Quantum optimization method and device for Rydberg atom array and storage medium

Abstract

The invention relates to a quantum optimization method and device for a Rydberg atom array and a storage medium. The quantum optimization method of the Rydberg atom array comprises the following steps of expanding a source graph vertex through a cross lattice, replacing all substructures violating unit disc constraints in the cross lattice with logically equivalent replacement devices, and mapping the substructures into a real triangular lattice to generate a Rydberg atom array space layout diagram. According to the method, the vertexes of the source graph are expanded through the cross lattice, all substructures violating unit disc constraints in the cross lattice are replaced with logically equivalent replacement devices and mapped into the real triangular lattice, the Rydberg atomic array spatial layout graph can be generated, independent set constraint violations caused by long-tail interaction can be remarkably reduced through the triangular lattice, and the spatial layout of the Rydberg atomic array can be generated. Therefore, post-processing is reduced and even eliminated, the interpretability and the physical realizability of the result are improved, and the technical problems that the constraint violation rate of the quantum optimization result under KSG embedding is high and a large amount of post-processing is needed in the related technology are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of quantum computing, in particular to a Rydberg atom array quantum optimization method, device and storage medium. BACKGROUND

[0002] At present, in the field of quantum computing, Rydberg atom arrays have become the mainstream physical platform for solving combinatorial optimization problems (such as maximum independent set, graph coloring, etc.) due to their long-range interaction characteristics.

[0003] In related technologies, KSG (Kronecker Subgraph Embedding) is generally used as the embedding geometric structure to place Rydberg atoms. KSG provides an approximate geometric structure: in a square lattice, each vertex is not only connected to the nearest neighbor, but also includes the next nearest neighbor, i.e. diagonal connection, thereby forming a higher connectivity. The advantage of this method is that the geometric structure is clear, the vertex overhead is , and it is universal, so it has become one of the most common embedding methods.

[0004] However, in actual Rydberg arrays, the quantum optimization result under KSG embedding has a high constraint violation rate and requires a large amount of post-processing.

[0005] Therefore, it is necessary to design a new Rydberg atom array quantum optimization method to overcome the above problems. SUMMARY

[0006] The present application provides a Rydberg atom array quantum optimization method, device and storage medium, which can solve the technical problems of high constraint violation rate of quantum optimization result under KSG embedding and the need for a large amount of post-processing in related technologies.

[0007] In a first aspect, the present application provides a Rydberg atom array quantum optimization method, which includes the following steps: expanding the source graph vertices through the cross grid, replacing all substructures that violate the unit disc constraint in the cross grid with logically equivalent replacement devices, and mapping them to the real triangular lattice to generate a Rydberg atom array spatial layout graph.

[0008] In combination with the first aspect, in an implementation manner, the Rydberg atom array quantum optimization method includes the following steps: constructing a specific shape copy line for each vertex of the source graph and arranging an odd-length copy device to form a cross grid; According to the edge distribution in the source graph, all substructures in the cross grid that violate the unit circle constraint are replaced by the designed replacement device at the intersection points, and are mapped to the real triangular lattice to generate the spatial layout diagram of the Rydberg atom array.

[0009] In combination with the first aspect, in an implementation, the method further includes: constructing a specific shape replication line for each vertex of the source graph in a rectangular coordinate system, and arranging an odd number of replication devices to form a cross grid, including: allocating a unique horizontal slot for each vertex of the source graph; wherein the slot refers to a reserved position on a row of the cross grid, from which a specific shape replication line for the corresponding vertex can be generated, and the replication lines for different vertices do not overlap and can be ordered to intersect; calculating the extension range of each specific shape replication line in two different extension directions, so that it intersects at least once with the specific shape replication line of each adjacent vertex to carry the edge relationship of the source graph at the intersection point; placing an odd number of atoms on each specific shape replication line to form an odd number of replication devices; wherein the atoms at an odd sequence position on the specific shape replication line are logically equivalent to the vertices of the source graph.

[0010] In combination with the first aspect, in an implementation, the method further includes: converting the edge relationship of the source graph into a local constraint at the intersection point; identifying and replacing all substructures in the cross grid that violate the unit circle constraint using the designed replacement device; converting the vertex weights of the source graph into a detuning configuration of the spatial layout diagram of the Rydberg atom array, and mapping the atomic space coordinates in the rectangular coordinate system to the real triangular lattice to generate the spatial layout diagram of the Rydberg atom array and the detuning configuration.

