Molecular dynamics quantum variation simulation method and device

Abstract

The invention provides a molecular dynamics quantum variation simulation method and device.The molecular dynamics quantum variation simulation method comprises the steps that an initial quantum state of a target molecular system is measured to obtain a position basis vector set, then the position basis vector set and momentum basis vectors of the initial quantum state are used for constructing a parameterized quantum circuit, and the parameterized quantum circuit is obtained; and iteratively updating the parameters in the parameterized quantum circuit until the target evolution time is reached, thereby obtaining the final quantum state of the target molecular system. According to the method, a large-scale molecular simulation problem can be solved on a shallow quantum circuit, it is guaranteed that the circuit depth is only related to the Hamiltonian of a system and does not increase along with simulation time, and due to the fact that the depth of the quantum circuit is not related to evolution time, the quantum circuit depth can be improved. The method provided by the invention can be applied to simulating the processes of structural evolution, phase change and the like of the material under a long time scale.

Description

Technical Field

[0001] One or more embodiments of the present specification relate to the field of molecular dynamics, and in particular, to a method and apparatus for molecular dynamics quantum variational simulation. Background Art

[0002] Molecular dynamics simulation is one of the core issues in computational chemistry, and has its application scenarios in many fields such as chemistry, materials, and biology. For example, understanding the process of chemical reactions, designing and discovering new materials and drugs, and folding and binding of protein molecules all require a deep understanding of the behavior of molecules under different conditions. Traditional classical algorithms usually assume that the nuclei are stable and their positions are given, so they treat nuclear motion as parameters, thereby simplifying the problem to solving the motion state of electrons. In the calculation process, the information of dynamic coupling may be lost, resulting in failure to achieve the desired effect under high energy and high speed conditions.

[0003] Applying quantum computing to molecular dynamics simulations can more accurately characterize the structure of quantum systems, but current quantum simulation algorithms require quantum circuits of considerable depth. The increased complexity of quantum circuits also increases the demand for computing resources, posing a huge challenge to quantum computers. Summary of the invention

[0004] This application describes a molecular dynamics quantum variational simulation method and device that can solve the above-mentioned technical problems.

[0005] According to a first aspect, a molecular dynamics quantum variational simulation method is provided, the method comprising:

[0006] Sampling the results of measuring the initial quantum state of the target molecular system under the spatial basis vector to obtain a position basis vector set, wherein the position basis vector set is used to approximately describe the spatial distribution of the quantum state of the target molecular system;

[0007] Constructing a parameterized quantum circuit in a unitary format through the position basis vector set and the momentum basis vector of the initial quantum state of the target molecular system, wherein the parameterized quantum circuit includes parameters to be optimized and is used to simulate the evolution of the target molecular system in momentum and position space;

[0008] In each iteration time interval, the parameterized quantum circuit and the time-dependent Schrödinger equation are used to iteratively update the parameters to be optimized and the parameterized quantum circuit until the target evolution time is reached to obtain the final quantum state of the target molecular system.

[0009] Based on the above further embodiment, the parameterized quantum circuit includes a first quantum gate and a second quantum gate;

[0010] The first quantum gate uses a momentum basis vector to construct a rotation gate, and the rotation angle in the first quantum gate belongs to the parameter to be optimized;

[0011] The second quantum gate uses the position basis vector set to construct a rotation quantum gate, and the rotation angle in the second quantum gate belongs to the parameter to be optimized.

[0012] Based on the above further embodiment, the iterative updating of the parameters to be optimized and the parameterized quantum circuit using the parameterized quantum circuit and the time-dependent Schrödinger equation in each iterative time step specifically includes:

[0013] Before the start of each iterative time interval, using the parameterized quantum circuit to prepare the partial derivative state of the parameter to be optimized;

[0014] Obtaining the rate of change of the parameter to be optimized through the partial derivative state of the parameter to be optimized and the time-dependent Schrödinger equation;

[0015] The parameter to be optimized is iteratively updated using the change rate of the parameter to be optimized, and the parameterized quantum circuit is updated using the updated parameter to be optimized.

