Method for determining a configuration of a quantum processor for solving an optimization problem - Patents.com

By optimizing the positioning of neutral atoms in quantum processors using machine learning, the method addresses the complexity of QUBO problem-solving, enhancing performance and solution quality for QUBO problems.

JP2024541667AInactive Publication Date: 2024-11-08PASQAL SAS
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2024533061
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2021-12-03
Filing Date
2022-12-02
Publication Date
2024-11-08
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Existing quantum processors based on neutral atoms face limitations in solving Quadratic Unconstrained Binary Optimization (QUBO) problems due to the complexity of generating interaction Hamiltonians that scale linearly with the number of atoms, making them less effective compared to superconducting circuits.

Method used

A method for determining the configuration of a quantum processor using neutral atoms by minimizing the difference between the interaction and target Hamiltonians through machine learning tools, such as manifold learning, to optimize the positioning of neutral atoms for improved performance in solving QUBO problems.

Benefits of technology

The method enhances the performance of quantum processors using neutral atoms by aligning the interaction Hamiltonian closer to the target Hamiltonian, resulting in better solutions for QUBO problems, especially when the QUBO matrix elements vary significantly.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024541667000001_ABST
    Figure 2024541667000001_ABST
Patent Text Reader

Abstract

The present invention relates to a method for determining a configuration of a quantum processor for solving an optimization problem, the quantum processor including neutral atoms that can be manipulated to form qubits, the quantum processor having an interaction Hamiltonian that depends on the positions of the neutral atoms, the method comprising: a. determining a target Hamiltonian for a quantum processor according to a QUBO matrix, the QUBO matrix describing an optimization problem in the form of a quadratic unconstrained binary optimization; b. determining positions of neutral atoms of the quantum processor to minimize a difference between an interaction Hamiltonian of the quantum processor and a target Hamiltonian, the determining positions including using a machine learning tool capable of determining positions of atoms according to data from the target Hamiltonian; Includes.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a method for determining a configuration of a quantum processor for solving an optimization problem. The present invention further relates to a computer program product. The present invention further relates to an associated readable storage medium. [Background technology]

[0002] Quantum computing is an interesting and fast-growing technology, especially for solving combinatorial optimization problems. Such problems arise in all industrial sectors (logistics, energy, telecommunications, IT, banking / finance, etc.) and are generally not amenable to the computational capabilities of classical computers.

[0003] Combinatorial optimization problems are amenable to being reformulated in the form of a Quadratic Unconstrained Binary Optimization (QUBO) problem. A QUBO problem is a problem in which the cost function C(x)=x T The string of bits that minimizes Mx

number

[0004] Methods for solving QUBO problems using quantum processors are known from the prior art. Such methods use quantum adiabatic algorithms, for example, as described in the paper "A Quantum Adiabatic Evolution Algorithm Applied to Random Instances of an NP-Complete Problem" by Edward Farhi, Jeffrey Goldstone, Sam Gutmann, Joshua Lapan, Andrew Lundgren, and Daniel Preda, Science, 292(5516), pp. 472-476, April 2001, doi: 10.1126 / science.1057726. Other methods are based on the Quantum Approximate Optimization Algorithm (QAOA) as described in the article by Edward Farhi, Jeffrey Goldstone, and Sam Gutmann, "A Quantum Approximate Optimization Algorithm", arXiv e-prints, art. arXiv:1411.4028, November 2014.

[0005] Implementing such a solving method involves constructing a Hamiltonian for the quantum processor that is closely related to the cost function C. More specifically, to implement such a method, a target Hamiltonian H for the quantum processor is constructed. c of, <x|H c We try to construct the quantum state as |x>=C(x).

number

[0006] The target Hamiltonian H c Means to accurately generate have been developed for quantum processors based on superconducting quantum circuits or processors, such as the D-wave machines.

[0007] However, such measures are not available for other quantum processor technologies, in particular quantum processors based on neutral atoms with analogue operations. Indeed, the number of degrees of freedom possible with an arrangement of N atoms in a plane or in space scales linearly with N, and therefore the complexity of satisfying the constraints imposed during the generation of a matrix of size 2N becomes rapidly quadratic.

