Task scheduling method and device for industrial internet equipment, equipment and medium

Abstract

The invention discloses a task scheduling method and device for industrial internet equipment, equipment and a medium, and relates to the technical field of cloud computing, and the method comprises the steps: forming edge computing tasks into a corresponding target task set, and determining the task information of the edge computing tasks; target edge servers are determined, the target edge servers form a corresponding edge server set, and edge server information of the target edge servers is obtained; and constructing a nonlinear optimization model, and determining a task scheduling decision by using the nonlinear optimization model and a preset hybrid genetic taboo search algorithm, so as to process the edge calculation task according to the task scheduling decision by using the target edge server. According to the method and the device, the target task set and the edge server set are constructed, and the corresponding relationship between the edge computing tasks and the edge servers is determined by using the information of the target task set and the edge server set, so that the problem of low resource utilization rate caused by heterogeneity and dispersity of computing resources is solved.

Description

Technical Field

[0001] The present invention relates to the field of cloud computing technology, and in particular to a task scheduling method, device, equipment and medium for industrial Internet equipment. Background Art

[0002] In the context of the rapid development of today's industrial Internet and artificial intelligence technology, the demand for computing resources by industrial field equipment is growing. A large number of artificial intelligence technologies such as graphic recognition and data analysis are widely used to improve industrial production efficiency. However, the traditional cloud computing architecture cannot meet the real-time needs of the industrial field due to reasons such as long transmission links and homogeneous server resources. The edge computing network architecture is a new network architecture proposed on the basis of cloud computing.

[0003] Edge computing network architecture is a new network architecture proposed on the basis of cloud computing. Edge computing refers to the deployment of servers with certain computing resources on the edge side based on the cloud computing network architecture, while using the scattered terminal computing resources on the edge side to achieve deep integration of cloud, network, edge and terminal, and provide sufficient computing resources and computing services for industrial field equipment. Task scheduling and resource allocation is one of the important research issues in the field of cloud computing. Its scheduling algorithm is related to the utilization efficiency of cloud computing resources and the execution efficiency of industrial equipment tasks. At present, industrial field equipment has different computing resources required due to different manufacturers and interface protocols. At the same time, the computing power, heterogeneity, and dispersion of computing resources such as servers also bring huge challenges to task scheduling, resulting in low resource utilization efficiency. In addition, since the algorithm in the edge network architecture does not consider the heterogeneity of task equipment and server resources, it is not suitable for edge computing task scheduling in the field of industrial Internet. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a task scheduling method, device, equipment and medium for industrial Internet equipment, which can determine the correspondence between edge computing tasks and edge servers by constructing a target task set and an edge server set, and using the information of the above two, and use edge servers to process edge computing tasks based on the correspondence, thereby solving the problem of low resource utilization caused by the heterogeneity and dispersion of computing resources, and can effectively handle the heterogeneity of task equipment and the heterogeneity of server resources through a hybrid genetic taboo search algorithm, which provides a new research idea for edge computing task scheduling in the field of industrial Internet. The specific scheme is as follows:

[0005] In a first aspect, the present application provides a task scheduling method for an industrial Internet device, comprising:

[0006] Obtain edge computing tasks of each industrial Internet device preset in the industrial Internet within a target time period, group the edge computing tasks into a corresponding target task set, and determine task information of the edge computing tasks in the target task set; the target time period is a time period determined according to the initial moment of scheduling the edge computing tasks;

[0007] Determine a target edge server from the preset industrial Internet, group the target edge servers into a corresponding edge server set, and obtain edge server information of the target edge server in the edge server set;

[0008] A nonlinear optimization model is constructed according to preset decision variables, and a target correspondence relationship between the edge computing task and the target edge server is determined by using the nonlinear optimization model and a preset hybrid genetic taboo search algorithm according to the task information, the edge server information, and the preset decision variables; the preset hybrid genetic taboo search algorithm is an algorithm constructed according to a preset genetic algorithm and a preset taboo search algorithm;

[0009] A task scheduling decision corresponding to the edge computing task is constructed according to the target corresponding relationship, so as to utilize the target edge server to process the edge computing task according to the task scheduling decision.

[0010] Optionally, the determining the target correspondence between the edge computing task and the target edge server by using the nonlinear optimization model and a preset hybrid genetic taboo search algorithm includes:

[0011] Determine the initial correspondence between the edge computing task and the target edge server by using the nonlinear optimization model, and construct a corresponding population through the initial correspondence;

[0012] Determine the maximum number of iterations corresponding to the population, and after initializing the population, determine the individual fitness corresponding to each of the initial corresponding relationships in the population according to a preset fitness determination rule;

[0013] Based on the current individual fitness, a preset selection operation, a preset crossover operation and a preset mutation operation are sequentially performed on the population to obtain an operation result of the population;

[0014] updating the individual fitness according to the operation result, adjusting the population according to a preset local taboo search rule based on the updated individual fitness, and after the population adjustment is completed, increasing the current number of iterations by one;

[0015] Determine whether the current number of iterations is less than the maximum number of iterations;

[0016] If the current number of iterations is less than the maximum number of iterations, jump to the step of sequentially performing a preset selection operation, a preset crossover operation and a preset mutation operation on the population based on the current individual fitness;

[0017] If the current number of iterations is not less than the maximum number of iterations, the target correspondence is determined according to the currently adjusted population.