[0011] In combination with the first aspect, in an implementation, the method further includes: traversing all edges of the source graph, locating the intersection point of the corresponding two replication devices, and introducing an exclusion constraint at the intersection point; for vertex pairs that do not exist, the intersection point remains unconstrained.

[0012] In combination with the first aspect, in an implementation, the method further includes: scanning the cross grid to identify all substructures that violate the unit circle constraint; According to the local logical relationship, the corresponding replacement device is selected for replacement, and it is ensured that the pins of the replacement device are aligned with the connection points of the original structure. If the alignment cannot be achieved, further fine adjustment is performed. If two replacement devices share the same pin vertex, the vertex weights are added when merging.

[0013] In combination with the first aspect, in an implementation, the fine adjustment step includes: Maintaining the odd length constraint of each copy device; Bending, extending or shortening the copy line of a specific shape to align with the pins of the replacement device; Verifying that all connections remain complete, without disconnection or incorrect connection.

[0014] In combination with the first aspect, in an implementation, the conversion of the source graph vertex weight into the detuning configuration of the Rydberg atom array spatial layout graph and the mapping of the atomic space coordinates in the rectangular coordinate system to the real triangular lattice to generate the Rydberg atom array spatial layout graph and the detuning configuration includes: Calculating the corresponding detuning bias for each source graph vertex weight; Adding the detuning bias to the base weight of the vertex logically equivalent to the vertex; Applying the mapping relationship to map the atomic space coordinates in the rectangular coordinate system to the real triangular lattice; Outputting the mapped Rydberg atom array spatial layout graph and the detuning configuration.

[0015] In a second aspect, an embodiment of the present application provides a quantum optimization device for a Rydberg atom array, which includes a processor, a memory, and a quantum optimization program for a Rydberg atom array stored on the memory and executable by the processor. When the quantum optimization program for a Rydberg atom array is executed by the processor, the steps of the quantum optimization method for a Rydberg atom array described above are implemented.

[0016] In a third aspect, an embodiment of the present application provides a computer readable storage medium having a quantum optimization program for a Rydberg atom array stored thereon. When the quantum optimization program for a Rydberg atom array is executed by a processor, the steps of the quantum optimization method for a Rydberg atom array described above are implemented.

[0017] The technical solutions provided by the embodiments of the present application have the following beneficial effects: By unfolding the source graph vertices in the cross graph, replacing all substructures that violate the unit disk constraint in the cross graph with logically equivalent replacement devices, and mapping to a real triangular lattice, a Rydberg atom array spatial layout graph can be generated. The triangular lattice can significantly reduce the independent set constraint violations caused by long tail interactions, thereby reducing or even eliminating post-processing, improving the interpretability and physical realizability of the results, and solving the technical problems of high constraint violation rate of quantum optimization results under KSG embedding and the need for a large amount of post-processing in related technologies. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 A flowchart of the present application for mapping a general MIS problem to a Rydberg atom array on a triangular lattice; Figure 1 In (a), an example problem is shown, (b) shows the creation of a cross graph, (c) shows the replacement of devices, and (d) shows the final triangular lattice embedding. Figure 2 Several replacement devices are shown in the example problem based on Figure 1 In (a), a cross device is shown, (b) shows a cross edge device, and (c) shows a T-shaped connector device. Figure 2 In (a), the difference in intensity performance after quantum annealing between triangular lattice embedding and KSG embedding is shown, and (b) shows the difference in violation rate after quantum annealing between triangular lattice embedding and KSG embedding. Figure 3 Several replacement devices are shown in the example problem based on Figure 4 Several replacement devices are shown in the example problem based on Figure 5 A hardware structure diagram of a quantum optimization device for a Rydberg atom array involved in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0019] In order to better enable persons skilled in the art to understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by persons skilled in the art without creative labor fall within the scope of protection of the present application.