[0016] Based on the above further embodiment, obtaining the rate of change of the parameter to be optimized through the partial derivative state of the parameter to be optimized and the time-dependent Schrödinger equation specifically includes:

[0017] Using the parameterized quantum circuit to prepare a parameterized quantum state;

[0018] Substituting the parameterized quantum state into the time-dependent Schrödinger equation to obtain a linear equation, wherein the linear equation reflects the change of the parameter to be optimized over time, and the linear equation includes an evolution matrix and an evolution vector, wherein the evolution matrix is ​​composed of the partial derivative state of the parameter to be optimized, and the evolution vector is composed of the partial derivative state of the parameter to be optimized and a Hamiltonian;

[0019] Measuring the evolution matrix and the evolution vector using the partial derivative state of the parameter to be optimized;

[0020] Substituting the evolution matrix and the evolution vector into the linear equation, the change rate of the parameter to be optimized is obtained.

[0021] Based on the above further embodiment, the linear equation is expressed as , the evolution matrix , the evolution vector ;

[0022] R and I represent the real part and imaginary part respectively, H is the Hamiltonian, yes The partial derivative state of the parameter to be optimized, yes The partial derivative states of the parameters to be optimized.

[0023] Based on the above further embodiment, the parameterized quantum circuit , is the rotation angle in the first quantum gate, , is the position basis vector set, is the momentum basis vector, is the position basis vector, which belongs to the position basis vector set.

[0024] Based on the above further embodiment, the iterative updating of the parameter to be optimized by using the change rate of the parameter to be optimized specifically includes:

[0025] The gradient method is used to iteratively update the parameter to be optimized using the rate of change of the parameter to be optimized.

[0026] According to a second aspect, a molecular dynamics quantum variational simulation device is provided, the device comprising:

[0027] A first processing module is used to sample the results of measuring the initial quantum state of the target molecular system under the space basis vector to obtain a position basis vector set, wherein the position basis vector set is used to approximately describe the spatial distribution of the quantum state of the target molecular system;

[0028] A second processing module is used to construct a parameterized quantum circuit in a unitary format through a set of momentum basis vectors and position basis vectors of the initial quantum state of the target molecular system, wherein the parameterized quantum circuit includes parameters to be optimized and is used to simulate the evolution of the target molecular system in momentum and position space;

[0029] The third processing module is used to iteratively update the parameters to be optimized and the parameterized quantum circuit using the parameterized quantum circuit and the time-dependent Schrödinger equation in each iterative time step until the target evolution time is reached to obtain the final quantum state of the target molecular system.

[0030] Based on the above further embodiment, the parameterized quantum circuit includes a first quantum gate and a second quantum gate;

[0031] The first quantum gate uses a momentum basis vector to construct a rotation gate, and the rotation angle in the first quantum gate belongs to the parameter to be optimized;

[0032] The second quantum gate uses the position basis vector set to construct a rotation quantum gate, and the rotation angle in the second quantum gate belongs to the parameter to be optimized.

[0033] Based on the above further embodiment, the third processing module is specifically used to prepare a parameterized quantum state using the parameterized quantum circuit;

[0034] Substituting the parameterized quantum state into the time-dependent Schrödinger equation to obtain a linear equation, wherein the linear equation reflects the change of the parameter to be optimized over time, and the linear equation includes an evolution matrix and an evolution vector, wherein the evolution matrix is ​​composed of the partial derivative state of the parameter to be optimized, and the evolution vector is composed of the partial derivative state of the parameter to be optimized and a Hamiltonian;

[0035] Measuring the evolution matrix and the evolution vector using the partial derivative state of the parameter to be optimized;

[0036] Substituting the evolution matrix and the evolution vector into the linear equation, the change rate of the parameter to be optimized is obtained.

[0037] Based on the above further embodiment, the linear equation is expressed as , the evolution matrix , the evolution vector ;

[0038] R and I represent the real part and imaginary part respectively, H is the Hamiltonian, yes The partial derivative state of the parameter to be optimized, yes The partial derivative states of the parameters to be optimized.

[0039] Based on the above further embodiment, the parameterized quantum circuit , is the rotation angle in the first quantum gate, , is the position basis vector set, is the momentum basis vector, is the position basis vector, which belongs to the position basis vector set.

[0040] The gradient method is used to iteratively update the parameter to be optimized using the rate of change of the parameter to be optimized.

