Control execution method and equipment for quantum eigenstate solving task

Abstract

The embodiment of the invention provides a control execution method and equipment for a quantum eigenstate solving task, the solving task is carried out by using a variable component sub-eigenstate solving system of classical computational simulation, the variable component sub-eigenstate solving system comprises a plurality of modules designed and decoupled based on a workflow architecture, and the method is executed through a task control system. The method comprises: receiving a first configuration operation of a first user; calling an interface of at least one module according to input data specified by the first configuration operation so as to construct a first workflow task; and scheduling and executing the first workflow task to obtain task output. Therefore, different workflow tasks are generated by freely combining and configuring any module in the plurality of modules, and the workflow tasks are scheduled and executed to obtain the solving result, so that the calculation speed and the performance of the variable component sub-eigenstate solving system can be improved.

Description

Technical Field

[0001] The present application relates to quantum computing, and more particularly to a method and device for controlling and executing a quantum eigenstate solution task. Background Art

[0002] The variational quantum eigensolver (VQE) is increasingly used in the field of quantum computing. It is becoming a powerful tool for research in multiple disciplines such as quantum chemistry, materials science, and nuclear physics. By accurately simulating and analyzing complex quantum systems, VQE brings new solutions to these fields.

[0003] In the current variational quantum eigenstate solution, the processing of Hamiltonians, the preparation of quantum states, and the use of optimizers are usually implemented in a highly coupled manner and need to be completed within the same classical computing framework. This makes it difficult to achieve parallel processing and flexible management of computing resources for different computing tasks involved in VQE (such as Hamiltonian optimization, quantum state parameter update, etc.), affecting the overall computing speed and performance. Summary of the invention

[0004] The present application provides a control execution method and device for a quantum eigenstate solution task, which can improve the computational speed and performance of variational quantum eigenstate solution.

[0005] In a first aspect, the present application provides a control execution method for a quantum eigenstate solution task, wherein the solution task is performed using a variational quantum eigenstate solution system simulated by classical computing, wherein the variational quantum eigenstate solution system includes a plurality of modules designed and decoupled based on a workflow architecture, wherein the plurality of modules include a Hamiltonian module for constructing a Hamiltonian, a quantum circuit module for simulating a quantum circuit, and an optimizer module for performing quantum state parameter iteration; the method is executed by a task control system, and the method includes:

[0006] receiving a first configuration operation of a first user, wherein the first configuration operation at least specifies at least one module among the plurality of modules and input data;

[0007] Calling an interface of at least one module according to input data, thereby constructing a first workflow task;

[0008] The first workflow task is scheduled to be executed to obtain a task output, where the task output is a final state or intermediate state result of the variational quantum eigenstate solution system.

[0009] Therefore, the computing speed and performance of the variational quantum eigenstate solution system are improved by flexibly configuring and calling multiple decoupled modules to build workflow tasks, and flexibly scheduling and managing workflow tasks to complete the solution of variational quantum eigenstates.

[0010] In a second aspect, the present application provides a control execution device for a quantum eigenstate solution task, wherein the solution task is performed using a variational quantum eigenstate solution system simulated by classical computing, wherein the variational quantum eigenstate solution system includes a plurality of modules designed and decoupled based on a workflow architecture, wherein the plurality of modules include a Hamiltonian module for constructing a Hamiltonian, a quantum circuit module for simulating a quantum circuit, and an optimizer module for performing quantum state parameter iteration; the device includes:

[0011] A receiving module, configured to receive a first configuration operation of a first user, wherein the first configuration operation at least specifies at least one module among the plurality of modules and input data;

[0012] A construction module, used to call an interface of at least one module according to input data, so as to construct a first workflow task;

[0013] The scheduling execution module is used to schedule and execute the first workflow task to obtain the task output, where the task output is the final state or intermediate state result of the variational quantum eigenstate solution system.

[0014] In a third aspect, the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method described in the first aspect or any possible implementation of the first aspect.

[0015] It can be understood that the beneficial effects of the second to third aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. 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.

[0017] Figure 1 A control execution framework diagram of a quantum eigenstate solution task provided in an embodiment of the present application;

[0018] Figure 2 A diagram of a control execution method for a quantum eigenstate solution task provided in an embodiment of the present application;

[0019] Figure 3a A workflow task diagram of a variational quantum eigenstate solution system provided in an embodiment of the present application;

[0020] Figure 3bA workflow task diagram of a variational quantum eigenstate solution system provided in an embodiment of the present application;

[0021] Figure 4 The figure is a framework diagram of a control execution device for a quantum eigenstate solution task provided in an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0023] The variational quantum eigenstate solver is a quantum algorithm designed to simulate and solve the ground state energy problem of a quantum system. Its goal is to approximately solve the lowest energy state of a quantum system by calculating the expected value of the Hamiltonian.