[0020] From optimization problems to Rydberg quantum systems: Many NP-hard problems can be naturally mapped to the ground state solving process of a Rydberg atom array. Specifically, the classical Hamiltonian ground state of a Rydberg atom array naturally corresponds to the maximum weighted independent set problem on a unit disk graph: where the blocking effect implements the independence constraint, and the local detuning is used to represent the vertex weight. On this basis, by setting the Rabi frequency with detuned segmented linear pulses, which can gradually guide the system to approach the ground state of the Hamiltonian of the problem, thereby realizing quantum solution to the original optimization problem.

[0021] King Subgraph (KSG) based embedding: existing research generally uses KSG as the embedding geometry to place Rydberg atoms. KSG provides an approximate geometry: in the square lattice, each vertex is not only connected to the nearest neighbor, but also includes the next nearest neighbor, i.e. the diagonal connection, thereby forming a higher connectivity. In order to realize the embedding of an arbitrary graph in a finite geometry, the above embedding expands the vertices of the graph by constructing a cross lattice on the lattice using a replication device: each source graph vertex is replaced by an "L" shaped linear device composed of an odd number of atoms, which extends in the vertical and horizontal directions, and redistributes the connectivity relationship of the edges at the intersection and applies appropriate devices. The advantage of this method is that the geometry is clear, the vertex overhead is , and it is universal, so it has become one of the most common embedding methods at present.

[0022] However, KSG has a high violation rate: due to physical effects such as long-tail interactions, the measured results often appear to violate the independence constraint, and the solution even exceeds the size of the maximum independent set. It must be corrected by a large amount of post-processing (such as deleting or adding points), which increases the additional complexity and weakens the physical interpretability of the results. Device design is difficult and has poor scalability: the weight configuration of existing devices (such as cross devices) mainly relies on exhaustive or semi-manual methods. When the device size is slightly larger, the computational amount of this method increases sharply, which is almost infeasible. This makes it very difficult to systematically design and expand complex devices.

[0023] Embodiments of the present application provide a quantum optimization method, device and storage medium for a Rydberg atom array, which can solve the technical problems of high quantum optimization result constraint violation rate and the need for a large amount of post-processing under KSG embedding in related technologies.

[0024] First, some technical terms in the present application are explained and described to facilitate understanding by those skilled in the art.

[0025] UDG: Unit Disk Graph.

[0026] MIS: Maximum Independent Set.

[0027] MWIS: Maximum Weighted Independent Set.

[0028] To make the purpose, technical solutions and advantages of the present application clearer, the embodiments of the present application will be described in further detail below with reference to the drawings.

[0029] In one aspect, embodiments of this application provide a quantum optimization method for Rydberg atom arrays.

[0030] In one embodiment, reference is made to Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the quantum optimization method for Rydberg atom arrays in this application. Figure 1 As shown, the quantum optimization method for Rydberg atomic arrays includes: expanding the vertices of the source graph through a cross-lattice, replacing all substructures in the cross-lattice that violate the unit disk constraint with logically equivalent replacement devices, and mapping them to the real triangular lattice to generate a spatial layout diagram of the Rydberg atomic array.

[0031] See Figure 1 As shown, Figure 1 (a) provides an example problem, namely, how to solve it. The largest independent set of a graph. Figure 1 In (b), each vertex is unfolded into a replica device on a triangular grid and forms a cross grid. The vertices are numbered and their correspondence in the source graph are marked. The red edges correspond to the edges in the source graph. Figure 1 In (c), substructures in the cross grid that do not satisfy the unit disk constraint are replaced with logically equivalent devices. Figure 1 (d) shows the final triangular lattice embedding diagram, whose ground-state configuration of the numbered nodes can be directly mapped back to the solution of the original problem. Replacement devices, for example... Figure 2 Cross-connectors, cross-side devices, or T-shaped connectors, etc.

[0032] Traditional KSG embedding, although theoretically possessing While KSG embedding in quantum optimization suffers from high vertex overhead, in practical Rydberg arrays, the constraint violation rate of quantum optimization results is high, requiring extensive post-processing. This embodiment addresses this issue with a framework based on triangular lattice embedding using a cross-lattice. By expanding the vertices of the source graph through the cross-lattice, all substructures violating the unit disk constraint in the cross-lattice are replaced with logically equivalent replacement devices and mapped onto a real triangular lattice. This generates a spatial layout diagram of the Rydberg atom array, achieving highly robust embedding. The triangular lattice significantly reduces independent set constraint violations caused by long-tail interactions. Numerical results show that the constraint violation rate is reduced by nearly two orders of magnitude, thereby reducing or even eliminating post-processing, improving the interpretability and physical realizability of the results, and solving the technical problem of high constraint violation rate and extensive post-processing required in related technologies for quantum optimization results under KSG embedding.

