Task scheduling execution method and system based on hybrid computing environment

Abstract

The invention provides a task scheduling execution method and system based on a hybrid computing environment. The task scheduling execution method comprises the steps of receiving a task, performing task type division and quantum coding on the task to form quantum coding information for representing the task, and determining an initial quantum state of the task; evaluating the quantum enhancement priority of the task based on the quantum coding information, and converting a quantum enhancement priority evaluation result of the task into task priority information for a classical task scheduling execution method; and scheduling and executing the task based on the task priority information. According to the method, the dual advantages of fusion quantum computing and classical computing are utilized, effective management of complex tasks is achieved, the isomerized computing environment and task requirements are met, the problem of low-efficiency transfer caused by insufficient integration in the current hybrid computing environment is solved, and the efficiency is improved. And the comprehensive performance, the resource use efficiency and the task scheduling capability of the hybrid computing environment are effectively improved.

Description

Technical Field

[0001] The present disclosure relates to the field of quantum computing, and particularly to a task scheduling and execution method and system based on a hybrid computing environment. Background Art

[0002] With the booming development of fields such as artificial intelligence, big data analysis, and scientific simulation, the types and complexities of computing tasks have shown great diversity. The high heterogeneity of computing resources provides opportunities to improve the processing efficiency of specific tasks, while also increasing the complexity of resource management and task scheduling, and posing higher requirements for energy efficiency. Optimizing energy use while ensuring computing performance has become an important consideration in the design of modern computing systems.

[0003] The rapid development of quantum computing technology provides new ideas for solving some classical computing problems. However, how to effectively integrate quantum computing resources into the existing computing environment and how to perform efficient task scheduling in a hybrid computing environment remain an open research field. In such a technical background, traditional task scheduling and execution methods are difficult to meet the needs of modern computing environments. Traditional scheduling methods lack the ability to handle highly heterogeneous resources, are difficult to effectively manage complex task dependencies, and cannot fully utilize emerging quantum computing resources. For example, the scheduling method based on classical machine learning has the advantages of being able to handle complex patterns, being highly adaptable, and can be implemented on existing hardware, but it lacks the use of quantum computing to improve the scheduling rate and has problems with low scheduling efficiency when dealing with very large task sets. Another example is the distributed quantum-inspired task scheduling and execution method, which has the advantage of being able to be implemented on existing hardware, but still has the problem of not being able to achieve complete quantum acceleration. The feasibility of this technology is relatively strong, but the system performance still does not solve the problem of high-efficiency scheduling of complex tasks.

[0004] The existing technology still has problems such as insufficient integration of quantum computing and classical computing, lack of quantum-enhanced decision-making ability, inefficient priority evaluation in a hybrid computing environment, limited heterogeneous resource coordination ability, and insufficient utilization of quantum computing. Therefore, there is a need for a task scheduling and execution method and system that can make full use of quantum computing resources and effectively integrate quantum and classical computing resources to achieve high-efficiency scheduling of complex tasks. Summary of the Invention

[0005] In view of the problems existing in the prior art, the present disclosure provides a task scheduling and execution method and system based on a hybrid computing environment, which solves the problem of low-efficiency scheduling caused by insufficient integration in the current hybrid computing environment, and effectively improves the comprehensive performance, resource utilization efficiency, and task scheduling ability of the hybrid computing environment.

[0006] In the first aspect of the present disclosure, a task scheduling and execution method based on a hybrid computing environment is provided. The task scheduling and execution method includes:

[0007] Receive a task, classify the task type and perform quantum encoding on the task to form quantum encoding information representing the task, and determine the initial quantum state of the task;

[0008] Evaluate the quantum enhancement priority of the task based on the quantum encoding information, and convert the evaluation result of the quantum enhancement priority of the task into task priority information for the classical task scheduling execution method;

[0009] Schedule and execute the task based on the task priority information.

[0010] Optionally, evaluating the quantum enhancement priority of the task based on the quantum encoding information includes:

[0011] Superimpose the initial quantum states of all the tasks to form a quantum superposition state containing the quantum encoding information of all the tasks.

[0012] Optionally, evaluating the quantum enhancement priority of the task based on the quantum encoding information further includes:

[0013] Determine a priority evaluation function through a quantum gate and perform quantum parallel evaluation based on the quantum state of the task;

[0014] Enhance the amplitude of the quantum state of high-priority tasks and reduce the amplitude of the quantum state of low-priority tasks through quantum interference;

[0015] Perform quantum measurement to collapse the quantum state of the task after quantum interference to a classical bit state;

[0016] Post-process the priority result of the task determined by quantum measurement using the classical task scheduling execution method to generate task priority information for the classical task scheduling execution method.

[0017] Optionally, the influencing factors of the priority evaluation function include at least one of task execution time, resource requirements, task dependency, and task urgency.

[0018] Optionally, classifying the task type and performing quantum encoding on the task to form quantum encoding information representing the task and determining the initial quantum state of the task includes:

[0019] Classify the task into a quantum computing task, a classical computing task, or a hybrid task of quantum computing and classical computing according to the task type;

[0020] Match the corresponding quantum encoding to the task according to the task type of each task to form the quantum encoding information and determine the initial quantum state.

[0021] The matching of the corresponding quantum encoding for the tasks according to the task types of the respective tasks includes at least one of the following:

[0022] Encoding the quantum computing tasks in quantum states;

[0023] Encoding the classical computing tasks in quantum superposition states;

[0024] Encoding the hybrid tasks of quantum computing and classical computing in quantum states and quantum superposition states simultaneously.