[0024] At present, the variational quantum eigenstate solution system is mainly implemented through classical computing. The entire calculation process includes: simulating quantum circuits to generate parameterized quantum states, calculating the expectation value of the Hamiltonian based on the quantum state, and adjusting the parameters of the quantum state based on the expectation value using a classical optimization algorithm. The above steps are iterated by repeatedly executing them to approach the lowest energy state of the quantum system.

[0025] In the variational quantum eigenstate solution system, the Hamiltonian is used to describe the energy structure of the quantum system and guide the preparation and optimization process of the quantum state. Therefore, the variational quantum eigenstate solution system may involve switching different Hamiltonians to adjust the quantum circuit; it may also involve using electronic circuits to simulate different types of quantum circuits, such as UCCSD circuits or Hardware-Efficient circuits, and may involve selecting different optimizers for parameter iteration to reach convergence as quickly as possible.

[0026] As mentioned above, since the modules used in the solution process are usually implemented in a highly coupled manner, on the one hand, each solution needs to completely execute the entire calculation process of the variational quantum eigenstate solution within the classical computing framework. The user can only obtain the final solution result generated by repeated iterations based on the set Hamiltonian, and cannot observe and intervene in the intermediate states that appear in the iterative process in a timely manner;

[0027] On the other hand, each calculation process can only select a single Hamiltonian, a single type of quantum circuit, and a single optimizer for solving, and cannot achieve parallel computing of multiple tasks (such as Hamiltonian optimization, quantum state parameter update, etc.) and flexible management of computing resources;

[0028] On the other hand, it is also not possible to perform separate calculations on a certain module or a combination of some modules, and to select, for example, the Hamiltonian, quantum circuit type or optimizer based on the calculation results to increase the iteration rate.

[0029] Therefore, it is necessary to better support the implementation of quantum algorithms in the variational quantum eigenstate solution system, especially for tasks such as large-scale and complex quantum chemical simulations.

[0030] Workflow is a system design and management method used to coordinate and automate various tasks and steps in a business process. The Workflow architecture provides a systematic way to automate and optimize complex business processes by defining and managing tasks, processes, rules, data flows, and events. It decomposes the system into independent modules through decoupling, allowing each module to handle its tasks independently, and formulates rules for data input and output for each module. Modules interact through interfaces formulated using these rules, and modules can be flexibly combined through these interfaces to generate different workflow tasks.

[0031] In this way, when a part of the system needs to be updated or modified, it will not affect other parts, thereby reducing the overall complexity of the system, improving the flexibility and maintainability of the system, and generating customized workflow tasks according to user needs, thereby achieving efficient management and performance optimization of complex processes.

[0032] The key principles of the Workflow architecture include:

[0033] Module decoupling: Each module focuses on a specific task and interacts through well-defined interfaces without relying on the internal implementation of other modules.

[0034] Flexible combination of modules: Modules can be dynamically combined and configured according to business needs. Through predefined rules and interfaces, different modules can be flexibly arranged and rearranged to adapt to different workflows and application scenarios.

[0035] Single data flow: Data is transferred between tasks and modules in a predefined path and order. After each module processes the data, it passes the result to the next module, forming a continuous data processing chain. This design ensures the traceability and consistency of data flow, reduces data redundancy and repeated processing, and improves the efficiency and accuracy of data processing.

[0036] In view of this, the embodiment of the present application uses a variational quantum eigenstate solution system simulated by classical computing to solve the quantum eigenstate solution task. The variational quantum eigenstate solution system includes multiple modules designed and decoupled based on the workflow architecture. The task control system is also used to configure and call at least one module in multiple modules to build a workflow task according to the user's configuration operation, and schedule the execution of the workflow task to complete the variational quantum eigenstate solution. Thus, by flexibly configuring and calling multiple modules after decoupling to build workflow tasks, and flexibly scheduling and managing workflow tasks to complete the solution of variational quantum eigenstates, the computing speed and performance of the variational quantum eigenstate solution system are improved.

[0037] For example, Figure 1 A control execution framework diagram of a quantum eigenstate solution task provided in an embodiment of the present application is shown in FIG.