[0018] Optionally, before determining the target correspondence between the edge computing task and the target edge server by using the nonlinear optimization model and a preset hybrid genetic taboo search algorithm according to the task information, the edge server information, and the preset decision variables, the method further includes:

[0019] Determine the resource matching degree between the edge computing task and the target edge server according to the memory capacity required for the task, the hard disk capacity required for the task, and the computing resources required for the task in the task information and the server memory capacity, the server hard disk capacity, and the server computing resources in the server information;

[0020] The response delay of the edge computing task is determined based on the expected task completion time, task data volume, task calculation volume in the task information and the data transmission speed, server processing speed and initial time in the server information.

[0021] Optionally, constructing a nonlinear optimization model according to preset decision variables includes:

[0022] The nonlinear optimization model is constructed according to the preset decision variables, the resource matching degree and the response delay, so as to determine the target correspondence between the edge computing task and the target edge server according to the task information, the edge server information and the preset decision variables using the nonlinear optimization model and the preset hybrid genetic taboo search algorithm.

[0023] Optionally, determining the target correspondence between the edge computing task and the target edge server by using the nonlinear optimization model and a preset hybrid genetic taboo search algorithm according to the task information, the edge server information, and the preset decision variables includes:

[0024] Determine the target decision variable value corresponding to the edge computing task by a preset decision variable determination rule, and determine the target correspondence between the edge computing task and the target edge server by using the nonlinear optimization model and the preset hybrid genetic taboo search algorithm according to the task information, the edge server information, and the target decision variable value;

[0025] The step of determining the target decision variable value corresponding to the edge computing task by using a preset decision variable determination rule includes:

[0026] If the edge computing task is assigned to the target node of the target edge server, determining the preset decision variable as the first decision variable value;

[0027] If the edge computing task is not assigned to the target node of the target edge server, the preset decision variable is determined as the second decision variable value.

[0028] Optionally, the processing the edge computing task by using the target edge server according to the task scheduling decision includes:

[0029] According to the preset hybrid genetic taboo search algorithm and the task scheduling decision, a task processing server corresponding to the edge computing task is determined from a number of target edge servers, so that the task processing server can be used to process the corresponding edge computing task; the edge computing task corresponds one-to-one to the industrial Internet device in the preset industrial Internet.

[0030] Optionally, the processing the edge computing task by using the target edge server according to the task scheduling decision includes:

[0031] Using the task processing server to receive corresponding edge computing tasks in sequence;

[0032] Determine a task receiving order of the plurality of edge computing tasks, and process the edge computing tasks in sequence according to the task receiving order using the task processing server and the task scheduling decision.

[0033] In a second aspect, the present application provides a task scheduling device for an industrial Internet device, comprising:

[0034] A task information determination module is used to obtain edge computing tasks of each industrial Internet device preset in the industrial Internet within a target time period, and to group the edge computing tasks into a corresponding target task set, and to determine the task information of the edge computing tasks in the target task set; the target time period is a time period determined according to the initial moment of scheduling the edge computing tasks;

[0035] A server information acquisition module, used to determine a target edge server from the preset industrial Internet, group the target edge servers into a corresponding edge server set, and obtain edge server information of the target edge server in the edge server set;

[0036] A model building module, used to build a nonlinear optimization model according to preset decision variables, and use the nonlinear optimization model to determine the target correspondence between the edge computing task and the target edge server according to the task information, the edge server information, the preset decision variables and a preset hybrid genetic taboo search algorithm; the preset hybrid genetic taboo search algorithm is an algorithm built according to a preset genetic algorithm and a preset taboo search algorithm;

[0037] The task processing module is used to construct a task scheduling decision corresponding to the edge computing task according to the target corresponding relationship, so as to use the target edge server to process the edge computing task according to the task scheduling decision.

[0038] In a third aspect, the present application provides an electronic device, including:

[0039] Memory, used to store computer programs;

[0040] A processor is used to execute the computer program to implement the aforementioned task scheduling method for industrial Internet devices.

[0041] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the aforementioned task scheduling method for industrial Internet devices.