[0025] Optionally, the scheduling and execution of the tasks based on the task priority information includes:

[0026] Receiving the task priority information, allocating resources correspondingly according to the task priority information, and classifying and executing the tasks according to the task types;

[0027] Generating a task processing statistical report by analyzing the task priority information.

[0028] Optionally, the task scheduling and execution method further includes monitoring and feedback, where the monitoring and feedback continuously monitors the execution status of the tasks and dynamically adjusts the resource allocation of the tasks based on the execution status of the tasks.

[0029] The second aspect of the present disclosure provides a task scheduling and execution system based on a hybrid computing environment, including:

[0030] A task management module that divides the tasks into task types and performs quantum encoding, forms quantum encoding information for representing the tasks, and determines the initial quantum states of the tasks;

[0031] A quantum-enhanced scheduling decision module that evaluates the quantum-enhanced priorities of the tasks based on the quantum encoding information and converts the evaluation results of the quantum-enhanced priorities of the tasks into task priority information for the classical task scheduling and execution method;

[0032] A task scheduling module that schedules and executes the tasks based on the task priority information.

[0033] The third aspect of the present disclosure provides a computer-readable storage medium that stores a computer program for running the task scheduling and execution method based on a hybrid computing environment, where the computer program causes a computer to execute the task scheduling and execution method as described above.

[0034] The task scheduling and execution method and system based on a hybrid computing environment in the present disclosure have the following advantages:

[0035] (1) Through task management, quantum-enhanced scheduling decision-making, resource management, and optimization, a task scheduling and execution system based on a hybrid computing environment is formed. Among them, task management is used to receive and initialize tasks, quantum-enhanced scheduling decision-making is used to evaluate task priorities and make scheduling decisions, and resource management and optimization are used to implement execution decisions and manage computing resources, solving the problem that existing scheduling systems are difficult to effectively integrate and manage quantum computing resources and classical computing resources, and effectively improving the comprehensive performance, resource utilization efficiency, and task scheduling ability of the hybrid computing environment.

[0036] (2) The core introduces quantum-enhanced scheduling decision-making. By extracting priority-related indicators in tasks, a priority function is determined, and a quantum-enhanced scheduling decision-making module (QEPA: Quantum-Enhanced Priority Assessment) and logical method are designed and implemented. This module / method can execute a large number of complex task information within an extremely short time scale, assign reasonable priorities to each task, and ensure the optimal utilization of system resources.

[0037] (3) Based on the priority evaluation function, by taking advantage of batch processing such as quantum parallel processing and superposition state characteristics, an efficient and accurate task scheduling scheme is determined, which is particularly suitable for processing large-scale and complex-dependent task sets. At the same time, by making full use of quantum characteristics and using task scheduling and execution methods such as quantum entanglement and interference, quantum characteristic optimization scheduling is realized, which can construct an optimal scheduling scheme in a larger solution space, giving full play to the advantages of quantum computing while optimizing the overall efficiency of the system.

[0038] (4) By dividing tasks into quantum computing tasks, classical computing tasks, and hybrid tasks, corresponding quantum state encoding is performed and resource scheduling and management strategies are adaptively designed, endowing the task scheduling and execution system / scheme with stronger heterogeneous resource management capabilities to ensure the coordinated and efficient integrated control of quantum processors and various classical computing resources. Through encoding quantum computing tasks with the quantum state |ψ>, encoding classical computing tasks with the quantum superposition state | >, and encoding hybrid tasks with a dual encoding method, the quantum state representation of various types of tasks is realized.

[0039] (5) Further supplemented by a monitoring and feedback module, the resource utilization of the overall execution queue is monitored in real time to ensure that the computing resources of the entire system meet the execution requirements of the current tasks, and can be flexibly adjusted according to the execution situation to ensure the maximum utilization of resources and further improve the system scheduling and execution efficiency.

[0040] (6) By monitoring information such as resource utilization and queue execution status, adaptively adjusting resource allocation, a scheduling mechanism based on dynamic adaptation to the resource status and task characteristics in a hybrid computing environment is realized, significantly enhancing the flexible response ability and operation efficiency of the system / method; at the same time, by dynamically allocating quantum computing and classical computing resources, the flexibility of the system / method is greatly improved, and it can better adapt to heterogeneous computing environments and task requirements.

[0041] (7) The system / solution of the present disclosure covers the entire workflow from task reception, quantum state preparation to classical post-processing of quantum measurement results and then to task execution, realizing the efficient integration of quantum tasks and classical tasks, and having more profound guiding significance for the implementation and optimization of actual systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a flowchart of the task scheduling and execution method in the present disclosure;

[0043] Figure 2 is a flowchart of the work process of quantum-enhanced priority evaluation in the present disclosure;

[0044] Figure 3 is a flowchart of the work process of task management in the present disclosure;

[0045] Figure 4 is a task dependency logic diagram in the present disclosure;

[0046] Figure 5 is a flowchart of the work process of task scheduling and execution in the present disclosure;

[0047] Figure 6 is a flowchart of the work process of task scheduling, execution and monitoring feedback in the present disclosure;

[0048] Figure 7 is a block diagram of the task scheduling and execution system in the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] Next, the technical solutions in the embodiments of the present disclosure will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present disclosure without creative efforts shall fall within the protection scope of the present disclosure.

[0050] Based on the problems existing in the prior art, such as insufficient integration of quantum and classical resources, lack of quantum-enhanced decision-making ability, inefficient priority evaluation in a hybrid computing environment, and limited heterogeneous resource coordination ability.

[0051] Such as Figure 1As shown in the figure, the first aspect of the present disclosure provides a task scheduling and execution method based on a hybrid computing environment. The task scheduling and execution method includes:

[0052] Step 1, task management includes: receiving a task, classifying the task by task type and performing quantum encoding to form quantum encoding information representing the task, and determining the initial quantum state of the task.

