) is the simple Lorentz norm.
[0088] For the quantum part, under the external electromagnetic effect, the modified Schrödinger equation can be written as follows:
[0089]
[0090] where ψ is the wave function of the particle; m * , and e are the electron mass, the reduced Planck constant and the electron charge, respectively, and V is the confining potential. At the same time, the quantum current generated by particle motion can be expressed as:
[0091]
[0092] Formulas (1)-(4) can construct a self-consistent quantum-electromagnetic multi-physics coupling system to simulate the interaction process between light and matter.
[0093] Step 102: According to the iterative formulas, the stability condition and filter radius of the CFL under the external electromagnetic action are determined.
[0094] In formulas (1) and (2), the partial derivative of the potential function (A-φ) in time is approximated by the second-order difference, and its discrete form can be written as follows:
[0095]
[0096] Where n represents the sampling time.
[0097] Without loss of generality, without considering the current source, formula (5) can be expressed in the following matrix form:
[0098] P n+1 =[M r ]P n +[M s ]P n-1 (6)
[0099] in
[0100] P=(A x ,A y ,A z ,φ) T (7)
[0101]
[0102]
[0103]
[0104] in
[0105] A three-dimensional plane wave is defined as:
[0106]
[0107]
[0108]
[0109]
[0110] Where Φ represents the amplitude of each field component, ζ is the growth factor, and k u (u=x,y,z) represents the wave number.
[0111] make Its discrete form in space can be expressed as
[0112]
[0113] The median function in formula (5) is spatially discretized as shown in formulas (11)-(14), and its spatial first-order and second-order partial derivatives are approximated by q-order differences (with A x Direction as an example):
[0114]
[0115]
[0116] Among them C s and D s is the q-order juxtaposed difference coefficient of the first-order spatial partial derivative and the second-order spatial partial derivative, and determines the coefficient C of the q-order juxtaposed difference s and D s , Table 1 and Table 2 only give the differential coefficients when q=2,4,6,8.
[0117] Table 1 q-order juxtaposed difference coefficients of the first-order derivative
[0118]
[0119] Table 2 Q-order juxtaposed difference coefficients of the second-order derivative
[0120]
[0121] Using the same differential approximation method, other electromagnetic components (A y , A z The spatial discretization formulas for φ and φ) are similar to those of formulas (16) and (17).
[0122] Next, substitute formulas (7)-(14), (16)-(17) and the discrete formulas of electromagnetic components in other field directions into formula (6) and simplify to obtain:
[0123] [ζ 2 -M r ζ-M s ]P n-1 =MP n-1 =0 (18)
[0124] in:
[0125]
[0126]
[0127]
[0128] In order to obtain a non-zero solution to formula (18), the determinant of its coefficient matrix M must be 0. By simplifying formula (19), we can obtain:
[0129]
[0130] According to formula (22), the value of the growth factor ζ is
[0131]
[0132] In order to ensure the stability of the numerical algorithm, the modulus of the growth factor ζ must be less than or equal to 1. Obviously, the situation of ζ1|=0, |ζ2|=1 is established, and in order to make |ζ 3,4 |=|ζ 5,6 |=|ζ 7,8 |≤1, must satisfy:
[0133]
[0134] Substituting formula (21) into formula (24), we get:
[0135]
[0136] Pick The maximum value of is β, that is
[0137]
[0138] We can get the inequality about the time step Δt:
[0139]
[0140] Wherein, Δx = Δy = Δz = Δ. Formula (27) is the time stability condition for numerically solving the potential function equation (Formula (1) and Formula (2)) using the HO-FDTD (2,q) method.