[0033] The key advantage of triangular lattice embedding lies in its significantly reduced violation rate of independence constraints. In this embodiment, we introduce a geometric quality factor to measure the geometric robustness of different lattice structures in unit disk diagram (UDG) embedding. It is defined as: ,in, denotes the minimum distance between all unconnected vertices on the UDG, denotes the maximum distance between all connected vertices. Since the Rydberg interaction scales as , the relevant energy scale separation is . The larger , the more distinct the distance between edges and non-edges. In a triangular lattice: the nearest-neighbor distance is the lattice constant , and the next-nearest-neighbor distance is . Thus we have , which gives , . In a KSG: the neighbors include axial points (distance ) and diagonal points (distance ), thus . The nearest connected points are two-lattice away axial points, , which gives . It is clear that the triangular lattice outperforms the KSG in the geometric quality factor, with an energy scale separation about three times higher. This means that under the same physical conditions, the triangular lattice embedding can significantly reduce the independence constraint violations caused by long-tail interactions, improving the stability and interpretability of the embedding scheme. Figure 3 The results of quantum annealing simulation on the mapped graph ( Figure 1 (d) ) are shown in this embodiment, comparing the performance difference between triangular lattice embedding and KSG embedding. Figure 3 (a) gives the piecewise linear pulses used in the quasi-adiabatic evolution, including the Rabi frequency and the detuning . The shaded area represents the operation interval allowed by the Rabi frequency, i.e. , where . The choice of the lattice constant makes equal to the geometric mean of the two limits. Under the same pulse control conditions, the triangular lattice provides a wider operating energy window, making the evolution process better separate the interaction strengths of edges and non-edges. Figure 3 (b) shows that compared with KSG embedding, its violation rate is reduced by nearly two orders of magnitude in most time ranges, thereby reducing the dependence on post-processing and improving the interpretability of the results and the robustness of experimental implementation.

[0034] The triangular lattice mapping scheme described in this embodiment is applicable to MIS / MWIS problems with arbitrary connectivity, with a mapping vertex overhead of , where is the source graph vertex size. This method can be further extended to other NP-hard problems such as graph coloring and domination set.

[0035] Further, in one embodiment, the step of expanding the source graph vertices through a cross-lattice, replacing all substructures in the cross-lattice that violate the unit disk constraint with logically equivalent replacement devices, and mapping them to a real triangular lattice to generate a Rydberg atom array spatial layout diagram may include: S1: Construct a specific shape copy line for each vertex of the source image and arrange an odd number of copy devices to form a cross grid. In this step, the specific shape copy line can be, for example, an L-shaped copy line or an "I"-shaped copy line. It can be set according to requirements and is not limited here.

[0036] S2: At the intersection, based on the edge distribution in the source diagram, use the designed replacement device to replace all substructures in the intersection lattice that violate the unit disk constraint, and map them to the real triangular lattice to generate the Rydberg atom array spatial layout diagram.

[0037] In this embodiment, each vertex in the source graph is first replaced with a replica device of an odd number of nodes to form a two-dimensional cross lattice. Then, at the cross points, according to the edge distribution in the source graph, the original structure is replaced with pre-designed replacement devices (such as cross devices, cross-edge devices, T-shaped connectors, etc.) to ensure compliance with the geometric constraints of the triangular lattice. The final generated triangular lattice embedding maintains a one-to-one correspondence with the source graph MIS / MWIS problem.

[0038] In some embodiments, constructing a specific shape copy line for each vertex of the source image and arranging an odd number of copy devices to form a cross grid may include: S11: For each vertex of the source graph in a Cartesian coordinate system Assign a unique horizontal slot; where a slot refers to a reserved position on a row of intersecting grids, from which a specific shape copy line of the corresponding vertex can be generated, and the copy lines of different vertices do not overlap and can intersect in an orderly manner.

[0039] S12: Calculate the extension range of each specific shape copy line in two different extension directions, so that it intersects the specific shape copy lines of all adjacent vertices at least once, so as to carry the edge relationship of the source graph at the intersection.