[0053] Step 2, quantum-enhanced priority evaluation includes: evaluating the quantum-enhanced priority of the task based on the quantum encoding information, and converting the evaluation result of the quantum-enhanced priority of the task into task priority information for a classical task scheduling and execution method.

[0054] Step 3, task scheduling and execution includes: scheduling and executing the task based on the task priority information.

[0055] The present disclosure utilizes the dual advantages of integrating quantum computing and classical computing to achieve effective management of complex tasks, meet the heterogeneous computing environment and task requirements, solve the problem of low-efficiency mobilization caused by insufficient integration in the current hybrid computing environment, and effectively improve the comprehensive performance, resource utilization efficiency, and task scheduling ability of the hybrid computing environment.

[0056] In some embodiments of the present disclosure, evaluating the quantum-enhanced priority of the task based on the quantum encoding information includes:

[0057] Superposing the initial quantum states of all tasks to form a quantum superposition state containing the quantum encoding information of all tasks.

[0058] In some embodiments of the present disclosure, evaluating the quantum-enhanced priority of the task based on the quantum encoding information further includes:

[0059] Determining a priority evaluation function through a quantum gate and performing quantum parallel evaluation based on the quantum state of the task;

[0060] Enhancing the amplitude of the quantum state of high-priority tasks and reducing the amplitude of the quantum state of low-priority tasks through quantum interference;

[0061] Performing quantum measurement to collapse the quantum state of the task after quantum interference to a classical bit state;

[0062] Using a classical task scheduling and execution method to post-process the priority result of the task determined by the quantum measurement to generate task priority information for the classical task scheduling and execution method.

[0063] Specifically, as Figure 2 shown, the quantum-enhanced priority evaluation in step 2 includes:

[0064] Step 201, quantum state superposition includes: superposing the initial quantum states of all tasks to form a quantum superposition state containing the quantum encoded information of all tasks;

[0065] In some embodiments of the present disclosure, a Hadamard gate (H) is used to form a quantum superposition state containing the quantum encoded information of all tasks.

[0066] Step 202, quantum parallel evaluation includes: determining the priority evaluation function Q(|ψ>) through quantum gate operations, and simultaneously performing parallel evaluation on all tasks based on the quantum states of the tasks using quantum computing methods;

[0067] In some embodiments of the present disclosure, the priority evaluation function Q(|ψ>) is determined through quantum gate operations. Optionally, a multi-controlled quantum gate is used to implement the priority evaluation function Q(|ψ>), and CNOT gates and Toffoli gates are used to determine the conditional logic in the implementation process of the priority evaluation function Q(|ψ>), and a conditional operation circuit is constructed through CNOT gates and Toffoli gates to implement it. Specifically, a CNOT gate is used to implement the state transition of a single condition, a Toffoli gate (CCX) is used to implement the state transition of a double condition, and a complex conditional judgment circuit is constructed by combining multiple CNOT gates and Toffoli gates to determine the priority evaluation function Q(|ψ>).

[0068] In some embodiments of the present disclosure, the influencing factors of the priority evaluation function include at least one of task execution time, resource requirements, task dependencies, and task urgency.

[0069] The above influencing factors are represented by independent unitary matrices in the priority evaluation function Q(|ψ>). When evaluating the task priority, the unitary matrices of task execution time, resource requirements, task dependencies, and task urgency are considered to determine the priority evaluation function, ensuring the reliability of the priority evaluation result.

[0070] In some embodiments of the present disclosure, the priority evaluation function Q(|ψ>) may include the following form: Q(|ψ>) = Σ(i = 1 to n)α i U i (x i ); where x i represents different factors affecting the task priority. In addition to the task execution time (t), resource requirements (r), dependencies (d), and urgency (u), it may also include the importance weight (w) of the task, historical execution success rate (s), energy consumption (e), waiting time (q), and other influencing factors, which can be specifically set according to task requirements.

[0071] Step 203, quantum interference includes: using the quantum interference effect to optimize the quantum state amplitudes of tasks with different priorities. Specifically, enhance the amplitude of the quantum state of high-priority tasks and reduce the amplitude of the quantum state of low-priority tasks.

[0072] In some embodiments of the present disclosure, in step 203, quantum interference realizes the optimization of the quantum state amplitudes of tasks with different priorities through quantum gate operations, such that the quantum state components representing high-priority tasks undergo constructive interference, while the quantum state components of low-priority tasks undergo destructive interference. For example, quantum gates such as controlled-phase gates, controlled-rotation gates, etc. can be used to achieve quantum interference, and the operation parameters of the quantum gates can be set according to the characteristics of the tasks (such as urgency, resource requirements, etc.). After the quantum interference operation, the amplitude of the quantum state of high-priority tasks is enhanced, while the amplitude of the quantum state of low-priority tasks is weakened. The task scheduling execution method of the present disclosure can utilize quantum computing to simultaneously process the priorities of all tasks, and use quantum parallelism to accelerate the task priority evaluation process, improving the quantum-enhanced scheduling decision-making efficiency.

[0073] In some embodiments of the present disclosure, the quantum gate operations include at least one of the following ways:

[0074] Multi-controlled NOT gate (MCX), used to achieve state flipping under multiple control conditions.

[0075] Multi-controlled phase gate (MCP), used to adjust the phase of the quantum state according to multiple conditions.

[0076] Controlled-swap gate (CSWAP), used to conditionally swap the state parameters of two qubits.

[0077] Rotation gates (Rx, Ry, Rz), used to precisely adjust the amplitude and phase of the quantum state.