[0141] For the modified Schrödinger equation under the action of an external electromagnetic field, formula (3) can be simplified into the following form:
[0142]
[0143] Formula (27) can be written as an eigenvalue problem, namely
[0144]
[0145] Here λ is the eigenvalue:
[0146]
[0147] The second-order difference approximation of the time partial derivative of formula (29) is:
[0148]
[0149] In order to ensure the stability of the system, the growth factor g Need to meet:
[0150] |g|=|ψ n+1 / ψ n |≤1 (32)
[0151] Substituting formula (32) into formula (31), we get:
[0152]
[0153] In order to make |g|≤1, we must have
[0154]
[0155] Consider the case where the wave function is a plane wave, that is By discretizing it in space, we can get:
[0156]
[0157] The same processing method as the electromagnetic part is used to approximate the first-order and second-order spatial partial derivatives of the wave function by q-order difference (taking the x direction as an example):
[0158]
[0159]
[0160] The wave function is treated the same way in other directions.
[0161] According to equations (36) and (37), λ in equation (30) can be derived as follows:
[0162]
[0163] At the same time, according to inequality (34) The inequality about the time step can be written as follows:
[0164]
[0165] in
[0166]
[0167]
[0168] Formula (39) is the time stability condition for numerically solving the modified Schrödinger equation (Formula (3)) using the HO-FDTD (2,q) method. It can be seen that the value of the time step is not only related to the grid size, but also to the field components A, φ and the potential well V. If the influence of the external electromagnetic field is not considered (A = 0 and φ = 0), and according to formula (26) Can get
[0169]
[0170] Formula (42) is the time stability condition for numerically solving the Schrödinger equation using the HO-FDTD(2,q) method without external electromagnetic action.
[0171] For electromagnetic systems, it can be seen from formula (27) that the time stability condition must be met to solve the potential function equation using the HO-FDTD(2,q) method. However, if in formula (25), it is assumed that the wave function is restricted to a k max In the spherical region with radius
[0172]
[0173] Assuming Δx=Δy=Δz=Δ, substitute formula (43) into formula (25) to obtain:
[0174]
[0175] Where d = k max Δ / 2.
[0176] make
[0177]
[0178] Then the inequality about the time step can be written as follows:
[0179]
[0180] With θ, The changing curve is Figure 2 As shown, through numerical solution, it is found that for q = 4, 6, 8, All at θ = 0.304π, The maximum value Q is obtained when max (0.304π,0.25π). max Substituting (0.304π, 0.25π) into formula (46), the derived inequality about the time step is:
[0181]
[0182]
[0183] Among them, Δt CFL is the maximum discrete time step of the traditional HO-FDTD(2,q) method to solve the potential function equation, and CE represents the multiple by which the stability condition of the traditional HO-FDTD(2,q) algorithm is extended. Substituting into formula (46), and substituting q = 4, 6, 8 into formulas (26) and (45), respectively, we can obtain the numerical variation curve of CE as follows: Figure 3 As shown, it can be found that for q = 4, 6, 8, the value of CE is greater than 1. This shows that the temporal stability condition of the traditional HO-FDTD (2, q) method can be further extended through spatial filtering.
[0184] For quantum systems, using a similar approach as above, the time stability condition for numerically solving the Schrödinger equation using the HO-FDTD(2,q) method (Equation (42)) can be expanded to:
[0185]
[0186]
[0187] For formulas (48) and (50), when the value of CE is determined, k can be obtained max The value of Δ is used to determine the filter radius R. The calculation formula for R is:
[0188]
[0189] Step 103: performing a spatial filtering operation on the filtering radius to obtain a filtering radius that satisfies a CFL stability condition, specifically including:
[0190] According to the filter radius, a spatial filtering operation is added in each numerical iteration process to filter out spatial high-frequency components generated by adopting a large time step that does not meet the CFL stability condition, so as to obtain a filter radius that meets the CFL stability condition.
[0191] Taking the electromagnetic system as an example, when CE em After the value of is determined, the filter radius R can be determined according to formula (51): Figure 4 For different CE em The filter radius is determined by the value of . It can be found that CE em The larger the value of , the smaller R is, which indicates that more high-frequency components need to be filtered out to ensure the stability of the numerical algorithm.
[0192] Step 104: performing numerical dispersion calculation analysis on the quantum electromagnetic coupling system according to the filter radius that satisfies the CFL stability condition to obtain a numerical dispersion error.