[0041] According to a third aspect, a computer storage medium is provided, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by one or more processors, the molecular dynamics quantum variational simulation method as described in any one of the above technical solutions is implemented.

[0042] According to a fourth aspect, an electronic device is provided, comprising a memory and one or more processors, wherein the memory stores a computer program, and when the computer program is executed by the one or more processors, the molecular dynamics quantum variational simulation method as described in any one of the above technical solutions is implemented.

[0043] In the above-mentioned system and method provided in the embodiments of this specification, the architecture of quantum computing variational method is adopted to solve the molecular dynamics problem, and the shallow quantum circuit can be used to simulate the evolution of the actual quantum system for a long time, which alleviates the demand for quantum device resources. By measuring the sampling method, the largest partial component of the atomic nucleus wave function under the basis vector is obtained to characterize the behavior of the atomic nucleus, which can further avoid the influence of unnecessary terms on the depth of the quantum circuit. The Hamiltonian and molecular structure are introduced into the parameterized quantum circuit design, taking into account the quantum effects between molecules, and the calculation accuracy is higher. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 A schematic diagram showing a flow chart of a molecular dynamics quantum variational simulation method provided in an embodiment of this specification; Figure 2 A schematic diagram of the quantum circuit structure provided in the embodiments of this specification is shown; Figure 3 A schematic diagram showing a flow chart of a molecular dynamics quantum variational simulation method provided in an embodiment of this specification; Figure 4 A schematic diagram showing a molecular dynamics quantum variational simulation device provided in an embodiment of this specification. DETAILED DESCRIPTION

[0046] The solution provided in this specification is described below in conjunction with the accompanying drawings.

[0047] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0048] In the description of the embodiments of the present application, words such as "exemplary", "for example" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary", "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "for example" or "for example" is intended to present related concepts in a concrete way.

[0049] In the description of the embodiments of the present application, the term "and / or" is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, B exists alone, and A and B exist at the same time. In addition, unless otherwise specified, the term "plurality" means two or more.

[0050] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprises", "has" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0051] In the classical molecular system dynamics simulation process, the nucleus is assumed to be stable, its position is considered to be given, the nuclear motion is regarded as a parameter, and the problem is simplified to solving the motion state of the electron. The specific process is as follows: first, an initial molecular configuration is selected, the Hamiltonian is determined, the nucleus in the Hamiltonian is regarded as stationary according to the Born-Oppenheimer approximation, the molecular Hamiltonian is separated from the Hamiltonian as a function of the nucleus and electron positions, the Schrödinger equation corresponding to the molecular Hamiltonian is solved, and the electron orbit is obtained as a function of the nucleus position as a parameter. The potential energy distribution of the nucleus is calculated using the obtained electron density distribution, the classical motion of the nucleus in the potential energy surface is calculated, the nucleus position is updated, and the above steps are repeated until the target evolution time is obtained. The dynamic simulation process of the molecular system is obtained. This method can be used to simulate the dynamic behavior of molecular systems within a certain period of time, but the Born-Oppenheimer approximation method is limited to high energy and high speed conditions. In addition, traditional methods usually consider the coupling between nuclei and electrons to be static, that is, the nucleus position is regarded as fixed during the calculation process, and the information of dynamic coupling may be lost during the calculation process. And because simulations on long time scales can lead to numerical instability or excessive computational costs, classical methods are usually limited to simulations on short time scales.

[0052] The current technical solution for using quantum computing to solve dynamic problems is the product formula method. Its main idea is to decompose the Hamiltonian of the multi-body problem into local Hamiltonians. The process includes decomposing the total Hamiltonian of the system into the sum of a series of commutative local Hamiltonians, and using the Trotter-Suzuki decomposition to convert the time evolution operator Approximately in the product form of local Hamiltonians, a quantum circuit is constructed based on the decomposed Hamiltonian. Each local Hamiltonian corresponds to a local quantum gate in the circuit. The optimized quantum circuit is run on a quantum computer to simulate the evolution of the system's quantum state over time until the target evolution time. However, this method usually requires very deep quantum circuits. Considering the first-order slicing method in the product formula method, the number of quantum gates required is proportional to the square of the evolution time. The increased complexity of quantum circuits also increases the demand for computing resources. Especially on current quantum computing devices, due to the presence of noise and a limited number of quantum bits, these additional gate operations may cause decoherence and error accumulation of the quantum state.