[0008] Thus, the performance of quantum processors based on neutral atoms is limited in solving QUBO problems. However, by using neutral atoms exactly, such processors have many advantages over processors based on artificial atoms such as superconducting circuits or silicon spin qubits. [Prior art documents] [Non-patent literature]

[0009] [Non-Patent Document 1] Edward Farhi, Jeffrey Goldstone, Sam Gutmann, Joshua Lapan, Andrew Lundgren, Daniel Preda, "A Quantum Adiabatic Evolution Algorithm Applied to Random Instances of an NP-Complete Problem", Science, 292(5516), pp. 472-476, April 2001, doi: 10.1126 / science.1057726 [Non-Patent Document 2] Edward Farhi, Jeffrey Goldstone, Sam Gutmann, “A Quantum Approximate Optimization Algorithm”, arXiv e-prints, art. arXiv:1411.4028, November 2014 [Non-Patent Document 3] Loic Henriet, Lucas Beguin, Adrien Signoles, Thierry Lahaye, Antoine Browaeys, Georges-Olivier Reymond, Christophe Jurczak, "Quantum computing with neutral atoms." Quantum, 4:327, September 2020, 2020. ISSN 2521-327X. Paragraph II of doi:10.22331 / q-2020-09-21-327 Summary of the Invention [Problem to be solved by the invention]

[0010] Therefore, there is a need for methods to improve the performance of quantum processors based on neutral atoms to solve QUBO problems. [Means for solving the problem]

[0011] To this end, the subject of the present specification is a method for determining a configuration of a quantum processor for solving an optimization problem, the quantum processor comprising neutral atoms suitable for being manipulated to form qubits, the quantum processor having an interaction Hamiltonian describing an interaction between the qubits, the interaction Hamiltonian being a function of the positioning of the neutral atoms, the method comprising the following steps implemented by a computer: a. determining a target Hamiltonian for a quantum processor according to a QUBO matrix, the QUBO matrix describing an optimization problem in the form of a quadratic unconstrained binary optimization; b. determining positions of neutral atoms of the quantum processor to minimize a difference between an interaction Hamiltonian of the quantum processor and a target Hamiltonian, the determining positions including the use of a machine learning tool capable of determining positions of the atoms according to data from the target Hamiltonian, the determined positions defining a processor configuration suitable for use in solving the optimization problem; Includes.

[0012] The present invention further relates to a method for solving an optimization problem by a quantum processor, said method of solving comprising the steps of the method for determining a configuration of a quantum processor as described above, wherein the method of solving comprises the step of positioning neutral atoms of the quantum processor according to the positions determined in order to solve the optimization problem.

[0013] According to a particular embodiment, the solving method further comprises a step of solving the optimization problem using a quantum approximate optimization algorithm called QAOA, or a quantum adiabatic algorithm, using neutral atoms of the quantum processor positioned according to the determined positions.

[0014] According to other particular embodiments, the determination and / or solving method further comprises one or more of the following features, taken individually or according to all technically possible combinations: The learning tool is a manifold machine learning tool suitable for performing a non-linear reduction in the dimensionality of the data at the input of the tool. - Multiple machine learning tools based on multi-dimensional positioning algorithms. - The QUBO matrix is ​​constructed in the form of a symmetrical real matrix, and the step of determining the position is Determining a matrix, called the distance matrix, as a function of the QUBO matrix, where each coefficient at position i, j of the distance matrix corresponds to a distance between qubit i and qubit j; and Learning tools to determine the positions of neutral atoms according to the distance matrix Including, the step of determining the position comprises the determination of an intermediate matrix as a function of the QUBO matrix, the intermediate matrix having coefficients equal to zero on its diagonal and coefficients otherwise the same as the QUBO matrix, and the distance matrix is ​​obtained as a function of the intermediate matrix; The coefficients of the distance matrix are obtained according to the following formula:

number

[0015] The invention further relates to a computer program product having stored thereon a computer program comprising program instructions, which, when loaded into a data processing unit, results in the implementation of the steps of determining the target Hamiltonian and of determining the position of the method as defined above, when the computer program is implemented in the data processing unit.

[0016] The present specification further relates to a readable storage medium having stored thereon a computer program comprising program instructions, which is loaded into a data processing unit and, when the computer program is implemented in the data processing unit, results in the implementation of the steps of determining the target Hamiltonian and of determining the position of the method as defined above.