[0042] In the present application, the industrial Internet first obtains the edge computing tasks of each industrial Internet device in the preset industrial Internet within the target time period, and organizes the edge computing tasks into a corresponding target task set, and determines the task information of the edge computing tasks in the target task set; the target time period is a time period determined according to the initial moment of scheduling the edge computing task, and then determines the target edge server from the preset industrial Internet, organizes the target edge servers into a corresponding edge server set, and obtains the edge server information of the target edge server in the edge server set, and then constructs a nonlinear optimization model according to preset decision variables, and determines the target correspondence between the edge computing task and the target edge server according to the task information, the edge server information, and the preset decision variables using the nonlinear optimization model and a preset hybrid genetic taboo search algorithm; the preset hybrid genetic taboo search algorithm is an algorithm constructed according to a preset genetic algorithm and a preset taboo search algorithm, and finally, a task scheduling decision corresponding to the edge computing task is constructed according to the target correspondence, so as to use the target edge server to process the edge computing task according to the task scheduling decision. In this way, the present application obtains the task information and edge server information by respectively utilizing the edge computing tasks and edge servers to construct the corresponding target task set and edge server set, and obtains the correspondence between the edge computing tasks and the edge servers by utilizing the task information and the edge server information, thereby solving the problem of heterogeneity between the server computing resources and the task demand resources. According to the above correspondence, a corresponding edge server is assigned to each edge computing task to process it, thereby solving the problem of low resource utilization efficiency caused by server dispersion. The heterogeneity of task equipment and server resources is processed by a hybrid genetic taboo search algorithm, which provides a new research idea for edge computing task scheduling in the field of industrial Internet. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 A flowchart of a task scheduling method for industrial Internet equipment disclosed in this application;

[0045] Figure 2 A flow chart of a method disclosed in the present application for determining a target correspondence relationship between an edge computing task and a target edge server by using a preset hybrid genetic taboo search algorithm;

[0046] Figure 3 A genetic algorithm flow chart disclosed in this application;

[0047] Figure 4 A flowchart of a taboo search algorithm disclosed in this application;

[0048] Figure 5 This is a structural diagram of a task scheduling device for an industrial Internet device disclosed in this application;

[0049] Figure 6 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] In the field of industrial Internet, due to the different interface protocols and computing resources required by different manufacturers of industrial field equipment, the scheduling of edge computing tasks is inconvenient. To this end, the embodiment of the present application provides a task scheduling method for industrial Internet devices, which can use task information and edge server information to obtain the correspondence between edge computing tasks and edge servers, solve the problem of mismatch between server computing resources and task demand resources, and have heterogeneity, thereby solving the problem of low resource utilization efficiency caused by server dispersion.

[0052] See also Figure 1 As shown, an embodiment of the present invention discloses a task scheduling method for an industrial Internet device, including:

[0053] Step S11, obtain the edge computing tasks of each industrial Internet device in the preset industrial Internet within the target time period, and organize the edge computing tasks into a corresponding target task set, and determine the task information of the edge computing tasks in the target task set; the target time period is a time period determined according to the initial moment of scheduling the edge computing tasks.

[0054] In this embodiment, obtaining edge computing tasks for each industrial Internet device preset in the target time period may specifically include: setting the initial scheduling time w of the edge computing task, obtaining edge computing tasks for each industrial Internet device preset in the target time period before time w, for example, within 1 hour, stopping obtaining edge computing tasks at time w, and uniformly scheduling and auctioning the edge computing tasks collected in the target time period before time w. Before the auction, first determine all the tasks that need to be auctioned this time, that is, obtain the edge computing tasks for each industrial Internet device preset in the target time period before time w, organize the above edge computing tasks into a corresponding target task set, that is, set N, and obtain the task information of all edge computing tasks in set N. It should be noted that the task information in the above set N includes but is not limited to the data volume of task i∈N. , the amount of calculation , expected completion time , the required memory size , required hard disk capacity , the required computing module resources Among them, computing module resources Indicates the task's demand for multiple computing module resources such as GPU (Graphics Processing Unit) and NPU (Neural network Processing Unit), which are specifically expressed as follows:

[0055] ;

[0056] in is the number of types of computing resources, Representation Task exist The computing demand on the class resources. When the task does not require the computing resources of this class, the computing demand value for this class of resources is expressed as 0.

[0057] Table 1 Task parameter setting table

[0058]

[0059] As shown in Table 1 above, in a specific implementation, the initial time of the edge computing task scheduling is set to 8:00:00, and the 100 edge computing tasks collected within 1 hour before this time are uniformly scheduled, and the task numbers are 1-100 in sequence. The task information includes the data volume, in kB (Kilobyte); the computing volume, in MI; the memory capacity, in MB; the hard disk capacity, in MB; the number of computing resource modules required by the task is 1, in FLOPS (Floating Point Operations Per Second); and the expected completion time of the task. By determining the edge computing tasks that need to be scheduled and auctioned and grouping the above tasks into a set, the task information of the edge computing tasks can be uniformly obtained, thereby improving work efficiency.

[0060] Step S12: determine a target edge server from the preset industrial Internet, organize the target edge servers into a corresponding edge server set, and obtain edge server information of the target edge server in the edge server set.