[0193] The following numerical dispersion analysis is performed through quantum system and electromagnetic system respectively:
[0194] Quantum Systems:
[0195] Without loss of generality, without considering the influence of external electromagnetic field (A = 0 and φ = 0), the dispersion relation of equation (3) solved by high-order FDTD (2, q) method can be written as follows:
[0196]
[0197] In the formula,
[0198] In formula (52), replace Δt qm Convert to Δt qm_SF , and substituting equations (43) and (49) into equation (52), we can obtain the numerical dispersion relation (Δx=Δy=Δz=Δ) for solving the Schrödinger equation using the SF-HO-FDTD(2,q) method:
[0199]
[0200] The analytical dispersion relation of the Schrödinger equation is:
[0201]
[0202] Definition of numerical dispersion error Err qm :
[0203]
[0204] Definition: Δ=B / k max, the number of points per wavelength is given by 2π / B. Substituting equations (49), (53) and (54) into equation (55) yields (V = 0):
[0205]
[0206] Figure 5 When CE = 3, the numerical dispersion error of the SF-HO-FDTD(2,q) method for solving the Schrödinger equation. It can be seen that when the time step of the SF-HO-FDTD(2,q) method (q = 4, 6, 8) method is 3 times the maximum time step of the traditional HO-FDTD(2,q) method (Δt em_SF =Δt em_CFL ×3), the SF-HO-FDTD(2,q) method still has a low numerical dispersion error. At the same time, the numerical dispersion error of the SF-FDTD(2,2) method when q=2 is also given ( Figure 5 .(a)), which shows that the use of spatial differential approximation method with higher computational accuracy can reduce the numerical dispersion error.
[0207] Figure 6 The numerical dispersion error (CE=1,5,7,10) of the SF-HO-FDTD(2,4)(q=4) method at different time steps is given. It can be found that as the CE increases, the numerical error becomes larger, which is consistent with the actual situation. But it is worth noting that even when CE=10, the SF-HO-FDTD(2,4) method still has a lower numerical dispersion error.
[0208] Electromagnetic system:
[0209] In formula (22), let the growth factor ζ = e jωΔt , where ω represents the angular frequency, we can get:
[0210]
[0211] in, According to the method introduced in the quantum part, the numerical dispersion error of solving the potential function equation (Equations (1) and (2)) using the SF-HO-FDTD(2,q) method can be obtained:
[0212]
[0213] Figure 7 When CE = 3, the numerical dispersion error of the SF-HO-FDTD(2,q) method in solving the potential function equation. It can be seen that when the time step of the SF-HO-FDTD(2,q) method (q = 4, 6, 8) method is 3 times the maximum time step of the traditional HO-FDTD(2,q) method (Δt qm_SF=Δt qm_CFL ×3), the SF-HO-FDTD(2,q) method still has a low numerical dispersion error. At the same time, the numerical dispersion error of the SF-FDTD(2,2) method when q=2 is also given ( Figure 7 .(a)), which is the same as the conclusion of the quantum part. The use of the spatial differential approximation method with higher computational accuracy can reduce the numerical dispersion error.
[0214] Figure 8 The numerical dispersion error (CE=1,5,7,10) of the SF-HO-FDTD(2,4)(q=4) method at different time steps is given. It can be found that as the CE increases, the numerical error becomes larger, which is consistent with the actual situation. But it is worth noting that even when CE=10, the SF-HO-FDTD(2,4) method still has a lower numerical dispersion error.
[0215] Through the numerical dispersion analysis of the SF-HO-FDTD(2,q) method, it can be found that the SF-HO-FDTD(2,q) method has a lower numerical dispersion error. At the same time, similar to the numerical characteristics of the traditional HO-FDTD(2,q) method, the use of higher-order numerical difference approximation methods will reduce the numerical dispersion error. However, in practical applications, higher-order numerical approximation methods often reduce numerical calculation efficiency because they have higher numerical complexity. Therefore, in general numerical simulations, it is necessary to ensure the calculation accuracy while taking into account the numerical calculation efficiency. .