[0017] Other characteristics and advantages of the invention will become apparent on reading the following description of an embodiment of the invention, given purely by way of example, and on reference to the following drawings, in which: [Brief description of the drawings]

[0018] [Figure 1] 1 is a flowchart of an example of an implementation of a method for determining a configuration of a quantum processor for solving an optimization problem. [Diagram 2] FIG. 2 illustrates a specific example of an implementation of the method shown in FIG. 1. [Diagram 3] 1 is an example of a graph comparing the positions of neutral atoms determined by the method of the present invention with their positions on a triangular array for a first family of QUBO matrices, which groups QUBO matrices that all have equal coefficients. [Figure 4] FIG. 4 is an example of a graph similar to that shown in FIG. 3 for the second family of QUBO matrices, grouping together QUBO matrices that have coefficients with small differences between them. [Diagram 5] FIG. 5 is an example of a graph similar to that shown in FIGS. 3 and 4, plotted for the third family of QUBO matrices, with QUBO matrices that have coefficients with large differences between them grouped together. [Figure 6] 1 is an example of a graph showing the performance of a quantum processor for solving the QUBO problem from the QAOA algorithm when the neutral atoms of the processor are positioned according to positions determined by the method of the present invention, and when the positions are on a triangular array. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0019] The decision method described below in the description is implemented for a quantum processor, also called a quantum computer.

[0020] A quantum processor is a network of qubits, as well as hardware for manipulating the qubits. A quantum processor is suitable for performing quantum operations on the qubits.

[0021] Quantum processors use the quantum properties of matter, such as superposition and entanglement, to perform operations on data. Unlike conventional computers based on transistors, which operate on binary data (encoded as bits, i.e., 0 or 1), quantum processors operate on qubits, whose quantum states can take on a continuum of values, rather than being discrete.

[0022] More specifically, a qubit refers to a two-level quantum mechanical system. For example, a qubit includes two fundamental quantum states I0> and I1>, which represent the possible quantum states of the qubit. According to the principle of superposition in quantum mechanics, any superposition of the form aI0> + bI1> (where a and b are complex numbers and aa* + bb* = 1) is a possible quantum state of the qubit.

[0023] A quantum processor envisaged in the implementation of the method comprises neutral atoms adapted to be manipulated to form qubits, the manipulation being carried out by a light beam such as a laser.

[0024] A neutral atom is an atom that is electrically neutral. Such a neutral atom is excited to the Rydberg level, for example, during use in a quantum processor. A neutral atom is, for example, a rubidium atom.

[0025] A quantum processor has an interaction Hamiltonian H that describes the behavior of the qubits, specifically their interactions. In the case of a quantum processor using neutral atoms, the interaction Hamiltonian H is a function of the positions of the neutral atoms. For example, the interaction Hamiltonian corresponds to the Ising model. In such a case, the interaction Hamiltonian H is given, for example, by the following equation:

number

number

[0026] Such quantum processors using neutral atoms are, for example, those described in paragraph II of the article by Loic Henriet, Lucas Beguin, Adrien Signoles, Thierry Lahaye, Antoine Browaeys, Georges-Olivier Reymond, and Christophe Jurczak, Quantum computing with neutral atoms. Quantum, 4:327, September 2020, 2020. ISSN 2521-327X. doi:10.22331 / q-2020-09-21-327.

[0027] Preferentially, at least some steps of the method for determining the configuration of a quantum processor are performed by another computer (classical or quantum), i.e. by a computing machine interacting with a computer program product.

[0028] In such a case, the computing device comprises, for example, a processor including a data processing unit, a memory, a data storage medium reader and possibly a human machine interface such as a screen and a display.

[0029] The computer program product includes a data storage medium. The data storage medium is a medium that is readable by a computer, typically by a data processing unit. The readable storage medium is a medium suitable for storing electronic instructions and is easily coupled to a bus of a computer system. A computer program including program instructions is stored in the data storage medium.

[0030] The computer program can be loaded into a data processing unit and is suitable for causing the implementation of several steps of the method for determining the configuration of a quantum processor when the computer program is implemented in the processing unit of the computing device.

[0031] An example of a method for determining a configuration of a quantum processor for solving an optimization problem will now be described with reference to FIG. 1 and the flow chart shown in FIGS. 2 to 6 illustrating some steps of the method.