[0040] S13: Place an odd number of atoms on each specific shape replication line to form an odd-length replication device; wherein, the atoms in the odd-order positions on the specific shape replication line are logically equivalent to the vertices of the source graph.

[0041] In this embodiment, step S1, namely the construction of the cross lattice, involves constructing an L-shaped replication line for each source map vertex and arranging odd-length atomic chains. Step S1 includes steps S11, S12, and S13. Step S11 involves constructing an L-shaped replication line for each vertex in a Cartesian coordinate system. allocating a unique horizontal slot, where "slot" refers to a reserved position on a row of the cross grid, from which a specific shape copy line (preferably L-shaped copy line in this embodiment) of the corresponding vertex can be generated, ensuring that copy lines of different vertices do not overlap and can be ordered to cross. Step S12, i.e. calculating the extension range of each L-shaped copy line in the horizontal and vertical directions, so that it at least crosses the copy lines of all adjacent vertices, thereby carrying the edge relationship of the source graph at the intersection. Step S13, i.e. placing an odd number of atoms on each L-shaped copy line to form an odd-length copy device; the odd-numbered atoms on the L-shaped copy line are logically equivalent to the vertices of the source graph and are used to maintain one-to-one correspondence in the calculation problem.

[0042] Further, in an embodiment, the replacing all substructures in the cross grid that violate the unit disc constraint at the intersection points with the designed replacement device according to the edge distribution in the source graph, and mapping to the real triangular lattice to generate the spatial layout diagram of the Rydberg atom array, can include: S21: converting the edge relationship of the source graph into local constraints at the intersection points.

[0043] S22: identifying and replacing all substructures in the cross grid that violate the unit disc constraint with the designed replacement device.

[0044] S23: converting the vertex weight of the source graph into the detuning configuration of the spatial layout diagram of the Rydberg atom array, and mapping the atomic space coordinates in the rectangular coordinate system to the real triangular lattice to generate the spatial layout diagram of the Rydberg atom array and the detuning configuration.

[0045] In this embodiment, after constructing the cross lattice, edge redistribution (i.e. step S21), device replacement and UDG repair (i.e. step S22), weight encoding, and coordinate transformation and physical mapping (i.e. step S23) are performed.

[0046] On the basis of the above technical solutions, illustratively, the converting the edge relationship of the source graph into local constraints at the intersection points includes: traversing all edges of the source graph, locating the intersection points corresponding to the two copy devices, and introducing mutual exclusion constraints at the intersection points, and for vertex pairs without edges, the intersection points remain unconstrained.

[0047] In this embodiment, S21 includes: S211: traversing all edges of the source graph , locating the intersection points corresponding to the two copy devices.

[0048] S212: introducing "mutual exclusion constraints" at the intersection points: the logically equivalent atoms on the two copy devices cannot be selected at the same time, corresponding to the edge relationship in the source graph. For vertex pairs without edges, the intersection points remain unconstrained.

[0049] Further, in an embodiment, the identifying and replacing all sub-structures violating the unit disk constraint with the designed replacement devices can include: S221: Scanning the crossbar to identify all sub-structures violating the unit disk constraint.

[0050] S222: Replacing the sub-structures with the corresponding replacement devices according to the local logic relationship, and ensuring that the pins of the replacement devices are aligned with the connection points of the original structure, and if not, further fine-tuning is needed; if two replacement devices share the same pin vertex, the vertex weight is added when combined.

[0051] In this embodiment, when replacing the devices, the crossbar is first scanned to identify the sub-structures violating the UDG constraint of the triangular lattice; then the corresponding replacement devices (such as cross devices, edge-crossing devices, T-connection devices, etc., see Figure 4 ) are selected according to the local logic relationship; then the device replacement is performed to ensure that the pins (i.e., the logical vertices exposed and connectable by the device) of the new device (i.e., the replacement device) are aligned with the connection points of the original structure, and if not, further fine-tuning is needed in the next step; if two replacement devices share the same pin vertex, the device combination principle is applied: the vertex weight is added when combined to maintain the consistency of the energy function.

[0052] For example, the fine-tuning step includes: Step a: maintaining the odd length constraint of each replicated device.