[0216] Example 1: Dipole Source
[0217] In order to verify the correctness and effectiveness of the SF-HO-FDTD(2,q) method proposed in the present invention, the SF-HO-FDTD(2,q) method is first used to solve the potential function equation (Formula (1) and Formula (2)) to simulate the electromagnetic radiation problem of the electric dipole source in three-dimensional space. Without loss of generality, q=4 is selected in all examples of the present invention. The electric dipole is located at the center of the simulation space. The simulation area contains 70×70×70 grids, and the grid size Δx=Δy=Δz=5cm. The calculation space is truncated using the Mur absorption boundary condition. The detection point is 50cm away from the radiation source. The radiation source uses a Gaussian pulse in the form of
[0218] The potential function equation was numerically solved using the traditional HO-FDTD(2,4) method (CE=1) and the SF-HO-FDTD(2,4) method (CE=5,10) to obtain the potential function equation at the detection point A. z The time domain waveform is as follows Fig. 9It can be found that when the time step used by the SF-HO-FDTD(2,4) method is 10 times the maximum discrete time step of the traditional HO-FDTD(2,4) method, the calculation results are still in good agreement with the calculation results of the traditional HO-FDTD(2,4) method, which verifies the correctness of the SF-HO-FDTD(2,4) method in numerically solving the potential function equation.
[0219] Example 2: 3D Quantum Well
[0220] The second example is a three-dimensional quantum well model, which studies its eigenfrequency and eigenstate problems. The size of the three-dimensional quantum well model is L x =L y =L z =30nm, spatial step length Δx=Δy=Δz=1nm. The boundary condition ψ(0,y,z)=ψ(L x ,y,z)=ψ(x,0,z)=ψ(x,L y ,z)=ψ(x,y,0)=ψ(x,y,L z )=0, its potential energy V(x,y,z)=0. According to formula (42), the maximum time step of the traditional HO-FDTD(2,4) method for solving the Schrödinger equation can be obtained. Table 3 and Fig.10 The three-dimensional eigenfrequencies and eigenstates ψ calculated by the traditional HO-FDTD(2,4) method (CE=1) and the SF-HO-FDTD(2,4) method (CE=2,3,5) are given respectively. 2,2,2 Table 3 and Fig.10 It is shown that when the time step adopted by the SF-HO-FDTD(2,4) method is 5 times the maximum discrete time step of the traditional FDTD(2,4) method, relatively accurate numerical calculation results can still be obtained.
[0221] Table 3 Comparison of characteristic frequency results of three-dimensional quantum wells calculated by HO-FDTD (2,4) and SF-HO-FDTD (2,4)
[0222]
[0223] Example 3: Quantum-electromagnetic multi-physics coupling model
[0224] Quantum-electromagnetic multiphysics coupling models such as Fig.11 As shown in the figure, the electrons confined in the nanotubes will change their quantum states under the influence of an external electromagnetic field. At the same time, the electron wave packets will move. The quantum current generated by the movement will be superimposed on the electromagnetic excitation source and act on the electromagnetic system again. The interaction process between the electromagnetic system and the quantum system forms a self-consistent multi-physics coupling system. Fig.11In the quasi-one-dimensional model shown, the size of the electromagnetic space is 120Δx×120Δy×120Δz, and the calculation space is truncated using the Mur absorbing boundary condition. The spatial size of the nanotube is Δx×Δy×100Δz, and the potential well expression added in the z direction is:
[0225]
[0226] Where, V0 = 5000eV, z max =1nm.
[0227] In this potential well, by solving the time-invariant Schrödinger equation, the eigenwave functions of the ground state and the first excited state can be obtained, such as Fig.12 shown.