[0032] The determination method comprises a step 100 of obtaining a matrix, called the QUBO M-matrix, which describes the optimization problem in the form of quadratic unconstrained binary optimization. The optimization problem is, for example, a combinatorial optimization problem.

[0033] The obtaining step 100 is, for example, performed by a computer interacting with a computer program product, i.e., is computer-implemented. The matrix QUBO is, for example, communicated to the computer by an operator.

[0034] As explained above, the QUBO problem is a cost function C(x)=x where M is the QUBO matrix. T The string of bits that minimizes Mx

[0035]

number

[0036] , i.e., to find a vector with N dimensions of 0 and 1. The QUBO M matrix is ​​a symmetric real matrix.

[0037] The QUBO M matrix is ​​specific to the optimization problem under consideration.

[0038] The determination method is to determine the target Hamiltonian H for the quantum processor as a function of the QUBO matrix M. c Thus, the target Hamiltonian H c is also specific to the optimization problem in question.

[0039] The determining step 110 is, for example, performed by a computing device interacting with a computer program product, ie, is computer-implemented.

[0040] Target Hamiltonian H for quantum processors c teeth, <x|H c |x>=C(x), i.e. <x|H c |x>=x T It will look like Mx.

[0041] Using the previous notation for the occupation operator, we define the target Hamiltonian H c can also be expressed in the following form:

number

[0042] The quantum processor interaction Hamiltonian H has the form

number

[0043] The determination method is based on the interaction Hamiltonian H of the quantum processor and the target Hamiltonian H c 120. Determining 120 the positions of the neutral atoms of the quantum processor so as to minimize the difference between M i,j Closest to U i,j The positions of the atoms are determined in two-dimensional space (a plane) or in three-dimensional space.

[0044] The determining step 120 may be performed, for example, by a computing device interacting with a computer program product, ie, may be computer-implemented.

[0045] The determined locations define a configuration of quantum processors suitable for use in solving the optimization problem.

[0046] The step of determining the position is carried out by computing the target Hamiltonian H c The method includes using a machine learning tool suitable for determining the positions of atoms according to data generated from

[0047] The learning tool is preferentially a manifold machine learning tool suitable for performing a non-linear reduction of the dimensionality of the data at the input of the tool. Such a tool implements an algorithm based on unsupervised estimators to describe a dataset as a collector of small dimensions embedded in a high-dimensional space.

[0048] Preferentially, the learning tool is based on a multidimensional scaling algorithm. Such an algorithm serves to transform a proximity matrix (similarity or dissimilarity) between a set of N objects into the coordinates of those objects in a P-dimensional space. When P is set to 2 or 3, it is easy to observe the objects.

[0049] In a variant, the learning tool is based on another algorithm, for example an algorithm chosen from the following list: Isomap, Locally Linear Embedding, Modified Locally Linear Embedding, Hessian Eigenmapping, Spectral Embedding, Local Tangent Space Alignment, T-distributed Stochastic Neighbor Embedding.

[0050] An example of an implementation of step 120 for determining the position is shown in FIG. 2 (left column). In the above example, the QUBO M matrix has positive coefficients outside its diagonal. Preferentially, the QUBO M matrix is ​​formulated as a symmetric real matrix with positive terms. Nevertheless, the above example generalizes to any symmetric QUBO M matrix.

[0051] In such an example of implementation, the determining step 120 determines the intermediate matrix U as a function of the QUBO matrix M (a symmetric real matrix). 0 This includes the determination of:

[0052] intermediate matrix U 0 has coefficients equal to zero on its diagonal and otherwise has the same coefficients as matrix QUBO M. 0 is therefore of the form

number

number

[0053] In the above example, the decision step 120 determines whether the intermediate matrix U 0 As a function of 0 For example, the distance matrix D 0 The coefficient of [D 0 ] i,j is obtained according to the following formula:

number

[0054] Therefore, the distance matrix D 0 Each coefficient at position i, j [D 0 ] i,j corresponds to the distance between qubit i and qubit j.

[0055] In the above example, the decision step 120 is to determine the distance matrix D 0 It involves determining the positions of neutral atoms in space as a function of . More precisely, the learning tool returns at its output the coordinates of each atom in the considered space.