[0038] like Figure 1 As shown, the control execution framework includes a variational quantum eigenstate solution system and a task control system. The variational quantum eigenstate solution system includes multiple modules designed and decoupled based on a workflow architecture, and the multiple modules include a Hamiltonian module for constructing a Hamiltonian, a quantum circuit module for simulating a quantum circuit, and an optimizer module for performing quantum state parameter iteration.

[0039] The task control system includes a front-end interface and a back-end. The front-end interface (i.e., the client) can display user interaction units corresponding to multiple modules of the variational quantum eigenstate solution system to enable the user to perform configuration operations. The configuration operation at least specifies at least one module among the multiple modules, the parameter settings of the module, and the input data.

[0040] The backend (i.e., server side) includes task definition and configuration unit, task management unit, task scheduling unit, task execution unit, data processing unit, and task monitoring unit, which are responsible for implementing data processing, database management, and business logic. There is a communication interface between the front-end interface and the back-end, such as HTTP API, WebSocket, etc., for data exchange and accurate execution of instructions.

[0041] The task control system configures each module by passing input data to the variational quantum eigenstate solution system, and calls each module through the interface provided by each module.

[0042] The process of executing the quantum eigenstate solution task includes: first, users can flexibly perform configuration operations according to specific computing needs in a visual way in the front-end interface, including selecting the required module from multiple modules and setting specific parameters for each module. The front-end interface allows multiple users to perform configuration operations at the same time.

[0043] Secondly, after receiving the configuration operation of the user, the front-end interface of the task control system sends the configuration operation of the user to the back-end of the task control system through the communication interface.

[0044] Next, the task definition and configuration unit of the backend configures parameters of multiple modules in the variational quantum eigenstate solution system according to the user's configuration operations, and calls the interface provided by the module to construct a workflow task and send it to the task management unit.

[0045] The task management unit receives workflow tasks, identifies the priorities and resource requirements of workflow tasks, and sends the workflow tasks to the task scheduling unit while ensuring system stability and reliability.

[0046] The task scheduling unit dynamically arranges and allocates computing resources according to the priorities and resource requirements of workflow tasks. Its goal is to ensure that system resources are optimally utilized, thereby improving the efficiency of computing task execution and overall system performance.

[0047] The task execution unit receives the workflow tasks and computing resources assigned by the task scheduling unit, calls multiple modules according to the predetermined logic and configuration, executes the workflow tasks according to the predetermined code logic, and finally obtains the results.

[0048] The data processing unit transmits, organizes and analyzes the data generated during the execution of the quantum eigenstate solution task, and feeds back the analysis results and the execution results of the task execution unit to the user. Necessary data pre-processing or post-processing can also be performed before feedback to ensure the accuracy and availability of the data.

[0049] The task monitoring unit tracks the execution status and life cycle of workflow tasks in real time. It records the progress of workflow tasks, resource usage, and possible exceptions. All this information is recorded in detail in the log to ensure the transparency and traceability of the workflow task execution process, which is convenient for users to conduct subsequent review and optimization.

[0050] Therefore, through the modular design of the workflow architecture and the dynamic resource management mechanism, the efficiency, flexibility and scalability of the variational quantum eigenstate solution system are achieved, enabling the variational quantum eigenstate solution computational tasks to run reliably in complex computing environments.

[0051] Next, based on the above content, a control execution method for a quantum eigenstate solution task provided in an embodiment of the present application is introduced.

[0052] For example, Figure 2 A control execution method diagram of a quantum eigenstate solution task provided in an embodiment of the present application is shown in FIG. The control execution method can be run on any computing unit, server, device, or device cluster with computing and processing capabilities.

[0053] Before executing the control execution method, a variational quantum eigenstate solution system is simulated and implemented using classical computing resources. Classical computing resources refer to hardware platforms that rely on classical processing units (CPUs) to process information and perform computing tasks, such as servers, server clusters, etc. The variational quantum eigenstate solution system is used to complete the quantum eigenstate solution task. The simulated variational quantum eigenstate solution system includes multiple modules designed and decoupled based on a workflow architecture, including a Hamiltonian module for constructing a Hamiltonian, a quantum circuit module for simulating a quantum circuit, and an optimizer module for performing quantum state parameter iteration.

[0054] The above control execution method mainly includes the following execution steps:

[0055] Step S201: receiving a first configuration operation of a first user, where the first configuration operation at least specifies at least one module among a plurality of modules and input data.