[0228] In multi-physics simulation, in order to ensure the stability of the numerical results, the entire simulation system uses a unified time step, which is not long enough to satisfy the CFL conditions of both the electromagnetic system and the quantum system. According to the simulation parameters and the CFL stability conditions of the electromagnetic system, the maximum discrete time step Δt of the electromagnetic system -em =3.0×10 -5 fs. Similarly, according to the CFL condition of the quantum system, Δt -qm =4.3×10 -5 fs. It can be seen that in this simulation environment, the CFL stability condition of the quantum system is 15 times that of the electromagnetic system. Therefore, in order to ensure the stability of the numerical results, when the traditional HO-FDTD (2,4) method is used to numerically solve equations (1)-(3), the time step Δt used in the entire simulation system is 3.0×10 -5 When the SF-HO-FDTD(2,4) method is used to numerically solve equations (1)-(3), the time step used in the entire simulation system is 15 times, 30 times, and 60 times that of the traditional HO-FDTD(2,4) method (CE em =15,30,60). Based on this, in order to ensure the stability of the numerical results, when the SF-HO-FDTD(2,4) method is used to numerically solve formula (3), CE qm They are 1, 2, and 4 respectively.
[0229] Fig.13 is the time-varying process of the waveform of the electron wave packet in the nanotube, where Fig.13 (a) is the result calculated by the traditional HO-FDTD (2,4) method. Fig.13 (b), (c), and (d) show the SF-HO-FDTD (2,4) method with different time steps. Fig.13(a) It can be found that the electron wave packet exists in a single peak form at the initial moment (corresponding to the ground state), oscillates rapidly in the 2-8s period, and finally appears in a double peak form after 10s (corresponding to the first excited state). At the same time, in order to quantitatively describe the conversion rate of the electron quantum state, the conversion rate factor is defined as follows
[0230]
[0231] Among them, y k Represents the quantum state of an electron.
[0232] Fig.14 The figure shows the conversion rate of the electronic quantum state calculated by the SF-HO-FDTD(2,4) method and the traditional HO-FDTD(2,4) method with different time steps. It can be found that the perfect transition of the electronic quantum state can be achieved by numerically solving the multi-physics field coupling equations using the traditional HO-FDTD(2,4) method and the SF-HO-FDTD(2,4) method. Fig.13 and Fig.14 It can be found that the calculation results of the SF-HO-FDTD (2,4) method are in good agreement with those of the traditional HO-FDTD (2,4) method, which also verifies the correctness of the SF-HO-FDTD (2,4) method proposed in this invention for solving multi-physics equations.
[0233] In addition, the SF-HO-FDTD(2,4) method was used to numerically solve the multi-physics coupling system and simulate the multi-state conversion process of electronic quantum states. Fig.15 is the probability wave packet of the second excited state and the third excited state of the electron. Fig.16 and Fig.17 They are respectively the time-varying process of the waveform of the electron wave packet in the nanotube and the conversion rate of the electron quantum state. Fig.16 and Fig.17 It can be found that the use of the SF-HO-FDTD (2,4) method to solve the multi-physics coupling equation can achieve a perfect transition of the electron quantum state, and the calculation results are consistent with those of the traditional HO-FDTD (2,4) method, verifying the correctness of the SF-HO-FDTD (2,4) method of the present invention.
[0234] At the same time, the numerical calculation efficiency of the HO-FDTD (2,4) method and the SF-HO-FDTD (2,4) method is given for the numerical solution of the above multi-physics field coupling equations. As can be seen from Table 4, when CE em When =15, the calculation time of the SF-HO-FDTD(2,4) method is less than that of the traditional HO-FDTD(2,4) method, which indicates that the SF-HO-FDTD(2,4) method proposed in the present invention has higher calculation efficiency.