[0078] Step 204, quantum measurement includes: performing quantum measurement to collapse the quantum state of the task after quantum interference into a classical bit state, where the classical bit state is the bit state in a classical computer, i.e., 0 and 1.

[0079] In some embodiments of the present disclosure, when performing quantum measurement in step 204, the task priority information is encoded into the amplitude of the quantum state. The quantum state components corresponding to high-priority tasks correspond to larger amplitudes, so that in the result of quantum measurement, it is manifested as a high frequency of the state appearance, to determine the priority distribution of all tasks. In the result of quantum measurement, the state with the highest frequency of appearance corresponds to the highest-priority task, the second-highest frequency corresponds to the second-priority task, and so on, to achieve the priority sorting of all tasks and determine the priorities of all tasks.

[0080] For example, there are three tasks A, B, and C. After the quantum state superposition in step 201, the task state is |ψ>=0.6|A>+0.3|B>+0.1|C>. After multiple quantum measurements, the following results are obtained: A is 60%, B is 30%, and C is 10%. This reflects the priority order of the tasks as A > B > C.

[0081] Step 205, post-processing includes: using a classical task scheduling execution method to post-process the task priority results determined by quantum measurement to generate task priority information for classical task scheduling execution, where the presentation form of the task priority information includes, but is not limited to, a task priority list. In some embodiments of the present disclosure, the task priority information includes a priority phase value, eigenvalue information, priority sorting, priority distribution, priority accuracy, quantum state superposition information, and priority weight.

[0082] The present disclosure performs quantum parallel evaluation and quantum interference on the quantum states of tasks through quantum-enhanced priority evaluation to achieve quantum enhancement, enabling tasks with high priorities to be identified in the form of high-frequency occurrences, generating task priority information and efficiently executing it, and effectively improving the efficiency and quality of scheduling decisions using quantum computing.

[0083] In some embodiments of the present disclosure, the quantum parallel evaluation in step 202 includes the following steps:

[0084] Use quantum computing to simultaneously perform parallel evaluation on all tasks, and determine the priority evaluation function Q(|ψ>) through quantum gate operations;

[0085] Evaluate the task priority, and the priority evaluation function Q(|ψ>) is represented by the following relational expression:

[0086] Q(|ψ>)=α1U1(t)+α2U2(r)+α3U3(d)+α4U4(u).

[0087] Where |ψ> is the quantum state of the task, α1, α2, α2, α4 are weight coefficients, U1(t) is a unitary matrix related to the task execution time (t), U2(r) is a unitary matrix related to the resource requirement (r), U3(d) is a unitary matrix related to the task dependency relationship (d), and U4(u) is a unitary matrix related to the task urgency (u).

[0088] The above four unitary matrices include:

[0089] U1(t)=exp(-iH1t), where H1 is the Hamiltonian related to the task execution time, and the time-dependent phase evolution is implemented using CNOT and rotation gates.

[0090] U2(r) = exp(-iH2r), where H2 is the Hamiltonian related to resource requirements, and the Toffoli gate is used to implement the multi-condition judgment of resource requirements.

[0091] U3(d) = exp(-iH3d), where H3 is the Hamiltonian related to task dependencies, and the controlled-swap gate (CSWAP) is used to implement the state swap of dependencies.

[0092] U4(u) = exp(-iH4u), where H4 is the Hamiltonian related to task urgency, and the multi-controlled phase gate (MCP) is used to implement the phase adjustment of urgency.

[0093] In some embodiments of the present disclosure, the weight coefficient α i is determined through the following steps:

[0094] Initialization, including setting initial weights according to empirical values;

[0095] Normalization, including ensuring Σα i = 1;

[0096] Dynamic adjustment, including evaluating the importance of each factor according to historical task execution results and optimizing the task scheduling execution method to adjust the weights; for example, using gradient descent to optimize the task scheduling execution method and adjust the weights;

[0097] Weight verification, verifying the rationality of the weights through test data.

[0098] In some embodiments of the present disclosure, evaluating task priorities includes the following steps:

[0099] Parallel computing of the priority scores of each task, including: inputting the quantum states of all tasks into the priority evaluation function simultaneously and using quantum parallelism to calculate multiple tasks simultaneously;

[0100] Priority level division, including dividing priority levels according to the score range, where the priority levels include high priority (score ≥ 0.8), medium priority (0.4 ≤ score < 0.8), and low priority (score < 0.4);

[0101] Priority marking, including marking the corresponding priority level for each task and recording the priority score;

[0102] Result integration, including generating a task state containing priority information and entering quantum interference.

[0103] By constructing a priority evaluation function, determining evaluation criteria, and performing the priority evaluation of each task, a clear priority division result can be obtained, making the priority evaluation process more refined and the priority evaluation result more accurate.

[0104] In some embodiments of the present disclosure, steps 201 to 204 adopt a quantum computing processing method, and step 205 adopts a classical computing processing method to ensure that the quantum-enhanced decision-making results can be used in a classical computer environment.

[0105] In some embodiments of the present disclosure, task type division and quantum encoding are performed on a task to form quantum encoding information representing the task, and determining the initial quantum state of the task includes:

[0106] Dividing the task into a quantum computing task, a classical computing task, or a hybrid task of quantum computing and classical computing according to the task type;

[0107] According to the task type of each task, matching the corresponding quantum encoding to the task, forming quantum encoding information, and determining the initial quantum state.

[0108] In some embodiments of the present disclosure, matching the corresponding quantum encoding to the task according to the task type of each task includes at least one of the following:

[0109] Encoding the quantum computing task in a quantum state;

[0110] Encoding the classical computing task in a quantum superposition state;

[0111] Encoding the hybrid task of quantum computing and classical computing in a quantum state and a quantum superposition state simultaneously.