[0056] In one embodiment, a first configuration operation of a first user is received using a front-end interface of a task control system. The front-end interface includes user interaction units corresponding to multiple modules of a variational quantum eigenstate solution system, such as a visualized Hamiltonian module, a quantum circuit module, and an optimizer module.

[0057] Exemplarily, the first user drags, clicks, or inputs in a dialog box on the front-end interface by touching or operating the mouse, selects at least one user interaction unit corresponding to at least one module, and sets input data in the interface corresponding to the at least one user interaction unit.

[0058] Optionally, the task control system further includes a visualization observation tool. The first configuration operation further includes connecting the visualization observation tool to at least one module to observe the calculation result of at least one module in real time during the solution process.

[0059] For example, the first user can also connect a visualization observation tool to at least one module in the front-end interface to observe the calculation results of at least one module in real time during the solution process. The visualization observation tool analyzes the calculation results and can display the calculation results of at least one module in real time in the form of text or charts. The visualization observation tool is particularly important when iteration is required, as it allows users to observe the iteration results in real time and intervene in a timely manner.

[0060] By selecting the user interaction unit, at least one module for executing the quantum eigenstate solution task can be specified. By setting input data in the interface corresponding to the user interaction unit, the connection relationship between at least one module can be specified, and the initial parameters or operation configuration parameters (such as the number of operations) of at least one module can also be specified.

[0061] It can be understood that the first user needs to set input data in the interface corresponding to the user interaction unit according to the preset input and output rules of each module. For each module corresponding to the user interaction unit, the preset input and output rules of the module are determined when the variational quantum eigenstate solution system is decoupled based on the workflow architecture design.

[0062] First, when the Hamiltonian is decoupled from the variational quantum eigenstate solution system to obtain the Hamiltonian module, the Hamiltonian module is set to only accept user input, and the Hamiltonian is constructed based on the user input. The final output can be seamlessly connected to the Hamiltonian data of any other system or module. The user input of the Hamiltonian module is the quantum system data provided by the user.

[0063] For example, taking the application of the variational quantum eigenstate solution system in quantum chemistry or materials science as an example, the input data of the Hamiltonian module is set to the molecule type, the spatial structure and bond length of the molecule provided by the first user. The output data of the Hamiltonian module is the first Hamiltonian constructed according to the molecule type, the spatial structure and bond length of the molecule.

[0064] Optionally, Hamiltonian decoupling also involves re-decoupling the decoupled Hamiltonian module to obtain multiple Hamiltonian quantum modules. For example, when it is necessary to calculate larger molecules (such as proteins), because the Hamiltonian of the molecule is relatively large, it is difficult to use variational quantum eigenstates to solve system optimization. Therefore, the Hamiltonian is decomposed into multiple sub-operators, and multiple sub-operators are used to describe a Hamiltonian as a whole. Multiple sub-operators can be calculated in parallel to obtain expected values, which can greatly reduce the problem of long calculation time caused by the large Hamiltonian. Each sub-operator corresponds to a Hamiltonian quantum module, and a complete Hamiltonian is constructed by the user's input to the relevant Hamiltonian quantum module.

[0065] Secondly, when the quantum circuit is decoupled from the variational quantum eigenstate solution system to obtain a quantum circuit module, the quantum circuit module is set to receive the output of the Hamiltonian module or accept user input, and the output of the quantum circuit module is exported to other systems or modules for execution, thereby achieving effective integration with other parts.

[0066] For example, the input data of the quantum circuit module is set to the output data of the Hamiltonian module, or the second Hamiltonian provided by the first user, and the quantum circuit type and quantum gate information provided by the first user. The output data of the quantum circuit module is set to the first parameterized quantum state generated by the quantum circuit simulated by the quantum circuit type and the quantum gate, and the expected value of the first or second Hamiltonian calculated according to the first parameterized quantum state.

[0067] Quantum gate information includes quantum gate type, quantum gate parameters, quantum gate bits, etc. Quantum circuit types of quantum circuit modules include UCCSD circuits, Hardware-Efficient circuits, etc.

[0068] In the classical computing framework, the output of one module is the input of the next module, and there is data dependency between them. In the variational quantum eigenstate solution system based on the workflow architecture, users are allowed to set independent input data for each decoupled module according to preset tasks, without having to rely entirely on the output of the previous module.