[0053] In view of this, the present invention proposes a molecular dynamics simulation method based on the variational method, which can solve the problem that the complexity of quantum circuits increases with the evolution time. Figure 1 The schematic diagram of the concept of the dynamics simulation method of the present invention is shown. As shown in the figure, the initial quantum state of the target molecular system is first set, and its wave function is generally gathered in space in the form of wave packets. The initial quantum state of the target molecular system can be measured multiple times under the spatial basis vector, and the position eigenvalues ​​obtained by the multiple measurements are statistically analyzed to select the position basis vector with a higher frequency of occurrence to form a position basis vector set. . Design of parameterized quantum circuits in unitary format ,like Figure 2 It is shown that the quantum gate including momentum part in the parameterized quantum circuit , is the momentum basis vector, is the rotation angle in the quantum gate, the position part of the quantum gate , is the position basis vector, . The rotation angles are taken as parameters to be optimized.

[0054] Then, at one iteration time interval, the evolution matrix is ​​measured through the parameterized quantum circuit and the evolution vector , and obtain the rate of change of the parameter to be optimized , use the gradient method to iteratively update the parameters to be optimized, use the updated parameters to be optimized to update the parameterized quantum circuit until the target evolution time is reached, and obtain the final quantum state of the target molecular system, and use the final quantum state of the target molecular system to obtain the target evolution state information.

[0055] The present invention is based on the variational process of classical approximation and wave packets, and can handle large-scale molecular simulation problems on shallow quantum circuits. At the same time, in the design stage of parameterized quantum circuits, the phase in the Hamiltonian evolution term is taken into consideration, and a parameterized unitary transformation is designed to ensure that the circuit depth is only related to the system Hamiltonian and does not increase with the simulation time. Since the depth of the quantum circuit is independent of the evolution time, it can be used to simulate the structural evolution and phase change of materials under long time scales. In addition, the atomic nucleus wave function is approximately regarded as a localized wave packet, and the most likely position of the atomic nucleus under the spatial basis is obtained by measuring and sampling, thereby avoiding the construction of quantum circuits under all spatial basis vectors, and further reducing the number of layers of quantum circuits.

[0056] Combine the following Figure 3 The schematic diagram of the molecular dynamics quantum variational simulation method is described in detail, including:

[0057] 110. The results of measuring the initial quantum state of the target molecular system under the space basis vector are sampled to obtain a set of position basis vectors. The set of position basis vectors is used to approximately describe the spatial distribution of the quantum state of the target molecular system.

[0058] Before measuring the target molecular system, the initial quantum state of the target molecular system is first set, and the wave function is generally gathered in space as a wave packet. At the same time, the target evolution time T, the time interval Δt, and the total number of evolution steps n=T / Δt are set.

[0059] A set of spatial basis vectors are selected, usually these basis vectors are position eigenstates, and the initial quantum state is measured multiple times under these spatial basis vectors. Each measurement may obtain a specific position eigenvalue, that is, a specific spatial position. The results of multiple measurements are collected and statistically analyzed. The measurement results will give the probability distribution of finding the system under each position basis vector. According to the measurement results, those position basis vectors with higher frequency of occurrence are selected to form a position basis set, which can be used to approximately describe the spatial distribution of the initial quantum state of the molecular system.

[0060] Specifically, in molecular dynamics simulation problems, due to the complex interactions between the nucleus and the electrons, the Born-Oppenheimer approximation (BO approximation) is usually used to separate the electron Hamiltonian and solve the electron density distribution by considering the nucleus as stationary, which in turn constitutes the potential energy surface of the nucleus' motion. In this embodiment, by repeatedly measuring the initial state of the nucleus under the space basis vectors for many times, some sets of space basis vectors with larger components can be obtained: , which is used to approximate the behavior of the original nuclear wave function. The potential energy of the position where the nucleus appears with high probability It will largely determine how the state of the nucleus evolves.

[0061] 120. A parameterized quantum circuit is constructed by using a set of position basis vectors and a momentum basis vector of the initial quantum state of the target molecular system. The parameterized quantum circuit includes parameters to be optimized and is used to simulate the evolution of the target molecular system in momentum and position space.