[0061] In this embodiment, the target edge server is the scattered terminal computing resources at the edge of the cloud computing network architecture, and the corresponding central server is also there. In this embodiment, all available edge servers in this auction are first determined, and the target edge servers are grouped into a corresponding edge server set, that is, set , and get the collection The basic information of all edge servers in the server. It should be noted that the server information of the edge server includes but is not limited to the server Processing speed , data transmission speed , memory capacity , hard disk capacity , computing module resources . The vector Indicates the GPU, NPU and other computing module resources contained in the server node. The specific representation is as follows:

[0062] ;

[0063] in is the number of types of computing resources, Represents a server node exist The computing resource demand that can be provided per unit time on a computing resource of this type. When the server node does not have such resources, the value of this computing resource is expressed as 0.

[0064] Table 2 Server parameter setting table

[0065]

[0066] In a specific implementation, the number of edge-side server nodes is 5, and the server numbers are 1-5 in sequence, as shown in Table 2 above, and the server information includes: the processing speed of the server, in MIPS; the transmission speed of the server, in kB / s; the memory capacity of the server, in MB; the hard disk capacity of the server, in MB; the number of computing resource modules possessed by the server is 1, in FLOPS. In this embodiment, by determining that all available edge servers are required and forming the above edge servers into a set, the server information of the edge servers can be uniformly obtained, thereby improving work efficiency.

[0067] Step S13, constructing a nonlinear optimization model according to preset decision variables, and determining the target correspondence between the edge computing task and the target edge server according to the task information, the edge server information, and the preset decision variables using the nonlinear optimization model and a preset hybrid genetic taboo search algorithm; the preset hybrid genetic taboo search algorithm is an algorithm constructed according to a preset genetic algorithm and a preset taboo search algorithm.

[0068] In this embodiment, it is necessary to first set the decision variables and build a nonlinear optimization model based on the decision variables. Then, the edge computing task scheduling method based on the hybrid genetic taboo search algorithm is used to solve the nonlinear optimization model according to the task set N, the server set M and the basic information to determine the task scheduling decision. The above task scheduling decision is a task scheduling decision built according to the target correspondence between the edge computing task and the target edge server. Before determining the target correspondence between the edge computing task and the target edge server using the nonlinear optimization model and the preset hybrid genetic taboo search algorithm according to the task information, the edge server information, and the preset decision variables, it specifically also includes: according to the memory capacity required for the task in the task information, that is, , the hard disk capacity required for the task, that is , and the computing resources required for the task, that is, and the server memory capacity in the server information, i.e. , server hard disk capacity, that is , server computing resources, i.e. Determine the resource matching degree between the edge computing task and the target edge server, and calculate the expected completion time of the task according to the task information, that is, , the amount of task data, i.e. , the task computation amount, that is and the data transmission speed in the server information, i.e. , server processing speed, i.e. And the initial moment, that is, w determines the response delay of the edge computing task. Among them, the model of the above resource matching degree is as follows:

[0069] According to the memory capacity of the server node , hard disk capacity , computing module resources and the memory capacity required by the task , hard disk capacity , computing module resources Compute the resource matching degree between the task and the server node; Task Offloading to edge nodes The resource matching calculation formula is:

[0070] ;

[0071] The task response model corresponding to the above task response delay is as follows:

[0072] According to the data transmission speed of the server node , processing speed and the amount of data for the task The response delay of the computational task; Offloading to edge nodes The response delay can be calculated as:

[0073] ;

[0074] The specific meanings of the above decision variables are:

[0075] ;

[0076] The objective function of the above nonlinear optimization model is:

[0077] ;

[0078] The objective function is to minimize the maximum value of task response delay and resource matching degree in all server nodes.

[0079] The constraints of the above nonlinear optimization model are as follows:

[0080] Constraint 1: ;

[0081] Constraint 2: ;

[0082] Constraint three: ;

[0083] Constraint 4: ;

[0084] Constraint 5: ;

[0085] Among them, constraint one means that any task can only be assigned to one server. Constraint two means that each task can only be assigned to a server node that meets the memory requirements. Constraint three means that each task can only be assigned to a server node that meets the hard disk requirements. Constraint four means that each task can only be assigned to a server node that meets the resource requirements of the computing module. Constraint five represents the value range of the decision variable. In this embodiment, before determining the target correspondence between the edge computing task and the target edge server, the resource matching degree between the edge computing task and the target edge server and the response delay of the edge computing task are first determined, thereby improving the reliability of the correspondence determination process.

[0086] Step S14: construct a task scheduling decision corresponding to the edge computing task according to the target corresponding relationship, so as to utilize the target edge server to process the edge computing task according to the task scheduling decision.