[0112] Specifically, as Figure 3 shown, the task management in step 1 includes:

[0113] Step 101, task reception includes: the system runs and receives a computing task, and the task source includes but is not limited to user-submitted tasks, system-generated tasks automatically, data-driven tasks, real-time data stream processing tasks; the present disclosure does not make a special limitation on the task source.

[0114] Step 102, task classification includes: dividing the task into a quantum computing task, a classical computing task, or a hybrid task of quantum computing and classical computing according to the task type.

[0115] Step 103, task representation includes: according to the task type of each task, extracting the information features of the task, matching the corresponding quantum encoding to the task, forming quantum encoding information representing the task, and determining the initial quantum state.

[0116] Among them, matching the corresponding quantum encoding to the task includes: 1) encoding the quantum computing task in a quantum state |ψ>; 2) encoding the classical computing task in a quantum superposition state | >Perform encoding; 3) Encode the hybrid tasks of quantum computing and classical computing in the quantum state |ψ> and the quantum superposition state | > simultaneously. Among them, rotation gates (Rx, Ry, Rz) are used to implement the quantum encoding corresponding to the task matching.

[0117] In some embodiments of the present disclosure, the quantum encoding information includes features related to task priorities. Specifically, the features related to task priorities include the execution time, resource requirements, dependencies, urgency, etc. of the tasks.

[0118] In some embodiments of the present disclosure, for quantum computing tasks, the extracted information features include at least one of qubit requirements, gate operation complexity, depth of quantum computing method, quantum error correction requirements, quantum state retention time, number of quantum measurements, and interaction requirements with classical computing.

[0119] Specifically, the qubit requirements include the number of qubits required to execute a specific task. The gate operation complexity includes the number or type of quantum gate operations required to implement the task scheduling execution method. The depth of the quantum computing method includes the number of layers of logic gate operations from input to output in the quantum computing method. The quantum error correction requirements include the degree of dependence of the task scheduling execution method on quantum error correction, which can be represented by discrete values. For example, 0 represents low requirements and 1 represents high requirements. The number of quantum measurements includes the number of quantum measurements required during the execution of the task scheduling execution method. The number of quantum measurements can characterize the overall complexity and information loss amount of the task scheduling execution method to a certain extent. The interaction requirements with classical computing include the total execution time and resource consumption of the task scheduling execution method, which can be represented by discrete values.

[0120] In some embodiments of the present disclosure, for classical computing tasks, the extracted information features include at least one of processor requirements, memory requirements, estimated execution time, storage space requirements, I / O operation intensity, network bandwidth requirements, or encoding tools.

[0121] Specifically, the processor requirements include parameters such as the type of processor, the number of cores, and the clock frequency required to execute the computing tasks. The memory requirements include the maximum amount of memory required during the execution of the task scheduling and execution method. The estimated execution time includes the time required to estimate the task from start to finish based on the complexity of the task and the computing resources. The storage space requirements include the size of the storage space required during the execution of the task scheduling and execution method. The greater the storage space requirements, the higher the requirements for the capacity and read / write speed of the storage device. The I / O (input / output) operation intensity includes measuring the frequency and degree of input / output operations during the execution of the task scheduling and execution method. The network bandwidth requirements include the amount of network bandwidth required during the execution of the task scheduling and execution method. The encoding tools include information such as the software environment, encoded voice, and library files relied on by the task scheduling and execution method during encoding.

[0122] In some embodiments of the present disclosure, for hybrid tasks of quantum computing and classical computing, the extracted information features include at least one of quantum computing features (such as quantum computing task encoding), classical computing features (such as electronic computing task encoding), quantum and classical interaction frequencies, data conversion overheads, overall execution time estimates, and quantum and classical resource allocation ratios.

[0123] Specifically, the quantum computing task encoding includes the representation of quantum information such as quantum task scheduling and execution methods, quantum computing methods, and quantum states. The electronic computing task encoding includes the representation of classical computing information such as traditional programming languages, data structures, and task scheduling and execution methods. The quantum and classical interaction frequencies include the number of interactions and frequencies between quantum computing and classical computing in hybrid tasks. The data conversion overheads include the time and resources required to convert quantum data and classical data into each other. The overall execution time estimate includes the total time required for the hybrid task of quantum computing and classical computing from start to finish, specifically including quantum computing time, classical computing time, and data conversion and communication time, etc.

[0124] Step 104, task transmission includes: mapping the initial quantum state of each task to qubits using quantum amplitude encoding and transmitting it to the next step for processing. Among them, qubits are allocated for each task feature when performing task transmission in step 104; for example, there are n tasks, and each task contains m features, then a total of n*m qubits are required.

[0125] The present disclosure divides tasks into quantum computing tasks, classical computing tasks, or hybrid tasks of quantum computing and classical computing, matches different encoding methods to perform quantum state information processing accordingly, generates metadata and then makes quantum-enhanced scheduling decisions, improving the adaptability of the task scheduling and execution method, enabling the task scheduling and execution method to adapt to various modes and achieve high-efficiency execution.

[0126] In some embodiments of the present disclosure, before the task transmission in step 104, task dependency graph construction may also be included. The task dependency graph construction includes analyzing the dependency relationships between tasks in the system, constructing a directed acyclic graph (DAG) according to the dependency relationships, and the constructed directed acyclic graph can be encoded into the initial quantum state as part of the task features. For example, additional qubits can be used to represent the position of a task in the task dependency graph and its relationship with other tasks.