[0053] Step b: bending, extending, or shortening the specific shape of the replicated line to align with the pins of the replacement device.

[0054] Step c: verifying that all connections remain complete without disconnection or incorrect connection.

[0055] In this embodiment, the fine-tuning step, i.e., geometric fine-tuning and connection preservation, adjusts the shape of the replicated line to fit the device interface. Specifically, the odd length constraint of each replicated device is maintained, and the specific shape of the replicated line is bent, extended, or shortened to align with the pins of the replacement device, as shown in Figure 4 Finally, it is verified that all connections remain complete without disconnection or incorrect connection.

[0056] Further, in some optional embodiments, the conversion of the source graph vertex weight into the detuning configuration of the Rydberg atom array space layout diagram and the mapping of the atomic space coordinates in the rectangular coordinate system to the real triangular lattice to generate the Rydberg atom array space layout diagram and the detuning configuration includes: S231: calculating the corresponding detuning bias for each source graph vertex weight.

[0057] S232: Add the detuning bias to the base weight of the vertex logically equivalent to the vertex.

[0058] S233: Apply the set mapping relationship to map the atomic space coordinates in the rectangular coordinate system to the real triangular lattice.

[0059] S234: Output the mapped Rydberg atom array space layout diagram and the set detuning.

[0060] In this embodiment, step S23 includes the steps of weight encoding, coordinate transformation and physical mapping. The step of weight encoding, i.e., converting the source graph vertex weight into the detuning configuration of the Rydberg atom array space layout diagram, includes the following specific steps: for each source graph vertex weight , the corresponding detuning bias is calculated; the is added to the base weight of the vertex logically equivalent to the vertex, and the size of is controlled: it must be large enough to distinguish the target state from the non-target state (avoid degeneracy), and it must be smaller than the energy difference defined by the blocking effect inside the device to avoid damaging the device function. The step of coordinate transformation and physical mapping, i.e., converting the square grid representation to the real triangular lattice geometry, includes the following specific steps: applying the coordinate transformation formula to map the algorithm internal index position (i.e., atomic space coordinates) in the rectangular coordinate system to the real triangular lattice; then verifying that all edge lengths after transformation still satisfy the unit disc constraint; finally outputting all vertex coordinates after mapping and the mapping relationship for subsequent experimental implementation and measurement readout.

[0061] The quantum optimization method of the Rydberg atom array provided in the embodiments of the present application uses a triangular lattice embedding mapping algorithm, and the target is to: embed the maximum independent set / maximum weighted independent set problem of any general graph (i.e., source graph) into a triangular lattice unit disc graph, and realize the conversion of the logically equivalent quantum optimization problem. The input is: a graph : a general graph to be embedded, including a vertex set , an edge set , and a vertex weight set (optionally). The output is: (1) UDG on the triangular lattice: the embedded unit disc graph coordinates and edge relationship. (2) Vertex weight configuration Δ: the weight on each mapped atom. (3) Mapping relationship: the correspondence relationship between the source graph vertex and the logically equivalent atom in the embedded graph.

[0062] The present application aims to provide a general and scalable graph optimization embedding scheme on a triangular lattice: efficiently map the computational complex problems, especially the maximum independent set / maximum weighted independent set (MIS / MWIS) problem on general graphs to unit disk graphs (UDGs) and robustly implement quantum optimization on programmable Rydberg atom arrays with triangular lattice geometry. Design and automate search for key devices that meet geometric constraints: use an integer programming-based systematic search method.

[0063] In a second aspect, the embodiments of the present application provide a quantum optimization device of a Rydberg atom array. The quantum optimization device of the Rydberg atom array can be a personal computer (PC), a notebook computer, a server, or other device with data processing function.

[0064] Reference Figure 5 , Figure 5 The figure is a schematic diagram of the hardware structure of the quantum optimization device of the Rydberg atom array involved in the embodiments of the present application. In the embodiments of the present application, the quantum optimization device of the Rydberg atom array can include a processor, a memory, a communication interface, and a communication bus.

[0065] The communication bus can be of any type, used to interconnect the processor, the memory, and the communication interface.