[0087] In this embodiment, the target edge server is used to process the edge computing task according to the task scheduling decision, which can specifically include: determining the task processing server corresponding to the edge computing task from a number of target edge servers according to the preset hybrid genetic taboo search algorithm and the task scheduling decision, so as to use the task processing server to process the corresponding edge computing task; it should be noted that the edge computing task corresponds one-to-one to the industrial Internet device in the preset industrial Internet, that is, each industrial Internet device only generates one edge computing task, and each task can only be processed by one server node.

[0088] In this embodiment, the target edge server is used to process the edge computing task according to the task scheduling decision, including: using the task processing server to sequentially receive a number of corresponding edge computing tasks; determining the task reception order of a number of edge computing tasks, and using the task processing server to sequentially process the edge computing tasks according to the task scheduling decision in accordance with the task reception order. It can be understood that the edge server can receive multiple edge computing tasks, but can only process one task at a time, and the tasks are executed sequentially in the order of transmission. In this embodiment, by determining the correspondence between edge computing tasks and edge servers, and determining the processing order of multiple edge computing tasks by the edge server, the situation of task loss is avoided, and the reliability of edge computing task processing by the edge server is improved.

[0089] It can be seen that the embodiment of the present application obtains the correspondence between edge computing tasks and edge servers by utilizing task information and edge server information, thereby solving the problem of heterogeneity between server computing resources and task demand resources. By allocating a corresponding edge server to each edge computing task for processing according to the above correspondence, the problem of low resource utilization efficiency caused by server dispersion is solved. At the same time, the embodiment of the present application avoids task loss by determining the processing order of multiple edge computing tasks by the edge server, thereby improving the reliability of edge computing task processing by the edge server.

[0090] In order to further illustrate the process of determining the target correspondence between edge computing tasks and target edge servers using a nonlinear optimization model and a preset hybrid genetic taboo search algorithm in the aforementioned application embodiment, the present application also discloses a specific task scheduling method for industrial Internet devices to illustrate how to use a preset hybrid genetic taboo search algorithm to determine the target correspondence between edge computing tasks and target edge servers, see Figure 2 As shown, the embodiment of the present application also discloses a method for determining a target correspondence relationship between an edge computing task and a target edge server using a preset hybrid genetic taboo search algorithm, including:

[0091] Step S21: Determine the initial correspondence between edge computing tasks and target edge servers using a nonlinear optimization model, and construct a corresponding population based on the initial correspondence.

[0092] It can be understood that the genetic taboo algorithm is an optimization search algorithm. The traditional genetic algorithm can perform a global search in the feasible solution space to find the optimal solution or the approximate optimal solution by simulating the evolution process in nature. The hybrid genetic taboo search algorithm in this embodiment includes a genetic algorithm and a taboo search algorithm. In the problem of edge computing task scheduling, the hybrid genetic taboo search algorithm can effectively handle the heterogeneity of task equipment and the heterogeneity of server resources, while taking into account the response delay and resource matching of the task, which provides a new research idea for edge computing task scheduling in the field of industrial Internet.

[0093] In this embodiment, in the genetic algorithm stage, the genetic algorithm uses natural number encoding, the length of the solution is the number of tasks, each position in the solution represents an edge computing task, and the element at each position represents the serial number of the edge server to which the corresponding task is assigned. For example, a feasible solution to the problem of allocating five tasks to four servers can be encoded as

[0094] X=(1,2,1,3,4);

[0095] The corresponding meaning is: tasks 1 and 3 are assigned to server 1, task 2 is assigned to server 2, task 4 is assigned to server 3, and task 5 is assigned to server 4. The above X is a chromosome. It should be noted that the fitness of each chromosome is equal to the inverse of the maximum value of the comprehensive matching degree among all server nodes. The comprehensive matching degree is the sum of the task execution time and resource matching degree of all tasks on server node j. The execution time of all tasks on any server is the sum of the transmission time and running time of all tasks assigned to the server. Then the initial population is generated. First, the population size is recorded as , the initial population consists of a greedy solution and In the greedy solution, it is assumed that the tasks assigned to the same server are transmitted in order of priority from high to low. Figure 3 As shown, the set of all tasks is set to N, and the greedy solution generation scheme is as follows, including: step 1, calculating the priority of the task according to the expected completion time of the task, the task scheduling time, and the calculation amount of the task, and sorting all tasks in descending order according to the priority of the task; step 2, assigning the highest priority task in the set N to the available server with the smallest sum of task execution time and resource matching degree, and updating the set N; step 3, if the set N is empty, output the allocation result and terminate the greedy algorithm; otherwise, jump to step 1. For the remaining chromosomes, this embodiment adopts the method of randomly selecting any available server from the assignable servers to generate the initial greedy solution. The above steps of generating the initial greedy solution may specifically include: step 1, according to the constraints of memory resources, hard disk resources, and computing module resources, determine the set of server nodes that all tasks meet the resource constraints; step 2, randomly select an available server node for the task with the smallest task number in set N, and update set N; step 3, if set N is empty, output the allocation result and terminate the algorithm; otherwise, jump to step 2 in the steps of generating the initial greedy solution, the updated set N is the generated initial greedy solution, and the initial greedy solution and the above chromosomes are the generated initial population. In this embodiment, by describing the initial population generation step in detail, the process of constructing the corresponding population through the initial corresponding relationship is made more specific.