[0062] Specifically, the Hamiltonian evolves middle, The momentum term can be expressed in the momentum basis through quantum Fourier transform, and the potential energy term can be approximately restricted to the set of measurement results , reducing the potential energy term that needs to be realized.

[0063] The parameterized quantum circuit is designed with reference to the unitary operator of the Hamiltonian evolution, such as , which ensures that the circuit depth is only related to the system Hamiltonian and does not increase with the simulation time.

[0064] Specifically, in this embodiment, the parameterized quantum circuit includes a first quantum gate and a second quantum gate. The first quantum gate uses a momentum basis vector to construct a rotation gate, and the rotation angle in the first quantum gate belongs to the parameter to be optimized. The second quantum gate uses a position basis vector set to construct a rotation quantum gate, and the rotation angle in the second quantum gate belongs to the parameter to be optimized.

[0065] More specifically, if Figure 2 As shown, the parameterized quantum circuit , is the rotation angle in the first quantum gate, , is the set of position basis vectors, is the momentum basis vector, is the position basis vector, which belongs to the position basis vector set.

[0066] 130. In each iteration time interval, the parameterized quantum circuit and the time-dependent Schrödinger equation are used to iteratively update the parameters to be optimized and the parameterized quantum circuit until the target evolution time is reached to obtain the final quantum state of the target molecular system.

[0067] Step 130 specifically includes:

[0068] 131. Before each iteration time interval begins, use the parameterized quantum circuit to prepare the partial derivative state of the parameter to be optimized. .

[0069] 132. The rate of change of the parameter to be optimized is obtained through the partial derivative state of the parameter to be optimized and the time-dependent Schrödinger equation.

[0070] Step 132 specifically includes:

[0071] 1321. Using parameterized quantum circuits to prepare parameterized quantum states.

[0072] 1322. Substitute the parameterized quantum state into the time-dependent Schrödinger equation to obtain a linear equation, which reflects that the parameter to be optimized changes with time. The linear equation includes an evolution matrix and an evolution vector. The evolution matrix is ​​composed of the partial derivative state of the parameter to be optimized, and the evolution vector is composed of the partial derivative state of the parameter to be optimized and the Hamiltonian.

[0073] 1323. Using the partial derivative state of the parameter to be optimized to measure the evolution matrix and evolution vector;

[0074] 1324. Substitute the evolution matrix and evolution vector into the linear equation to obtain the rate of change of the parameter to be optimized.

[0075] Specifically, in this embodiment, the evolution is represented by parameterizing the quantum circuit. The quantum state of time , is a parameterized quantum circuit, is the initial state, calculated by the Schrödinger equation Satisfy the equations and design parameterized quantum circuit structures.

[0076] According to the time-dependent Schrödinger equation, substituting the parameterized quantum state into the equation, we can get the linear equation whose parameters change with time: ,in ,

[0077] , and Respectively represent the real part and imaginary part, is the Hamiltonian, yes The partial derivative state of the parameter to be optimized, yes The partial derivative state of the parameter to be optimized. The partial derivative of the parameterized quantum state composed of a single-bit rotation gate is expressed as , which can be measured on a quantum circuit. W(θ) is a transformation matrix related to the derivative of a single-bit rotation gate. For a common rotation gate, it can be expressed as the evolution of the gate's generator at a certain angle. The process of obtaining the derivative can also be obtained by applying a quantum gate on a quantum circuit and measuring it.

[0078] 133. The parameters to be optimized are iteratively updated using the rate of change of the parameters to be optimized, and the parameterized quantum circuit is updated using the updated parameters to be optimized.

[0079] In this embodiment, the gradient method can be used to iteratively update the parameters to be optimized using the change rate of the parameters to be optimized. For example, the formula , update the parameters to be optimized .

[0080] Through the above steps, we can finally After determining the quantum state of the molecular system at a certain moment, the observable to be measured can be measured to obtain relevant information about the target quantum state of the molecular system.