[0235] Table 4 Execution time of multi-state transitions of quantum states calculated by FDTD(2,4) and SF-FDTD(2,4) methods
[0236]
[0237] Compared with classical theory, the quantum-electromagnetic multi-physics coupling model based on semiclassical theory requires the joint solution of more mathematical equations, so it is difficult to implement numerically. Although the traditional FDTD(2,2) method and its improved method can effectively solve these multi-physics coupling equations, the traditional FDTD(2,2) method has large numerical errors and is prone to large error accumulation in long-term numerical iterations. Moreover, the traditional FDTD(2,2) method is an explicit algorithm, and its time step is limited by the grid size in the calculation space, resulting in poor computational efficiency. The HO-FDTD(2,q) method with high computational accuracy in space can effectively reduce the numerical dispersion error of the FDTD(2,2) method, but it is still an explicit algorithm and needs to meet the constraints of time stability conditions, so the computational efficiency needs to be further improved. Compared with the explicit method, the implicit FDTD method has the characteristics of unconditional stability, that is, the time step is not limited by the grid size. Using a larger time step in numerical simulation can improve computational efficiency. However, the implicit FDTD method often requires special time and space discretization of multi-physics equations, and the derived iterative formula is often more complex, and the numerical implementation is more difficult than the explicit method. The SF-HO-FDTD(2,q) method proposed in the present invention not only retains the high-precision characteristics of the traditional HO-FDTD(2,q) method in space, but also breaks through the limitations of its time stability conditions. In addition, the SF-HO-FDTD(2,q) method does not need to modify the numerical iteration formula of the traditional HO-FDTD(2,q) method, but only needs to add a filtering operation in each numerical iteration, so the numerical implementation is relatively easy. In general, the innovation points of the present invention mainly include the following two points: 1) The CFL stability conditions for processing the potential function equation and the modified Schrödinger equation based on the HO-FDTD(2,q) method are derived; 2) The SF method is applied to HO-FDTD(2,q), and the SF-HO-FDTD(2,q) method is proposed and applied to solve multi-physics coupling equations. The three numerical examples proposed in the paper, including the electric dipole radiation problem, the quantum well eigenstate problem and the multi-physical field coupling problem of electron quantum state regulation, all effectively demonstrate the correctness and effectiveness of the method proposed in this invention.
[0238] Fig.18 This is a system structure diagram of the invention for solving quantum electromagnetic problems based on high-order finite-difference time-domain spatial filtering. Fig.18 As shown, a system for solving quantum electromagnetic problems based on high-order finite-difference time-domain spatial filtering includes:
[0239] The quantum electromagnetic coupling system iterative formula determination module 201 is used to obtain the iterative formula of the vector magnetic potential, scalar potential, wave function and quantum current of the FDTD method by solving the self-consistent coupling equations of the D'Alembert equation and the Schrödinger equation under the external electromagnetic action for the electromagnetic part in the quantum electromagnetic coupling system;
[0240] A stability condition and filter radius determination module 202, for determining the stability condition and filter radius of the CFL under an external electromagnetic effect according to the iterative formulas;
[0241] A filter radius determination module 203 that satisfies the CFL stability condition is used to perform a spatial filtering operation on the filter radius to obtain a filter radius that satisfies the CFL stability condition;
[0242] The numerical dispersion error determination module 204 is used to perform numerical dispersion calculation analysis on the quantum electromagnetic coupling system according to the filter radius that meets the CFL stability condition to obtain the numerical dispersion error.
[0243] The quantum electromagnetic coupling system iterative formula determination module 201 specifically includes:
[0244] The iterative formula determination unit of the vector magnetic potential is used to determine the iterative formula of the vector magnetic potential in the quantum electromagnetic coupling system under the condition of uniform medium space:
[0245]
[0246] The iterative formula determination unit of the scalar potential is used to determine the iterative formula of the scalar potential in the quantum electromagnetic coupling system:
[0247]
[0248] Where J is the current source term, is the D'Alembert equation, It is a simple Lorentz gauge, where A is the vector magnetic potential, φ is the scalar electric potential, and c0 is the speed of light;
[0249] The wave function iteration formula determination unit is used to determine the wave function iteration formula of the particle under the external electromagnetic action:
[0250]
[0251] Where ψ is the wave function that describes the probability amplitude of the quantum state of a particle at position r and time t in quantum mechanics; m * , and e are the electron mass, reduced Planck constant and electron charge, respectively, V is the confining potential, is the modified Schrödinger equation;
[0252] Iterative formula for quantum current, used to determine the iterative formula for quantum current generated by particle motion:
[0253]
[0254] The stability condition of CFL under the electromagnetic action is:
[0255]
[0256]
[0257] Where Δt is the time step, Q max The wave function is restricted to a k-order spatial difference under q-order spatial difference. max is the maximum value of the trigonometric function part in a spherical region with a radius of , when the spatial step size is the same, q is the spatial differential accuracy, Δ is the discretization interval, Δt CFL is the maximum discrete time step of the traditional HO-FDTD(2,q) method to solve the Schrödinger equation. CE is the multiple by which the stability condition of the traditional HO-FDTD(2,q) algorithm is expanded. qm is the expansion multiple of the quantum system.