[0066] The communication interface includes input / output (I / O) interface, physical interface, and logical interface, etc. for realizing the interconnection of devices inside the quantum optimization device of the Rydberg atom array, and for realizing the interconnection of the quantum optimization device of the Rydberg atom array with other devices (such as other computing devices or user devices). The physical interface can be an Ethernet interface, a fiber interface, an ATM interface, etc.; the user device can be a display (Display), a keyboard (Keyboard), etc.

[0067] The memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0068] The processor can be a general-purpose processor, which can invoke a quantum optimization program of the array of Rydberg atoms stored in the memory and execute the quantum optimization method of the array of Rydberg atoms provided in the embodiments of the present application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the quantum optimization program of the array of Rydberg atoms is invoked can refer to the embodiments of the quantum optimization method of the array of Rydberg atoms of the present application, which will not be repeated here.

[0069] Those skilled in the art can understand that the hardware structure shown in the above Figure 5 The hardware structure shown in the above does not constitute a limitation on the present application, and can include more or fewer components than shown, or combine certain components, or different arrangement of components.

[0070] In a third aspect, the embodiments of the present application further provide a readable storage medium.

[0071] The readable storage medium of the present application stores a quantum optimization program of the array of Rydberg atoms, wherein when the quantum optimization program of the array of Rydberg atoms is executed by a processor, the steps of the quantum optimization method of the array of Rydberg atoms as described above are implemented.

[0072] The method implemented when the quantum optimization program of the array of Rydberg atoms is executed can refer to the embodiments of the quantum optimization method of the array of Rydberg atoms of the present application, which will not be repeated here.

[0073] It should be noted that the above sequence numbers of the embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0074] The terms "include" and "have" and any variations thereof in the specification and claims of the present application and the above-described drawings are intended to cover the non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but can optionally include other steps or units not listed or can optionally include other steps or units inherent to these processes, methods, products or devices. The terms "first", "second" and "third" and the like descriptions are used to distinguish different objects, and do not represent the order or limit the types of "first", "second" and "third".

[0075] In the description of the embodiments of the present application, "exemplary", "for example", "e.g." or "for instance" are used on the basis that a person of ordinary skill in the art will be able to bring to mind many examples of a given implementation as the implementation described in the embodiments of the present application is exemplary. Any embodiment or design scheme described as "exemplary", "for example" or "for instance" in the embodiments of the present application should not be interpreted as being more preferred than other embodiments or design schemes. Rather, the use of "exemplary", "for example", "e.g." or "for instance" is intended to present concepts in a particular manner.

[0076] In the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B; "and / or" in the text only describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can mean that A exists alone, A and B exist together, and B exists alone. In addition, in the description of the embodiments of the present application, "multiple" means two or more than two.

[0077] In some processes described in the embodiments of the present application, a plurality of operations or steps are included in a specific order, but it should be understood that these operations or steps can be executed or executed in parallel without the order in which they appear in the embodiments of the present application. The serial number of the operation is only used to distinguish different operations, and the serial number itself does not represent any execution order. In addition, these processes can include more or fewer operations, and these operations or steps can be executed in sequence or in parallel, and these operations or steps can be combined.

[0078] From the above description of the embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment method can be realized by means of software and necessary general hardware platform, of course, it can also be realized by hardware, but in many cases the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as a ROM / RAM, a magnetic disk, an optical disk) as described above, and includes a plurality of instructions for causing a terminal device to execute the method described in each embodiment of the present application.

[0079] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent process transformation using the content of the specification and drawings of the present application, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A quantum optimization method for a Rydberg atom array, characterized in that, The quantum optimization method for the Rydberg atom array includes the following steps: By expanding the vertices of the source graph through the cross-lattice, all substructures in the cross-lattice that violate the unit disk constraint are replaced with logically equivalent replacement devices and mapped to the real triangular lattice to generate the Rydberg atom array spatial layout diagram.

2. The quantum optimization method for Rydberg atom arrays as described in claim 1, characterized in that, The process of expanding the source graph vertices through a cross-lattice, replacing all substructures in the cross-lattice that violate the unit disk constraint with logically equivalent replacement devices, and mapping them to a real triangular lattice to generate a Rydberg atom array spatial layout diagram includes: For each vertex of the source image, construct a copy line of a specific shape and arrange an odd number of copy devices to form a cross grid; At the intersection, based on the edge distribution in the source diagram, all substructures that violate the unit disk constraint in the intersection lattice are replaced with pre-designed replacement devices and mapped to the real triangular lattice to generate the Rydberg atomic array spatial layout diagram.