[0127] In some embodiments of the present disclosure, for quantum computing tasks, qubit strings are used as the unique identifier for each task; quantum entanglement is utilized to characterize the dependency relationships of quantum computing tasks.

[0128] In some embodiments of the present disclosure, for classical computing tasks, it includes using quantum amplitude encoding to characterize the dependency relationships of classical computing tasks, and mapping the classical adjacency matrix to the amplitudes of the quantum state.

[0129] In some embodiments of the present disclosure, for hybrid tasks of quantum computing and classical computing, it includes combining the above two methods, using a hybrid quantum state to characterize the dependency relationships, where for the quantum computing part in the hybrid quantum state, an entangled state is used, and for the classical computing part, amplitude encoding is used to determine the dependency relationships of the hybrid tasks of quantum computing and classical computing.

[0130] In some embodiments of the present disclosure, as Figure 4 shown, the task dependency logic graph includes 4 tasks (A, B, C, D), where A depends on B and C, and D depends on C; then the encoding process includes:

[0131] The task identifiers (represented by 2 qubits) include A: |00>, B: |01>, C: |10>, D: |11>.

[0132] The dependency relationships (represented by 6 qubits for possible dependency pairs) include |ψ_dep> = |010100> (representing the dependency relationships of A - B, A - C, and C - D).

[0133] The task types (represented by 2 qubits for each task's type) include |ψ_type> = |00011011> (assuming that A and B are both quantum computing tasks, C is a classical computing task, and D is a hybrid task of quantum computing and classical computing).

[0134] The complete quantum state includes

[0135] In some embodiments of the present disclosure, scheduling and executing tasks based on task priority information includes:

[0136] Receive task priority information, perform resource allocation according to the task priority information, and classify and execute tasks according to the task type;

[0137] Generate a task processing statistical report by analyzing the task priority information.

[0138] Specifically, as Figure 5 shown, the scheduling and execution of tasks in step 3 include:

[0139] Step 301, data input includes: receiving task priority information, extracting task execution-related feature information, classifying and storing the feature information. Among them, the task execution-related feature information may include task ID, priority value, quantum state information, etc.

[0140] Step 302, resource allocation includes: allocating different types of resources according to the data input results, including quantum resources, classical resources, and hybrid resources.

[0141] Among them, the resource allocation can refer to the following parameters:

[0142] Task priority, specifically including: dividing the tasks to be executed into high, medium, and low priorities according to the task priority, and clarifying the quantity distribution of each priority.

[0143] Task type, specifically including: dividing the tasks to be executed into quantum computing tasks, classical computing tasks, and hybrid tasks of quantum computing and classical computing according to the task type, and clarifying the quantity of each task type and the corresponding priority of each task type.

[0144] Step 303, classified execution includes: allocating tasks to different execution queues according to the task type for classified execution. The execution queues are divided into a quantum execution queue, a classical execution queue, and a hybrid execution queue according to the task type. Among them, the quantum execution queue includes those for processing quantum computing tasks; the classical execution queue includes those for processing classical computing tasks; the hybrid execution queue includes those for processing hybrid tasks of quantum computing and classical computing.

[0145] In some embodiments of the present disclosure, when executing tasks, high-priority tasks preferentially obtain computing resources, and tasks with the same priority are executed according to the first-come-first-served principle, and task preemption can be performed. Specifically, high-priority tasks can preempt the computing resources of low-priority tasks.

[0146] In some embodiments of the present disclosure, as Figure 5 shown, step 3 further includes step 304, result analysis, where the result analysis includes: determining the main indicators of the scheduling result according to the task priority information, analyzing the priority sorting and the scheduling result, and identifying the task features that affect the priority. The main indicators include average priority, result standard deviation, etc.

[0147] In some embodiments of the present disclosure, by performing a weighted average analysis on the task types and priority evaluation results of each task, the average priority of each task type is determined. For example, in a quantum-enhanced scheduling strategy with 10 tasks, there are 4 quantum computing tasks and 6 classical computing tasks; the priority distribution has 3 high-priority tasks, 5 medium-priority tasks, and 2 low-priority tasks; it is determined that the average priority of the quantum computing tasks is 75 points, and the average priority of the classical computing tasks is 50 points, which indicates that the task scheduling execution method preferentially processes quantum computing tasks.

[0148] Compared with the traditional method, the efficiency evaluation result obtained by using the task scheduling execution method in the present disclosure reduces the task completion time by 8%. Based on the efficiency evaluation result, it can be concluded that the quantum computing tasks are given higher priority, and the quantum computing resources can be further increased to meet the needs of high-priority tasks.

[0149] In some embodiments of the present disclosure, as Figure 6 shown, the task scheduling execution method further includes step 4, monitoring and feedback, where the monitoring and feedback includes: continuously monitoring the execution status of the task, and dynamically adjusting the resource allocation of the task based on the execution status of the task.

[0150] By monitoring the running status of the execution queue, the resource utilization rate of each execution queue is determined in real time. When the resource utilization rate of the execution queue is high, the corresponding computing resources are updated in real time; when the resource utilization rate of the execution queue is low, the computing resources of the execution queue are transferred to other execution queues that need to supplement computing resources to meet the execution needs of all tasks, effectively improving the efficiency of task scheduling execution.

[0151] The monitoring and feedback can also feedback the current task information, and transmit the monitored task information to the quantum-enhanced priority evaluation for re-evaluation, which is used for further task planning and resource adjustment to achieve high resource utilization efficiency and task scheduling ability.