[0258] The filtering radius is:
[0259]
[0260] Among them, R is the filter radius, k max is the radius of the spherical region to which the wave function is confined.
[0261] The filter radius determination module 203 that satisfies the CFL stability condition specifically includes:
[0262] The spatial filtering operation unit is used to add a spatial filtering operation in each numerical iteration process according to the filtering radius, filter out the spatial high-frequency components generated by adopting a large time step that does not meet the CFL stability condition, and obtain a filtering radius that meets the CFL stability condition.
[0263] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.
[0264] The present invention uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only used to help understand the method and core ideas of the present invention. At the same time, for those skilled in the art, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.
Claims
1. A method for solving quantum electromagnetic problems based on high-order finite-difference time-domain method with spatial filtering, characterized in that: include: The electromagnetic part of the quantum electromagnetic coupling system is solved by solving the self-consistent coupling equations of the D'Alembert equation and the Schrödinger equation under external electromagnetic action to obtain the iterative formulas of the vector magnetic potential, scalar potential, wave function and quantum current of the FDTD method; According to the iterative formulas, the stability condition and filter radius of the CFL under the external electromagnetic action are determined; Performing a spatial filtering operation on the filtering radius to obtain a filtering radius that satisfies a CFL stability condition; According to the filter radius that satisfies the CFL stability condition, a numerical dispersion calculation analysis is performed on the quantum electromagnetic coupling system to obtain a numerical dispersion error; The electromagnetic part in the quantum electromagnetic coupling system is obtained by solving the self-consistent coupling equations of the D'Alembert equation and the Schrödinger equation under the external electromagnetic action to obtain the iterative formula of the vector magnetic potential, scalar potential, wave function and quantum current of the FDTD method, which specifically includes: In the case of uniform medium space, the iterative formula for determining the vector magnetic potential in the quantum electromagnetic coupling system is: Iterative formula for determining the scalar potential in a quantum electromagnetic coupling system: Where J is the current source term, is the D'Alembert equation, It is a simple Lorentz gauge, where A is the vector magnetic potential, φ is the scalar electric potential, and c0 is the speed of light; For the quantum part, under the external electromagnetic effect, the wave function iteration formula of the particle is determined: Where ψ is the wave function that describes the probability amplitude of the quantum state of a particle at position r and time t in quantum mechanics; m * , and e are the electron mass, the reduced Planck constant and the electron charge, respectively, V is the limiting potential, is the modified Schrödinger equation; Iterative formula for determining the quantum current generated by particle motion:
2. The method for solving quantum electromagnetic problems based on high-order finite-difference time-domain method using spatial filtering according to claim 1 is characterized in that: The stability condition of CFL under the electromagnetic action is: Where Δt is the time step, Q max The wave function is restricted to a k-order spatial difference under q-order spatial difference. max is the maximum value of the trigonometric function part in a spherical region with a radius of , when the spatial step size is the same, q is the spatial differential accuracy, Δ is the discretization interval, ΔtCFL is the maximum discrete time step of the traditional HO-FDTD(2,q) method for solving the Schrödinger equation, and CE represents the multiple by which the stability condition of the traditional HO-FDTD(2,q) algorithm is expanded. qm is the expansion multiple of the quantum system; The filtering radius is: Among them, R is the filter radius, k max is the radius of the spherical region to which the wave function is confined.