[0081] This embodiment uses the architecture of quantum computing variational method to solve molecular dynamics problems. It can use shallow quantum circuits to simulate the evolution of actual quantum systems over a long period of time, alleviating the demand for quantum device resources. By measuring the sampling method, the largest partial component of the atomic nucleus wave function under the basis vector is obtained to characterize the behavior of the atomic nucleus, which can further avoid the influence of unnecessary terms on the depth of the quantum circuit. The Hamiltonian and molecular structure are introduced into the parameterized quantum circuit design, taking into account the quantum effects between molecules, and the calculation accuracy is higher.

[0082] like Figure 4 The molecular dynamics quantum variation simulation device shown, a molecular dynamics quantum variation simulation device comprises:

[0083] A first processing module is used to obtain a position basis vector set according to the result obtained by measuring the initial quantum state of the target molecular system under the space basis vector, wherein the position basis vector set is used to approximately describe the spatial distribution of the quantum state of the target molecular system;

[0084] A second processing module is used to construct a parameterized quantum circuit through a momentum basis vector and a position basis vector set of the initial quantum state of the target molecular system, wherein the parameterized quantum circuit includes parameters to be optimized and is used to simulate the evolution of the target molecular system in momentum and position space;

[0085] The third processing module is used to iteratively update the parameters to be optimized and the parameterized quantum circuit using the parameterized quantum circuit and the time-dependent Schrödinger equation in each iterative time step until the target evolution time is reached to obtain the final quantum state of the target molecular system.

[0086] Based on the above further embodiment, the parameterized quantum circuit includes a first quantum gate and a second quantum gate;

[0087] The first quantum gate uses a momentum basis vector to construct a rotation gate, and the rotation angle in the first quantum gate belongs to the parameter to be optimized;

[0088] The second quantum gate uses the position basis vector set to construct a rotation quantum gate, and the rotation angle in the second quantum gate belongs to the parameter to be optimized.

[0089] Based on the above further embodiment, the third processing module is specifically used to prepare a parameterized quantum state using the parameterized quantum circuit;

[0090] Substituting the parameterized quantum state into the time-dependent Schrödinger equation to obtain a linear equation, wherein the linear equation reflects the change of the parameter to be optimized over time, and the linear equation includes an evolution matrix and an evolution vector, wherein the evolution matrix is ​​composed of the partial derivative state of the parameter to be optimized, and the evolution vector is composed of the partial derivative state of the parameter to be optimized and a Hamiltonian;

[0091] Measuring the evolution matrix and the evolution vector using the partial derivative state of the parameter to be optimized;

[0092] Substituting the evolution matrix and the evolution vector into the linear equation, the change rate of the parameter to be optimized is obtained.

[0093] Based on the above further embodiment, the linear equation is expressed as , the evolution matrix , the evolution vector ;

[0094] R and I represent the real part and imaginary part respectively, H is the Hamiltonian, yes The partial derivative state of the parameter to be optimized, yes The partial derivative states of the parameters to be optimized.

[0095] Based on the above further embodiment, the parameterized quantum circuit , is the rotation angle in the first quantum gate, , is the position basis vector set, is the momentum basis vector, is the position basis vector, which belongs to the position basis vector set.

[0096] The gradient method is used to iteratively update the parameter to be optimized using the rate of change of the parameter to be optimized.

[0097] The present invention also provides a computer storage medium, on which a computer program is stored. When the computer program is executed by one or more processors, the molecular dynamics quantum variational simulation method as described in any one of the above technical solutions is implemented.

[0098] The present invention also provides an electronic device, comprising a memory and one or more processors, wherein the memory stores a computer program, and when the computer program is executed by the one or more processors, the molecular dynamics quantum variational simulation method as described in any one of the above technical solutions is implemented.

[0099] The specific implementation methods described above further illustrate the purpose, technical solutions and beneficial effects of the present application in detail. It should be understood that the above description is only the specific implementation methods of the present application and is not intended to limit the scope of protection of the present application. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solutions of the present application should be included in the scope of protection of the present application.