[0056] Preferentially, the learning tool includes a parameter related to the dimension of the space and to the number of iterations performed by the learning tool to obtain the final positions of the neutral atoms. The determined positions of the atoms are thus the positions obtained at the end of the last iteration. In this case, the space is two or three dimensional.

[0057] Once the positions are determined, the corresponding interaction Hamiltonian H can be calculated using the formula explained above, i.e.

number

[0058] At the end of the method, U is a priori as close to U as possible. 0 , and the interaction Hamiltonian H and the target Hamiltonian H c The difference between is minimal.

[0059] In some cases, the determination method includes a step 130 of effectively positioning neutral atoms of a quantum processor according to the determined positions to solve the optimization problem.

[0060] The positioning step 130 is performed on the quantum processor in question, for example using a tool for manipulating neutral atoms. The tool is, for example, optical tweezers. In FIG. 2, in particular, the elements in the dotted box (left column) correspond to the implementation on the quantum processor considered. The quantum processor is then used to solve the QUBO problem. The solving method uses, for example, a quantum adiabatic algorithm or a quantum approximate optimization algorithm.

[0061] Thus, the method serves to determine a configuration of a quantum processor that uses neutral atoms to solve an optimization problem expressed in the QUBO format. The use of the learning tool is a step of determining the interaction Hamiltonian H of the quantum processor and the target Hamiltonian H c This makes it possible to determine the positions of the neutral atoms in the quantum processor such that the difference between θ and θ is minimal. The determined positions serve in particular to ensure that the states of the atoms useful for solving the QUBO problem are possible states.

[0062] As a result, the performance of such a quantum processor is improved while solving QUBO problems when the quantum processor is operating in analog mode.

[0063] Below we give examples of experiments performed to demonstrate the improvement in the performance of a quantum processor induced by the present method while solving the QUBO problem.

[0064] More specifically, Figs. 3 to 5 show the calculation of the target Hamiltonian H for a given QUBO M-matrix, for neutral atoms arranged in two different configurations, namely the positioning determined by the present method (curve C1 in the figure) and the so-called triangular positioning, where the atoms are placed on a triangular lattice (curve C2 in the figure). c The interaction part U 0 We aim to compare with the interaction part U of the interaction Hamiltonian H.

[0065] For size N ∈ {2,…,14}, 20 instances of each matrix family are approximated with the two aforementioned positioning configurations.

number

number

number

[0066] In triangular positioning, N atoms are 0 The positions determined by this method were obtained after 100 iterations by a machine learning tool.

[0067] To compare the relative performance of the two types of positioning,

number

[0068] In Figure 3, a matrix whose elements are all equal is considered. Such a matrix is ​​well represented by a triangular configuration, since it maximizes the number of pairs of atoms that can be considered nearest neighbors, and therefore the number of distances equal to the interval. In such cases, the positioning determined by the method does not produce a better solution when the method is limited to a fixed number of iterations.

[0069] In Fig. 4 we consider a matrix whose elements follow a uniform distribution between 0 and 1. The positioning determined by this method is better than the triangular positioning for a small number of atoms, but as N increases, the difference tends to decrease.

[0070] In FIG. 5, the matrix has a larger difference between its elements, i.e.

number

number

[0071] It is thereby clear from the example that the method makes it possible to provide a positioning configuration for neutral atoms that leads to a Hamiltonian approximation that is more relevant and therefore superior to the triangular positioning when the elements of the QUBO matrix differ by several orders of magnitude.

[0072] FIG. 6 shows a comparative example of the performance of a quantum processor for solving the QUBO problem using the QAOA algorithm when the neutral atoms of the processor are positioned according to positions determined by the method of the present invention and during triangular positions.

[0073] The complexity of the QAOA algorithm is given by the number of p layers it is applied to. By increasing p, the optimization efficiency will also increase. We consider here 5 QUBOs of size 10 to optimize. The QAOA algorithm is applied to each of the QUBOs, increasing the number of layers and with two different positioning configurations. In the first case, triangular positioning of the atoms is applied for the 5 QUBOs, excluding the rescaling factor. In the second case, the method is applied to each of the QUBOs and the atom positioning used is therefore specific to the QUBO being solved. The regularized cost function C(x) / Cmin(C(x) is negative,

number

[0074] FIG. 6 shows the distribution of scores achieved for five QUBOs with increasing values ​​of the number of layers. In FIG. 6, the triangular configuration corresponds to curve C4, and the intrinsic configuration determined by the method corresponds to curve C3. For both configurations, the average score increases with p as expected, but at a given p, QAOA achieves a better score using the configuration determined by the method. For example, for p=5, the average score for five QUBOs is about 0.71 for triangular positioning, but rises to 0.82 for intrinsic positioning. The method thus improves the performance of quantum processors when solving QUBO problems.