[0096] Step S22, determine the maximum number of iterations corresponding to the population, and after initializing the population, determine the individual fitness corresponding to each of the initial corresponding relationships in the population according to a preset fitness determination rule; based on the current individual fitness, perform a preset selection operation, a preset crossover operation, and a preset mutation operation on the population in turn to obtain the operation result of the population.

[0097] The roulette wheel selection method with the elite retention strategy is a common selection operation method in the hybrid genetic taboo search algorithm. The elite retention strategy can retain the best individuals in the population and enter the next generation of the population, thereby improving the efficiency of the algorithm in finding the optimal solution. This embodiment uses roulette and elite retention strategies for selection operations. The crossover operation is an important operation step in the hybrid genetic taboo search algorithm. Through the crossover operation, the search space of the population can be effectively expanded, and the efficiency of the algorithm in searching for the optimal solution can be improved. Common crossover methods in the hybrid genetic taboo search algorithm include single-point crossover, multi-point crossover, uniform crossover, arithmetic crossover and other crossover methods. The preset crossover operation in this embodiment is a multi-point crossover method. There are two methods for the mutation operation of the hybrid genetic taboo search algorithm. The first is to randomly generate a new gene to replace a gene of the original chromosome; the second is to select a mutant gene on the original chromosome and modify the gene to become a mutant gene. This embodiment selects the second operation method to perform the mutation operation. The probability of the mutation operation occurring is . The above mutation operation may specifically include: step 1, selecting the server node with the largest sum of task execution time and resource matching degree among all current server nodes; step 2, randomly selecting a task on the server node, and determining the set of executable servers for the task; step 3, selecting a server from the set of all servers that can execute the task, and adding the task to the task queue of the server node, that is, selecting a mutant gene in the original chromosome, modifying the gene, and obtaining the operation result of the population, and the above operation result is the mutant gene obtained after the gene is modified. The reliability of the operation result of the obtained population is guaranteed by the above preset selection operation, preset crossover operation and preset mutation operation.

[0098] Step S23, updating the individual fitness according to the operation result, and adjusting the population according to a preset local taboo search rule based on the updated individual fitness, and after the population adjustment is completed, increasing the current iteration number by one.

[0099] In the preset local taboo search rule, i.e., the taboo search algorithm stage, the above-mentioned population is adjusted according to the preset local taboo search rule based on the updated individual fitness, such as Figure 4As shown, it can specifically include: step 1, according to the population generated by the genetic algorithm mutation, select the individual with the largest fitness value as the initial solution, and give the initial parameters of the taboo search algorithm; step 2, determine whether the algorithm meets the preset termination condition, if so, output the tuned result, if not, continue to execute the subsequent steps; step 3, according to the neighborhood function of the current solution, generate several neighborhood solutions, and according to the fitness value of the solution, retain the neighborhood solutions that meet the preset conditions as candidate solutions; step 4, determine whether the candidate solution meets the contempt rule, if so, unban the solution and use it as the new current optimal solution, update the current optimal state, determine whether it meets the convergence criterion, if so, end; if not, jump to step 3; step 5, if the candidate solution does not meet the contempt criterion, select the optimal state among the candidate solutions as the new current solution, replace the emergency object that first enters the taboo table with the corresponding taboo object, and then jump to step 2; after the above population adjustment is completed, increase the current number of iterations by one.

[0100] Step S24: determine whether the current number of iterations is less than the maximum number of iterations, and determine the target correspondence between the edge computing task and the target edge server according to the corresponding judgment result.

[0101] In this embodiment, the target correspondence between the edge computing task and the target edge server is determined according to the corresponding judgment result, which may specifically include: if the current number of iterations is less than the maximum number of iterations, jumping to the steps of performing a preset selection operation, a preset crossover operation and a preset mutation operation on the population in sequence based on the current individual fitness; if the current number of iterations is not less than the maximum number of iterations, determining the target correspondence according to the currently adjusted population.

[0102] It can be seen that the embodiment of the present application constructs a population corresponding to the initial correspondence relationship through a preset genetic algorithm, and adjusts the above population through a preset taboo search algorithm, thereby avoiding the problem that the traditional genetic algorithm converges slowly and is prone to falling into local extreme values. A hybrid genetic taboo search algorithm is proposed, and the taboo table and contempt criterion in the taboo search algorithm are used to optimize and adjust the population after the iteration of the genetic algorithm to avoid it falling into the local optimal situation; and on the basis of retaining the global search capability of the genetic algorithm, it avoids falling into the local optimal situation, thereby ensuring the reliability of the correspondence between the edge computing task and the target edge server.