Claims

1. A quantum variational simulation method for molecular dynamics, characterized in that: The method comprises: Sampling the results of measuring the initial quantum state of the target molecular system under the spatial basis vector to obtain a position basis vector set, wherein the position basis vector set is used to approximately describe the spatial distribution of the quantum state of the target molecular system; Constructing a parameterized quantum circuit in a unitary format through the position basis vector set and the momentum basis vector of the initial quantum state of the target molecular system, wherein the parameterized quantum circuit includes parameters to be optimized and is used to simulate the evolution of the target molecular system in momentum and position space; In each iteration time interval, the parameterized quantum circuit and the time-dependent Schrödinger equation are used to iteratively update the parameters to be optimized and the parameterized quantum circuit until the target evolution time is reached to obtain the final quantum state of the target molecular system.

2. The method according to claim 1, characterized in that The parameterized quantum circuit includes a first quantum gate and a second quantum gate; The first quantum gate uses a momentum basis vector to construct a rotation gate, and the rotation angle in the first quantum gate belongs to the parameter to be optimized; The second quantum gate uses the position basis vector set to construct a rotation gate, and the rotation angle in the second quantum gate belongs to the parameter to be optimized.

3. The method according to claim 1, characterized in that The method of iteratively updating the parameter to be optimized and the parameterized quantum circuit by using the parameterized quantum circuit and the time-dependent Schrödinger equation in each iterative time step specifically includes: Before the start of each iterative time interval, using the parameterized quantum circuit to prepare the partial derivative state of the parameter to be optimized; Obtaining the rate of change of the parameter to be optimized through the partial derivative state of the parameter to be optimized and the time-dependent Schrödinger equation; The parameter to be optimized is iteratively updated using the change rate of the parameter to be optimized, and the parameterized quantum circuit is updated using the updated parameter to be optimized.

4. The method according to claim 3, characterized in that The method of obtaining the rate of change of the parameter to be optimized through the partial derivative state of the parameter to be optimized and the time-dependent Schrödinger equation specifically includes: Using the parameterized quantum circuit to prepare a parameterized quantum state; Substituting the parameterized quantum state into the time-dependent Schrödinger equation to obtain a linear equation, wherein the linear equation reflects the change of the parameter to be optimized over time, and the linear equation includes an evolution matrix and an evolution vector, wherein the evolution matrix is ​​composed of the partial derivative state of the parameter to be optimized, and the evolution vector is composed of the partial derivative state of the parameter to be optimized and a Hamiltonian; Measuring the evolution matrix and the evolution vector using the partial derivative state of the parameter to be optimized; Substituting the evolution matrix and the evolution vector into the linear equation, the change rate of the parameter to be optimized is obtained.

5. The method according to claim 4, characterized in that The linear equation is expressed as , the evolution matrix , the evolution vector ; R and I represent the real part and imaginary part respectively, H is the Hamiltonian, yes The partial derivative state of the parameter to be optimized, yes The partial derivative states of the parameters to be optimized.

6. The method according to claim 2, characterized in that The parameterized quantum circuit , is the rotation angle in the first quantum gate, , is the position basis vector set, is the momentum basis vector, is the position basis vector, which belongs to the position basis vector set.

7. The method according to claim 2, characterized in that The iterative updating of the parameter to be optimized by using the change rate of the parameter to be optimized specifically includes: The gradient method is used to iteratively update the parameter to be optimized using the rate of change of the parameter to be optimized.

8. A quantum variational simulation device for molecular dynamics, characterized in that: The device comprises: A first processing module is used to sample the results of measuring the initial quantum state of the target molecular system under the space basis vector to obtain a position basis vector set, wherein the position basis vector set is used to approximately describe the spatial distribution of the quantum state of the target molecular system; A second processing module is used to construct a parameterized quantum circuit in a unitary format through the position basis vector set and the momentum basis vector of the initial quantum state of the target molecular system, wherein the parameterized quantum circuit includes parameters to be optimized and is used to simulate the evolution of the target molecular system in momentum and position space; The third processing module is used to iteratively update the parameters to be optimized and the parameterized quantum circuit using the parameterized quantum circuit and the time-dependent Schrödinger equation in each iterative time step until the target evolution time is reached to obtain the final quantum state of the target molecular system.

9. A computer storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by one or more processors, the molecular dynamics quantum variational simulation method according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: It comprises a memory and one or more processors, wherein the memory stores a computer program, and when the computer program is executed by the one or more processors, the molecular dynamics quantum variation simulation method as claimed in any one of claims 1 to 7 is implemented.

Images