[0075] Those skilled in the art will appreciate that the above-described embodiments and variations may be combined to form new embodiments, where they are technically compatible.

Claims

1. 1. A method for determining a configuration of a quantum processor for solving an optimization problem, the quantum processor comprising neutral atoms adapted to be manipulated to form qubits, the quantum processor having an interaction Hamiltonian (H) describing interactions between qubits, the interaction Hamiltonian (H) being a function of positioning of the neutral atoms, the method comprising the following computer-implemented steps: a. A target Hamiltonian (H c ), the QUBO(M) matrix describing the optimization problem in the form of a quadratic unconstrained binary optimization. c ) b. The interaction Hamiltonian (H) of the quantum processor and the target Hamiltonian (H c ) to determine a position of the neutral atom in the quantum processor so as to minimize a difference between the target Hamiltonian (H c determining positions of the neutral atoms, the determined positions defining a processor configuration suitable for use in solving the optimization problem; A method comprising:

2. A method for solving an optimization problem by a quantum processor, the solving method comprising steps of a method for determining a configuration of a quantum processor as described in claim 1, the solving method comprising a step of positioning the neutral atoms of the quantum processor according to the positioning determined to solve the optimization problem.

3. 3. The method of claim 2, further comprising: solving the optimization problem using a quantum approximate optimization algorithm called QAOA or a quantum adiabatic algorithm with the neutral atoms of the quantum processor positioned according to the determined positioning.

4. 4. The method of claim 1, wherein the learning tool is a manifold machine learning tool suitable for performing a non-linear reduction in the dimensionality of the data at the input of the tool.

5. The method of claim 4 , wherein the multiple machine learning tools are based on a multi-dimensional positioning algorithm.

6. The QUBO(M) matrix is ​​constructed in the form of a symmetric real matrix, and the step of determining the location comprises: a. The distance matrix (D 0 ) by determining the distance matrix (D 0 ) at positions i, j ([D 0 ] i,j ) corresponds to the distance between qubit i and qubit j; and b. The distance matrix (D 0 determining the position of the neutral atom according to 6. The method according to any one of claims 1 to 5, comprising:

7. The step of determining the position includes the step of: 0 ), and determining the intermediate matrix (U 0 ) has coefficients equal to zero on its diagonal and otherwise the same coefficients as the QUBO(M) matrix, and the distance matrix (D 0 ) is the intermediate matrix (U 0 7. The method of claim 6, wherein the signal intensity is obtained as a function of

8. The distance matrix (D 0 ) of the coefficient ([D 0 ] i,j ) is obtained according to the following formula: [0010] During the ceremony, [D 0 ] i,j is the distance matrix D 0 is the coefficient in row i and column j of [U 0 ] i,j is the intermediate matrix U 0 is the coefficient in row i and column j of C 6 8. The method of claim 7, wherein: ##EQU1## is a coefficient that depends on the Rydberg level chosen as the excited state for the neutral atom of the quantum processor.

9. The target Hamiltonian (H c ) satisfies the following formula, [0025] During the ceremony, H c is the target Hamiltonian for the optimization problem under consideration, M is the QUBO matrix, x is the state vector, x T is the transpose of the vector x, <Φ| is the bra of Φ in bra-ket notation, |Ψ> is the ket of Ψ in bra-ket notation, 9. The method according to any one of claims 1 to 8.

10. A computer program product having stored thereon a computer program comprising program instructions, the computer program being loaded into a data processing unit and configured to calculate a target Hamiltonian (H c ) and determining the location.

11. A readable storage medium having stored thereon a computer program comprising program instructions, the computer program being loaded into a data processing unit and, when the computer program is implemented in the data processing unit, the computer program is adapted to calculate the target Hamiltonian (H c ) and determining a position.

Citation Information

Patent Citations

  • SSN2521-327X