[0069] Exemplarily, when the input data of the quantum circuit module is set to the second Hamiltonian provided by the first user, the first configuration operation may not specify the Hamiltonian module. By setting the input data of the quantum circuit module to the second Hamiltonian provided by the first user, the second Hamiltonian is subsequently used to solve the quantum eigenstate. In this case, only the solution performance of the quantum circuit module in the process of solving the quantum eigenstate can be observed, thereby improving the solution efficiency of the variational quantum eigenstate solution system.

[0070] Finally, when the optimizer is decoupled from the variational quantum eigenstate solver system to obtain the optimizer module, the user is allowed to match different optimization strategies to the optimizer module, and select or adjust different optimizers to meet specific computing tasks without modifying other parts of the variational quantum eigenstate solver system.

[0071] For example, the input data of the optimizer module is set to the output data of the quantum circuit module, or the first user provides the second parameterized quantum state and the expected value of the third Hamiltonian calculated according to the second parameterized quantum state, as well as the optimizer type, optimization iteration number and optimization constraint provided by the first user. The third Hamiltonian is the Hamiltonian set by the first user in the optimizer module. The output data of the optimizer module is set to the optimized quantum state parameters, and the minimized expected value.

[0072] Similar to the quantum circuit module, the output data of the optimizer module can be set as the output data of a quantum circuit module, or it can accept user input for specification.

[0073] Therefore, by setting the input and output modules of multiple modules after decoupling when performing module decoupling of the variational quantum eigenstate solution system based on workflow architecture design, users can flexibly configure and call multiple modules after decoupling in the front-end interface according to preset input and output rules to build different workflow tasks and improve the calculation speed of the variational quantum eigenstate solution system.

[0074] Step S202: Calling the interface of at least one module according to input data, thereby constructing a first workflow task.

[0075] In one embodiment, for multiple modules designed and decoupled based on the workflow architecture, they each have independent input and output rules. A module can be configured according to parameter settings based on the input and output rules, and an interface of a module can be called according to input data to independently construct a first workflow task and independently perform calculations. It is also possible to construct a first workflow task by flexibly combining multiple modules to jointly perform calculations.

[0076] The first workflow task includes multiple subtasks, each subtask is executed based on at least one module, and different subtasks have different input data; scheduling and executing the first workflow task includes: executing the multiple subtasks in parallel based on different input data.

[0077] For the multiple subtasks included in the first workflow task, use Figure 1 The task scheduling unit in the system dynamically arranges and allocates computing resources according to the priority and resource requirements of each subtask.

[0078] In one implementation, at least one module specified by the first configuration operation is a single module; constructing the first workflow task includes: taking input data as input, calling an interface of the single module; and output of the single module as task output.

[0079] Among them, the single module is a Hamiltonian module, the input data is the quantum system data provided by the first user; and the task output is the first Hamiltonian determined according to the quantum system data.

[0080] Or the single module is a quantum circuit module, the input data includes the second Hamiltonian, quantum circuit type and quantum gate information provided by the first user; the task output is the first parameterized quantum state generated by the quantum circuit simulated by the quantum circuit type and the quantum gate, and the expectation value of the second Hamiltonian calculated based on the first parameterized quantum state.

[0081] Or the single module is an optimizer module, the input data includes the second parameterized quantum state provided by the first user and the expected value of the third Hamiltonian calculated based on the second parameterized quantum state, as well as the optimizer type, the number of optimization iterations and the optimization constraints provided by the first user; the task output is the optimized quantum state parameters and the minimized expected value of the third Hamiltonian.

[0082] In another implementation, at least one module specified by the first configuration operation is a Hamiltonian module and a quantum circuit module; the input data includes quantum system data provided by the first user, as well as quantum circuit type and quantum gate information.

[0083] Constructing the first workflow task includes: taking quantum system data as input, calling the Hamiltonian module; taking quantum circuit type, quantum gate information, and target Hamiltonian output by the Hamiltonian module as input, calling the quantum circuit module; the task output is a first parameterized quantum state generated by a quantum circuit simulated by the quantum circuit type and the quantum gate, and an expected value of the target Hamiltonian calculated based on the first parameterized quantum state.

[0084] In another implementation, at least one module specified by the first configuration operation is a quantum circuit module and an optimizer module; the input data includes first data and second data, the first data includes a second Hamiltonian, a quantum circuit type, and quantum gate information provided by a first user, and the second data includes an optimizer type, a number of optimization iterations, and optimization constraints provided by the first user.