[0103] See also Figure 5 As shown, the embodiment of the present invention further discloses a data acquisition device, including:

[0104] The task information determination module 11 is used to obtain the edge computing tasks of each industrial Internet device preset in the industrial Internet within the target time period, and to form the edge computing tasks into a corresponding target task set, and to determine the task information of the edge computing tasks in the target task set; the target time period is a time period determined according to the initial time of scheduling the edge computing tasks;

[0105] A server information acquisition module 12 is used to determine a target edge server from the preset industrial Internet, group the target edge servers into a corresponding edge server set, and obtain edge server information of the target edge server in the edge server set;

[0106] A model building module 13 is used to build a nonlinear optimization model according to preset decision variables, and use the nonlinear optimization model to determine the target correspondence between the edge computing task and the target edge server according to the task information, the edge server information, the preset decision variables and a preset hybrid genetic taboo search algorithm; the preset hybrid genetic taboo search algorithm is an algorithm built according to a preset genetic algorithm and a preset taboo search algorithm;

[0107] The task processing module 14 is used to construct a task scheduling decision corresponding to the edge computing task according to the target corresponding relationship, so as to use the target edge server to process the edge computing task according to the task scheduling decision.

[0108] It can be seen that the present application obtains task information and edge server information by respectively utilizing edge computing tasks and edge servers to construct corresponding target task sets and edge server sets, and obtains the correspondence between edge computing tasks and edge servers by utilizing the task information and edge server information, thereby solving the problem of mismatch and heterogeneity between server computing resources and task demand resources. By allocating corresponding edge servers to each edge computing task for processing according to the above correspondence, the problem of low resource utilization efficiency caused by server dispersion is solved.

[0109] In some specific embodiments, the model building module 13 may specifically include:

[0110] A population determination unit, configured to determine an initial correspondence between the edge computing task and the target edge server using the nonlinear optimization model, and construct a corresponding population through the initial correspondence;

[0111] A fitness determination unit, used to determine the maximum number of iterations corresponding to the population, and after initializing the population, determine the individual fitness corresponding to each of the initial corresponding relationships in the population according to a preset fitness determination rule;

[0112] A population operation unit, used to sequentially perform a preset selection operation, a preset crossover operation and a preset mutation operation on the population based on the current individual fitness, to obtain an operation result of the population;

[0113] A population adjustment unit, configured to update the individual fitness according to the operation result, and adjust the population according to a preset local taboo search rule based on the updated individual fitness, and after the population adjustment is completed, increase the current iteration number by one;

[0114] An iteration number determination unit, used to determine whether the current iteration number is less than the maximum iteration number;

[0115] a step jump unit, configured to jump to the step of sequentially performing a preset selection operation, a preset crossover operation and a preset mutation operation on the population based on the current individual fitness if the current number of iterations is less than the maximum number of iterations;

[0116] A first correspondence determination unit, configured to determine the target correspondence according to the currently adjusted population if the current number of iterations is not less than the maximum number of iterations;

[0117] A second correspondence determination unit is used to construct the nonlinear optimization model according to the preset decision variables, the resource matching degree and the response delay, so as to determine the target correspondence between the edge computing task and the target edge server by using the nonlinear optimization model and the preset hybrid genetic taboo search algorithm according to the task information, the edge server information and the preset decision variables;

[0118] A correspondence determination submodule, used to determine the target decision variable value corresponding to the edge computing task by a preset decision variable determination rule, and determine the target correspondence between the edge computing task and the target edge server by using the nonlinear optimization model and the preset hybrid genetic taboo search algorithm according to the task information, the edge server information, and the target decision variable value;

[0119] In some specific embodiments, the model building module 13 further includes:

[0120] a matching degree determining unit, configured to determine a resource matching degree between the edge computing task and the target edge server according to the memory capacity required for the task, the hard disk capacity required for the task, and the computing resources required for the task in the task information and the server memory capacity, the server hard disk capacity, and the server computing resources in the server information;

[0121] A response delay determination unit is used to determine the response delay of the edge computing task based on the expected task completion time, task data volume, task calculation volume in the task information and the data transmission speed, server processing speed and initial time in the server information.

[0122] In some specific embodiments, the corresponding relationship determination submodule may specifically include:

[0123] a first decision variable value determining unit, configured to determine the preset decision variable as a first decision variable value if the edge computing task is assigned to a target node of the target edge server;

[0124] The second decision variable value determining unit is used to determine the preset decision variable as the second decision variable value if the edge computing task is not assigned to the target node of the target edge server.