[0152] As Figure 7 shown, the second aspect of the present disclosure provides a task scheduling execution system based on a hybrid computing environment. The task scheduling execution system includes:

[0153] A task management module, a quantum-enhanced scheduling decision module, and a task scheduling module. The task management module classifies the tasks by task type and performs quantum encoding to form quantum encoding information representing the tasks, and determines the initial quantum state of the tasks;

[0154] The quantum-enhanced scheduling decision module evaluates the quantum-enhanced priority of tasks based on quantum-encoded information and converts the evaluation result of the quantum-enhanced priority of tasks into task priority information for classical task scheduling execution methods;

[0155] The task scheduling module schedules and executes tasks based on the task priority information.

[0156] In some embodiments of the present disclosure, the quantum-enhanced scheduling decision module (QEPA) includes a quantum state superposition module, a quantum parallel evaluation module, a quantum interference module, a quantum measurement module, and a post-processing module.

[0157] The quantum-enhanced scheduling decision module performs a priority evaluation on tasks to generate a quantum-enhanced priority evaluation result; specifically, the quantum state superposition module superimposes the initial quantum states of all tasks to form a quantum superposition state containing the quantum-encoded information of all tasks; the quantum parallel evaluation module determines the priority evaluation function Q(|ψ>) through quantum gate operations and simultaneously performs a parallel evaluation on all tasks based on the quantum states of the tasks using quantum computing methods; the quantum interference module utilizes the quantum interference effect to optimize the quantum state amplitudes of tasks with different priorities, enhance the amplitudes of the quantum states of high-priority tasks, and reduce the amplitudes of the quantum states of low-priority tasks, so that high-priority tasks are identified in the form of high-frequency occurrences; the quantum measurement module collapses the quantum state with optimized quantum state amplitudes into a definite classical bit state to perform quantum measurement, and determines the priority order of each task according to the occurrence frequencies of each task in the classical bit state after multiple measurements; the post-processing module uses classical task scheduling execution methods to post-process the priority results of the tasks determined by quantum measurement to generate task priority information for classical task scheduling execution methods. Through the core priority evaluation of tasks by the quantum-enhanced scheduling decision module, the task priorities are determined and corresponding scheduling decisions are generated, thereby achieving efficient processing of complex tasks and ensuring the optimal utilization of system resources.

[0158] In some embodiments of the present disclosure, the task management module includes a task receiving module, a task classification module, a task representation module, and a task transmission module. Specifically, after the task scheduling execution system starts running, the task receiving module receives the computing tasks to be processed, and the task classification module classifies the tasks into quantum computing tasks, classical computing tasks, or hybrid tasks of quantum computing and classical computing according to the information processing method; the task representation module extracts the features of various tasks using corresponding quantum encoding methods according to different task types, determines the initial quantum states of all tasks, and thus obtains the quantum state task information that can be processed by the quantum-enhanced scheduling decision module.

[0159] Existing task scheduling and execution systems are difficult to effectively integrate and manage quantum computing resources and classical computing resources, and cannot fully exploit the potential of the hybrid computing environment. According to the differences in information processing methods, the present disclosure classifies tasks and matches different quantum state encodings, enabling each type of task to obtain an adaptable scheduling scheme in the system, thereby achieving an adaptive processing of tasks in the hybrid computing environment. Existing task scheduling and execution systems are difficult to achieve effective task allocation and resource coordination in a hybrid computing environment that includes quantum computing resources and classical computing resources, resulting in low processing efficiency. By setting up the above-mentioned modules, the present disclosure effectively combines quantum computing resources and classical computing resources, which is conducive to the efficient execution of hybrid computing tasks.

[0160] In some embodiments of the present disclosure, the task scheduling module includes a data input module, a resource allocation module, and a task execution module. The task scheduling module receives the task priority information output by the quantum-enhanced scheduling decision module, extracts the relevant execution information, and classifies and executes it.

[0161] Specifically, the data input module receives the task priority information output by the quantum-enhanced scheduling decision module and extracts the relevant execution information of the task. The resource allocation module allocates the tasks to be executed to the corresponding execution queues according to the key task characteristics. The task execution module implements the execution operations of various tasks. Through the task scheduling module, the execution decision tasks of the module can be optimized, and the computing resources can be adjusted accordingly, further improving the execution efficiency of the task scheduling and execution system and realizing resource management and optimization.

[0162] As Figure 7 shown, in some embodiments of the present disclosure, the task scheduling module further includes a result analysis module, which is configured to calculate and analyze the key task characteristics (such as average priority, standard deviation, etc.) that affect the task priority, and generate a task processing statistical report, which is conducive to optimizing the task scheduling and execution system.

[0163] The feedback of the result analysis module includes: generating a performance report in real time, identifying abnormal resource utilization, detecting task backlog, counting historical execution data, analyzing the task feature distribution, evaluating the effect of the scheduling strategy, or providing a basis for system parameter tuning, etc.

[0164] Traditional scheduling task scheduling and execution systems have limited computing power when dealing with complex optimization problems. On the one hand, traditional scheduling task scheduling and execution systems cannot execute tasks efficiently and accurately; the present disclosure introduces a quantum-enhanced scheduling decision module, which gives full play to the advantages of quantum computing. By extracting the priority characteristics of tasks and constructing a task priority function, the optimal scheduling scheme is determined, realizing the rapid execution of complex tasks and the flexible scheduling of computing resources. On the other hand, traditional task scheduling and execution systems are difficult to evaluate task priorities quickly and accurately, and cannot meet the processing of large-scale and highly relevant task sets; while the present disclosure makes full use of quantum characteristics and optimizes and coordinates the task scheduling process by means of the parallelism, superposition state, and entanglement of quantum computing, further improving the performance of the task scheduling and execution system. At the same time, through the setting of the task execution and result analysis parallel module, the resource allocation and execution module focuses on task scheduling and execution, and the result analysis module focuses on performance evaluation and optimization suggestions. The overall task scheduling and execution system has higher execution efficiency and can perform resource allocation without waiting for result analysis, and the analysis work and the execution process are carried out in parallel; and the result analysis module affects the task scheduling and execution system through the monitoring feedback module, and finally forms a closed-loop optimization system.