[0085] Constructing a first workflow task includes: taking the first data as input and calling the quantum circuit module; taking the second data and the output of the quantum circuit module as input and calling the optimizer module; the task output is the optimized quantum state parameters and the minimized expected value of the second Hamiltonian.

[0086] Based on the above operations, using Figure 1 The task execution unit receives the workflow tasks and computing resources assigned by the task scheduling unit, calls multiple modules according to a predetermined logic and configuration, and executes multiple subtasks included in the first workflow task in parallel according to a predetermined code logic, and finally obtains a result.

[0087] Step S203: Schedule and execute the first workflow task to obtain the task output, which is the final state or intermediate state result of the variational quantum eigenstate solution system.

[0088] In one embodiment, the final state result is the final execution result of the quantum eigenstate solution task, and the intermediate state result is the intermediate execution result of the quantum eigenstate solution task. The final state or intermediate state result is ultimately used to solve the variational quantum eigenstate.

[0089] Exemplarily, when executing a first workflow task, a stop execution instruction from a first user or a second user and a second configuration operation are received; according to the second configuration operation, the configuration of the module not specified in the first configuration operation and the calling of the module interface, as well as the disconnection of the module interface of the module in the module specified by the first configuration operation, or the update of input data are performed to construct a second workflow task; and the execution of the second workflow task is scheduled.

[0090] For example, during the execution of the first workflow task, the first user or the second user observes the final state result or the intermediate state result through a visualization observation tool, and concludes that the iterative change of the Hamiltonian measurement value is small, that is, it is close to the minimum. Then, there is no need to wait for the entire first workflow task to be completed, and an instruction to stop executing the first workflow task can be issued, and a second configuration operation can be performed, such as reducing the number of iterations in the optimizer module and constructing a second workflow task. By rescheduling the execution of the second workflow task, compared with the first workflow task, the solution of the variational quantum eigenstate can be completed as quickly as possible without reducing the calculation accuracy to obtain the solution result.

[0091] Therefore, on the one hand, users can calculate the workflow tasks consisting of a certain module alone or a combination of some modules according to the preset tasks, and optimize a certain module or part of the modules according to the calculation results, thereby reducing the dependency between the modules in the entire variational quantum eigenstate solution system and improving the flexibility of the variational quantum eigenstate solution system and the scalability of the quantum algorithm; on the other hand, a certain module or part of the modules can be optimized in a targeted manner to eliminate the interference of other modules, thereby making the tuning process simpler and helping to find the global optimal solution.

[0092] The following will take the preset task "In the field of quantum chemistry, for N hydrogen molecules, assume that the preset task is to find the Hamiltonian corresponding to the most stable hydrogen molecule" as an example to introduce the workflow tasks of the variational quantum eigenstate solving system.

[0093] For example, Figure 3a A workflow task diagram of a variational quantum eigenstate solution system provided in an embodiment of the present application is shown.

[0094] like Figure 3a As shown, in the front-end interface of the task control system, the first user specifies a single Hamiltonian module by dragging, clicking, or entering in a dialog box according to a preset task, adds one of the N hydrogen molecules by clicking the "ADD" button, and sets the spatial structure and bond length of the hydrogen molecule by clicking the "EDIT" button, thereby obtaining a first configuration operation.

[0095] In the back end of the task control system, the task definition and configuration unit configures a single Hamiltonian module of the variational quantum eigenstate solution system according to the first configuration operation, and calls the interface of the single Hamiltonian module to construct a first workflow task.

[0096] Click the "START" button to start the calculation of the first workflow task. Specifically, in the back end of the task control system, the task management unit, the task scheduling unit, and the task execution unit complete the resource allocation and scheduling execution of the first workflow task, and obtain the intermediate state result of the variational quantum eigenstate solution system, that is, the Hamiltonian constructed using the hydrogen molecule.

[0097] In the back end of the mission control system, the data processing unit parses the output data of a single Hamiltonian module, and uses the visualization observation tool "GraphShow" to display the properties related to the graph structure in the Hamiltonian constructed by the hydrogen molecule, and "TextView" to display the properties related to the performance description in the Hamiltonian for users to observe in real time.

[0098] The user can repeatedly perform the first workflow task creation and calculation for N hydrogen molecules, and output the Hamiltonians constructed for the N hydrogen molecules. Then, according to the Hamiltonian performance displayed by "GraphShow" and / or "TextView", the user can select the hydrogen molecule with the lowest molecular potential energy surface from the N hydrogen molecules, that is, the Hamiltonian established by the most stable molecule as the calculation result of the preset task.