3. The method for solving quantum electromagnetic problems based on high-order finite-difference time-domain method using spatial filtering according to claim 1 is characterized in that: The performing of a spatial filtering operation on the filtering radius to obtain a filtering radius that satisfies a CFL stability condition specifically includes: According to the filter radius, a spatial filtering operation is added in each numerical iteration process to filter out spatial high-frequency components generated by adopting a large time step that does not meet the CFL stability condition, so as to obtain a filter radius that meets the CFL stability condition.
4. A system for solving quantum electromagnetic problems based on high-order finite-difference time-domain spatial filtering, characterized in that: include: The module for determining the iterative formula of the quantum electromagnetic coupling system is used to obtain the iterative formula of the vector magnetic potential, scalar potential, wave function and quantum current of the FDTD method by solving the self-consistent coupling equations of the D'Alembert equation and the Schrödinger equation under the external electromagnetic action for the electromagnetic part of the quantum electromagnetic coupling system. A stability condition and filter radius determination module, used to determine the stability condition and filter radius of the CFL under external electromagnetic action according to each of the iterative formulas; A filter radius determination module that satisfies the CFL stability condition, used for performing a spatial filtering operation on the filter radius to obtain a filter radius that satisfies the CFL stability condition; A numerical dispersion error determination module is used to perform numerical dispersion calculation and analysis on the quantum electromagnetic coupling system according to the filter radius that satisfies the CFL stability condition to obtain a numerical dispersion error; The quantum electromagnetic coupling system iterative formula determination module specifically includes: The iterative formula determination unit of the vector magnetic potential is used to determine the iterative formula of the vector magnetic potential in the quantum electromagnetic coupling system under the condition of uniform medium space: The iterative formula determination unit of the scalar potential is used to determine the iterative formula of the scalar potential in the quantum electromagnetic coupling system: Where J is the current source term, is the D'Alembert equation, It is a simple Lorentz gauge, where A is the vector magnetic potential, φ is the scalar electric potential, and c0 is the speed of light; The wave function iteration formula determination unit is used to determine the wave function iteration formula of the particle under the external electromagnetic action: Where ψ is the wave function that describes the probability amplitude of the quantum state of a particle at position r and time t in quantum mechanics; m * , and e are the electron mass, the reduced Planck constant and the electron charge, respectively, V is the limiting potential, is the modified Schrödinger equation; Iterative formula for quantum current, used to determine the iterative formula for quantum current generated by particle motion:
5. The system for solving quantum electromagnetic problems based on high-order finite-difference time-domain method using spatial filtering according to claim 4 is characterized in that: The stability condition of CFL under the electromagnetic action is: Where Δt is the time step, Q max The wave function is restricted to a k-order spatial difference under q-order spatial difference. max is the maximum value of the trigonometric function part in a spherical region with a radius of , when the spatial step size is the same, q is the spatial differential accuracy, Δ is the discretization interval, ΔtCFL is the maximum discrete time step of the traditional HO-FDTD(2,q) method for solving the Schrödinger equation, and CE represents the multiple by which the stability condition of the traditional HO-FDTD(2,q) algorithm is expanded. qm is the expansion multiple of the quantum system; The filtering radius is: Where R is the filter radius, k max is the radius of the spherical region to which the wave function is confined.
6. The system for solving quantum electromagnetic problems based on high-order finite-difference time-domain method using spatial filtering according to claim 4, characterized in that: The filtering radius determination module satisfying the CFL stability condition specifically includes: The spatial filtering operation unit is used to add a spatial filtering operation in each numerical iteration process according to the filtering radius, filter out the spatial high-frequency components generated by adopting a large time step that does not meet the CFL stability condition, and obtain a filtering radius that meets the CFL stability condition.
Citation Information
Patent Citations
Spatial domain electromagnetic component determination method and system
CN111460593A
Electromagnetic problem determination method and system and storage medium
CN113468759A