[0165] In some embodiments of the present disclosure, the task scheduling and execution system further includes a monitoring feedback module. The monitoring feedback module continuously monitors the task execution situation of the task scheduling module and adjusts the allocation of computing resources according to the running situation. With the auxiliary cooperation of the monitoring feedback module, the allocation of computing resources can be adjusted according to the task execution situation, further improving the task scheduling and execution efficiency, ensuring that the task scheduling and execution system can run safely and reliably, and at the same time being beneficial to the scalability of the task scheduling and execution system.

[0166] In some embodiments of the present disclosure, the monitoring feedback module monitors the running state of the execution queue and obtains the resource utilization rate of each execution queue in real time. When the resource utilization rate of the execution queue is relatively high (for example, the resource utilization rate is ≥80%), the corresponding computing resources are replenished in a timely manner. When the resource utilization rate of the execution queue is relatively low (for example, the resource utilization rate is ≤30%), the resources of this queue can be moved to other execution queues that need to replenish computing resources, effectively improving the efficiency of the task scheduling and execution system.

[0167] Computing resources refer to the physical and logical components used to execute computing tasks and the operations they support, including qubits, quantum gates, quantum memories, quantum processors, etc.

[0168] In some embodiments of the present disclosure, the monitoring feedback module can also transmit the processed task information to the quantum-enhanced scheduling decision module for further task planning or resource adjustment.

[0169] The present disclosure also provides a computer-readable storage medium storing a computer program for running a task scheduling execution method based on a hybrid computing environment, wherein the computer program causes a computer to execute the task scheduling execution method as described above.

[0170] Finally, it should be noted that the above are only the preferred embodiments of the present disclosure and are not intended to limit the present disclosure. Although the present disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present disclosure shall be included within the protection scope of the present disclosure.

Claims

1. A task scheduling execution method based on a hybrid computing environment, characterized in that: The task scheduling execution method comprises: receiving a task, classifying the task into task types and performing quantum coding on the task to form quantum coding information for representing the task, and determining an initial quantum state of the task; Evaluate the quantum enhanced priority of the task based on the quantum coded information, and convert the quantum enhanced priority evaluation result of the task into task priority information for a classical task scheduling execution method; The tasks are scheduled and executed based on the task priority information.

2. The task scheduling execution method according to claim 1, characterized in that: Evaluating the quantum enhancement priority of the task based on the quantum encoded information includes: The initial quantum states of all the tasks are superimposed to form a quantum superposition state containing quantum coded information of all the tasks.

3. The task scheduling execution method according to claim 2, characterized in that: Evaluating the quantum enhancement priority of the task based on the quantum coded information further includes: Determine a priority evaluation function through a quantum gate, and perform quantum parallel evaluation based on the quantum state of the task; enhance the amplitude of the quantum state of the high priority task and reduce the amplitude of the quantum state of the low priority task through quantum interference; Perform quantum measurement to collapse the quantum state of the task after quantum interference into a classical bit state; The priority results of the tasks determined by quantum measurement are post-processed using a classical task scheduling execution method to generate task priority information for the classical task scheduling execution method.

4. The task scheduling execution method according to claim 3, characterized in that: The influencing factors of the priority evaluation function include at least one of task execution time, resource requirements, task dependencies and task urgency.

5. The task scheduling execution method according to claim 1, characterized in that: The step of dividing the task into task types and performing quantum coding to form quantum coding information for representing the task, and determining the initial quantum state of the task includes: Classifying the tasks into quantum computing tasks, classical computing tasks, or hybrid tasks of quantum computing and classical computing according to the task types; According to the task type of each task, the corresponding quantum code is matched to the task to form the quantum code information, and the initial quantum state is determined.

6. The task scheduling execution method according to claim 5, characterized in that: According to the task type of each task, matching the task with the corresponding quantum code includes at least one of the following: Encoding the quantum computing task in a quantum state; Encoding the classical computing task in a quantum superposition state; The hybrid task of quantum computing and classical computing is encoded simultaneously in quantum states and quantum superposition states.

7. The task scheduling execution method according to claim 1, characterized in that: The scheduling and executing of the task based on the task priority information includes: Receive the task priority information, allocate resources according to the task priority information, and perform classified execution of the tasks according to task types; By analyzing the task priority information, a task processing statistics report is generated.

8. The task scheduling execution method according to claim 1, characterized in that: The task scheduling execution method further includes monitoring feedback, wherein the monitoring feedback continuously monitors the execution status of the task and dynamically adjusts the resource allocation of the task based on the execution status of the task.

9. A task scheduling execution system based on a hybrid computing environment, characterized in that: include: A task management module, which performs task type classification and quantum coding on the tasks, forms quantum coding information for representing the tasks, and determines the initial quantum state of the tasks; A quantum enhanced scheduling decision module, which evaluates the quantum enhanced priority of the task based on the quantum coded information, and converts the quantum enhanced priority evaluation result of the task into task priority information for a classical task scheduling execution method; A task scheduling module schedules and executes the task based on the task priority information.

10. A computer-readable storage medium, characterized in that: It stores a computer program for running a task scheduling execution method based on a hybrid computing environment, wherein the computer program enables a computer to execute the task scheduling execution method as described in any one of claims 1-8.

Images