[0099] In addition, in the backend of the task control system, the task monitoring unit also records the progress of the first workflow task, resource usage and possible exceptions, and records this information in detail in the log, which is provided to the user for subsequent review and optimization.

[0100] Obviously, compared with the entire calculation process of the variational quantum eigenstate solution system, the calculation result of this preset task is an intermediate state result, which can be seamlessly connected to the Hamiltonian data of any other system or quantum circuit module.

[0101] In addition, the button "DELETE" is used to delete the added hydrogen molecules, and the button "STOP" is used to pause the Hamiltonian module calculation.

[0102] For example, Figure 3b A workflow task diagram of a variational quantum eigenstate solution system provided in an embodiment of the present application is shown.

[0103] like Figure 3b As shown, in the front-end interface of the task control system, the first user specifies N Hamiltonian modules by dragging, clicking, or entering in a dialog box according to a preset task, and for each Hamiltonian module, adds one of the N hydrogen molecules by clicking the "ADD" button, and sets the spatial structure and bond length of the hydrogen molecule by clicking the "EDIT" button, thereby obtaining a first configuration operation.

[0104] In the back end of the task control system, the task definition and configuration unit configures N Hamiltonian modules of the variational quantum eigenstate solution system according to the first configuration operation, and calls the interface of a single Hamiltonian module to construct a first workflow task including N subtasks.

[0105] Click the "START" button to start the calculation of the first workflow task. In the back end of the task control system, the task management unit, task scheduling unit, and task execution unit complete the resource allocation and parallel scheduling execution of the N subtasks in the first workflow task, and obtain the intermediate state results of the N variational quantum eigenstate solution system, that is, the N Hamiltonians constructed using N hydrogen molecules.

[0106] In the back end of the mission control system, the data processing unit parses the i-th (1≤i≤N) Hamiltonian module, and uses the visualization observation tool "GraphShowi" to display the properties related to the graph structure in the Hamiltonian constructed by the hydrogen molecule, and "TextViewi" to display the properties related to the performance description in the Hamiltonian for users to observe in real time.

[0107] The user can select the Hamiltonian performance displayed by "GraphShowi" and / or "TextViewi" and choose the hydrogen molecule with the lowest molecular potential energy surface from N hydrogen molecules, that is, the Hamiltonian established by the most stable molecule as the calculation result of the preset task.

[0108] Obviously, compared to Figure 3a The single solution shown in Figure 3b The parallel solution method shown in can better utilize the computing resources of the task control system and improve the computing efficiency, thereby improving the solution performance of the variational quantum eigenstate solution system.

[0109] For example, Figure 4 The control execution device of a quantum eigenstate solution task provided by an embodiment of the present application is shown in the figure. The solution task is performed using a variational quantum eigenstate solution system simulated by classical computing. The variational quantum eigenstate solution system includes multiple modules designed and decoupled based on a workflow architecture. The multiple modules include a Hamiltonian module for constructing a Hamiltonian, a quantum circuit module for simulating a quantum circuit, and an optimizer module for performing quantum state parameter iteration. Figure 4 As shown, the control execution device 400 includes:

[0110] The receiving module 410 is used to receive a first configuration operation of a first user, where the first configuration operation at least specifies at least one module among the multiple modules and input data.

[0111] The construction module 420 is used to call the interface of at least one module according to the input data, so as to construct the first workflow task.

[0112] The scheduling execution module 430 is used to schedule and execute the first workflow task to obtain a task output, where the task output is a final state or intermediate state result of the variational quantum eigenstate solution system.

[0113] Based on the method in the above embodiment, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program runs on a processor, the processor executes the method shown in the above embodiment.

[0114] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0115] The professionals should also be further aware that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0116] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0117] 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 method 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 within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A control execution method for a quantum eigenstate solution task, wherein the solution task is performed using a variational quantum eigenstate solution system simulated by classical computing, wherein the variational quantum eigenstate solution system comprises a plurality of modules designed and decoupled based on a workflow architecture, wherein the plurality of modules comprises a Hamiltonian module for constructing a Hamiltonian, a quantum circuit module for simulating a quantum circuit, and an optimizer module for performing quantum state parameter iteration; The method is performed by a task control system, and the method includes: receiving a first configuration operation of a first user, wherein the first configuration operation at least specifies at least one module among the plurality of modules and input data; Calling the interface of the at least one module according to the input data, thereby constructing a first workflow task; The first workflow task is scheduled to be executed to obtain a task output, where the task output is a final state or intermediate state result of the variational quantum eigenstate solution system.