3. The quantum optimization method for Rydberg atom arrays as described in claim 2, characterized in that, The process of constructing a specific shape copy line for each vertex of the source image and arranging an odd number of copy devices to form a cross grid includes: For each vertex of the source graph in a Cartesian coordinate system Assign a unique horizontal slot; where a slot refers to a reserved position on a row of intersecting grids, from which a specific shape copy line of the corresponding vertex can be generated, and the copy lines of different vertices do not overlap and can intersect in an orderly manner; Calculate the extent of each specific shape copy line in two different extension directions, so that it intersects with the specific shape copy lines of all adjacent vertices at least once, so as to carry the edge relationships of the source graph at the intersection point; An odd number of atoms are placed on each copy line of a specific shape to form a copy device of odd length; wherein the atoms at odd-numbered positions on the copy line of a specific shape are logically equivalent to the vertices of the source graph.

4. The quantum optimization method for Rydberg atom arrays as described in claim 2, characterized in that, The process involves replacing all substructures violating the unit disk constraint in the cross grid at the intersection points according to the edge distribution in the source diagram, using pre-designed replacement devices, and mapping them to the real triangular lattice to generate a Rydberg atom array spatial layout diagram, including: Convert the edge relationships of the source graph into local constraints of the intersection points; Identify and replace all substructures in the cross grid that violate the unit disk constraint using the designed replacement devices; The vertex weights of the source image are transformed into the detuned configuration of the Rydberg atomic array spatial layout diagram, and the atomic space coordinates in the Cartesian coordinate system are mapped to the real triangular lattice to generate the Rydberg atomic array spatial layout diagram and the detuned settings.

5. The quantum optimization method for Rydberg atom arrays as described in claim 4, characterized in that, The process of converting the edge relationships of the source graph into local constraints of the intersection points includes: Traverse all edges of the source graph, locate the intersection point of the two corresponding replicas, and introduce a mutual exclusion constraint at the intersection point. For vertex pairs without edges, the intersection point remains unconstrained.

6. The quantum optimization method for Rydberg atom arrays as described in claim 4, characterized in that, The process of identifying and replacing all substructures in the cross grid that violate the unit disk constraint using pre-designed replacement devices includes: Scan the intersecting grid to identify all substructures that violate the unit disk constraint; Select the corresponding replacement device based on the local logic relationship and ensure that the pins of the replacement device are aligned with the connection points of the original structure. If they cannot be aligned, further fine-tune them. If two replacement devices share the same pin vertex, the vertex weights are added when merging.

7. The quantum optimization method for Rydberg atom arrays as described in claim 6, characterized in that, The fine-tuning steps include: Maintain the odd-length constraint for each replicated device; Bending, extending, or shortening a copy line of a specific shape to align it with the pins of the replacement device; Verify that all connections remain intact, with no disconnections or faulty connections.

8. The quantum optimization method for Rydberg atom arrays as described in claim 4, characterized in that, The process of converting the vertex weights of the source image into the detuned configuration of the Rydberg atomic array spatial layout diagram, and mapping the atomic space coordinates in the Cartesian coordinate system to the real triangular lattice to generate the Rydberg atomic array spatial layout diagram and detuning settings includes: For each vertex weight in the source graph, calculate its corresponding detuning bias; Add the detuning bias to the base weight of the logically equivalent vertex; By applying the established mapping relationship, the atomic spatial coordinates in the Cartesian coordinate system are mapped to the real triangular lattice; Output the spatial layout of the Rydberg atom array after mapping and the setting of detuning.

9. A quantum optimization device for a Rydberg atom array, characterized in that, The quantum optimization device for the Rydberg atom array includes a processor, a memory, and a quantum optimization program for the Rydberg atom array stored in the memory and executable by the processor, wherein when the quantum optimization program for the Rydberg atom array is executed by the processor, it implements the steps of the quantum optimization method for the Rydberg atom array as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a quantum optimization program for a Rydberg atom array, wherein when the quantum optimization program for the Rydberg atom array is executed by a processor, it implements the steps of the quantum optimization method for a Rydberg atom array as described in any one of claims 1 to 8.

Images