2. According to the method of claim 1, the task control system has a front-end interface, the front-end interface includes user interaction units corresponding to multiple modules of the variational quantum eigenstate solution system, and the first configuration operation includes: Select at least one user interaction unit corresponding to the at least one module by dragging, clicking or dialog box input; Input data is set in an interface corresponding to the at least one user interaction unit.

3. The method according to claim 1, wherein the first workflow task comprises a plurality of subtasks, each subtask is executed based on the at least one module, and different subtasks have different input data; and the scheduling execution of the first workflow task comprises: Based on the different input data, the multiple subtasks are executed in parallel.

4. The method according to claim 1, wherein: The at least one module is a single module; the construction of the first workflow task includes: taking the input data as input, calling the interface of the single module; the output of the single module is used as the task output; Wherein, the single module is a Hamiltonian module, the input data is the quantum system data provided by the first user; the task output is the first Hamiltonian determined according to the quantum system data; or, The single module is a quantum circuit module, the input data includes the second Hamiltonian, quantum circuit type and quantum gate information provided by the first user; the task output is a first parameterized quantum state generated by a quantum circuit simulated by the quantum circuit type and quantum gate, and an expected value of the second Hamiltonian calculated based on the first parameterized quantum state; or, The single module is an optimizer module, the input data includes the second parameterized quantum state provided by the first user and the expected value of the third Hamiltonian calculated based on the second parameterized quantum state, as well as the optimizer type, number of optimization iterations and optimization constraints provided by the first user; the task output is the optimized quantum state parameters and the minimized expected value of the third Hamiltonian.

5. The method according to claim 4, wherein the variational quantum eigenstate solution system is applied in the field of quantum chemistry or materials science; The quantum system data includes the molecule type provided by the first user, and the spatial structure and bond length of the molecule.

6. The method according to claim 1, wherein: The at least one module is the Hamiltonian module and the quantum circuit module; the input data includes quantum system data provided by the first user, as well as quantum circuit type and quantum gate information; Construct the first workflow task, including: Taking the quantum system data as input, calling the Hamiltonian module; Taking the quantum circuit type, quantum gate information, and the target Hamiltonian output by the Hamiltonian module as input, calling the quantum circuit module; The task output is a first parameterized quantum state generated by a quantum circuit simulated according to the quantum circuit type and quantum gate, and an expectation value of the target Hamiltonian calculated according to the first parameterized quantum state.

7. The method according to claim 1, wherein: The at least one module is the quantum circuit module and the optimizer module; the input data includes first data and second data, the first data includes the second Hamiltonian, quantum circuit type and quantum gate information provided by the first user, and the second data includes the optimizer type, optimization iteration number and optimization constraint condition provided by the first user; Construct the first workflow task, including: Taking the first data as input, calling the quantum circuit module; taking the second data and the output of the quantum circuit module as input, calling the optimizer module; The task outputs are optimized quantum state parameters and minimized expectation values ​​of the second Hamiltonian.

8. The method according to claim 1, further comprising: When executing the first workflow task, receiving a stop execution instruction and a second configuration operation from the first user or the second user; According to the second configuration operation, configuration of the module not specified by the first configuration operation and calling of the module interface, as well as disconnection of the module interface of the module specified by the first configuration operation or update of input data are performed to construct a second workflow task; The second workflow task is scheduled for execution.

9. The method according to claim 1, wherein: The task control system also includes a visualization observation tool; the first configuration operation also includes connecting the visualization observation tool to the at least one module for real-time observation of the calculation results of the at least one module during the solution process.

10. A control execution device for a quantum eigenstate solution task, wherein the solution task is performed using a variational quantum eigenstate solution system simulated by classical computing, wherein the variational quantum eigenstate solution system comprises a plurality of modules designed and decoupled based on a workflow architecture, wherein the plurality of modules comprise a Hamiltonian module for constructing a Hamiltonian, a quantum circuit module for simulating a quantum circuit, and an optimizer module for performing quantum state parameter iteration; the device comprises: A receiving module, configured to receive a first configuration operation of a first user, wherein the first configuration operation at least specifies at least one module among the plurality of modules and input data; A construction module, configured to call an interface of the at least one module according to the input data, thereby constructing a first workflow task; The scheduling execution module is used to schedule the execution of the first workflow task to obtain a task output, where the task output is a final state or an intermediate state result of the variational quantum eigenstate solution system.

Images