[0125] In some specific embodiments, the task processing module 14 may specifically include:

[0126] A first task processing unit is used to determine a task processing server corresponding to the edge computing task from a number of target edge servers according to the preset hybrid genetic taboo search algorithm and the task scheduling decision, so as to use the task processing server to process the corresponding edge computing task; the edge computing task corresponds one-to-one to the industrial Internet device in the preset industrial Internet;

[0127] A task receiving unit, configured to use the task processing server to sequentially receive a corresponding number of the edge computing tasks;

[0128] The second task processing unit is used to determine the task receiving order of the plurality of edge computing tasks, and process the edge computing tasks in sequence according to the task receiving order using the task processing server according to the task scheduling decision.

[0129] Furthermore, the present application also discloses an electronic device. Figure 6 It is a structural diagram of an electronic device 20 shown according to an exemplary embodiment, and the content in the figure cannot be regarded as any limitation on the scope of use of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input and output interface 25 and a communication bus 26. Among them, the memory 22 is used to store a computer program, and the computer program is loaded and executed by the processor 21 to implement the relevant steps in the task scheduling method of the industrial Internet device disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment can specifically be an electronic computer.

[0130] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0131] In addition, the memory 22 as a carrier for resource storage may be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon may include an operating system 221, a computer program 222, etc., and the storage method may be temporary storage or permanent storage.

[0132] The operating system 221 is used to manage and control the hardware devices on the electronic device 20 and the computer program 222, which can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program that can be used to complete the task scheduling method of the industrial Internet device executed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can further include a computer program that can be used to complete other specific tasks.

[0133] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the task scheduling method of the industrial Internet device disclosed above is implemented. For the specific steps of the method, reference may be made to the corresponding contents disclosed in the above embodiments, and no further description will be given here.

[0134] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0135] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in 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 the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may 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.

[0136] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly 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.

[0137] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0138] The technical solution provided by the present application is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technicians in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A task scheduling method for industrial Internet equipment, characterized in that: include: Obtain edge computing tasks of each industrial Internet device preset in the industrial Internet within a target time period, group the edge computing tasks into a corresponding target task set, and determine task information of the edge computing tasks in the target task set; The target time period is a time period determined according to the initial time of scheduling the edge computing task; Determine a target edge server from the preset industrial Internet, group the target edge servers into a corresponding edge server set, and obtain edge server information of the target edge server in the edge server set; Construct a nonlinear optimization model according to preset decision variables, and determine the target correspondence between the edge computing task and the target edge server using the nonlinear optimization model and a preset hybrid genetic taboo search algorithm according to the task information, the edge server information, and the preset decision variables; The preset hybrid genetic taboo search algorithm is an algorithm constructed according to a preset genetic algorithm and a preset taboo search algorithm; A task scheduling decision corresponding to the edge computing task is constructed according to the target corresponding relationship, so as to utilize the target edge server to process the edge computing task according to the task scheduling decision.

2. The task scheduling method for industrial Internet devices according to claim 1 is characterized in that: The determining the target correspondence between the edge computing task and the target edge server by using the nonlinear optimization model and a preset hybrid genetic taboo search algorithm includes: Determine the initial correspondence between the edge computing task and the target edge server by using the nonlinear optimization model, and construct a corresponding population through the initial correspondence; Determine the maximum number of iterations corresponding to the population, and after initializing the population, determine the individual fitness corresponding to each of the initial corresponding relationships in the population according to a preset fitness determination rule; Based on the current individual fitness, a preset selection operation, a preset crossover operation and a preset mutation operation are sequentially performed on the population to obtain an operation result of the population; updating the individual fitness according to the operation result, adjusting the population according to a preset local taboo search rule based on the updated individual fitness, and after the population adjustment is completed, increasing the current number of iterations by one; Determine whether the current number of iterations is less than the maximum number of iterations; If the current number of iterations is less than the maximum number of iterations, jump to the step of sequentially performing a preset selection operation, a preset crossover operation and a preset mutation operation on the population based on the current individual fitness; If the current number of iterations is not less than the maximum number of iterations, the target correspondence relationship is determined according to the currently adjusted population.

3. The task scheduling method for industrial Internet devices according to claim 1 is characterized in that: Before determining the target correspondence between the edge computing task and the target edge server by using the nonlinear optimization model and the preset hybrid genetic taboo search algorithm according to the task information, the edge server information, and the preset decision variables, the method further includes: Determine the resource matching degree between the edge computing task and the target edge server according to the memory capacity required for the task, the hard disk capacity required for the task, and the computing resources required for the task in the task information and the server memory capacity, the server hard disk capacity, and the server computing resources in the server information; The response delay of the edge computing task is determined based on the expected task completion time, task data volume, task calculation volume in the task information and the data transmission speed, server processing speed and initial time in the server information.

4. The task scheduling method for industrial Internet equipment according to claim 3 is characterized in that: The method of constructing a nonlinear optimization model according to preset decision variables includes: The nonlinear optimization model is constructed according to the preset decision variables, the resource matching degree and the response delay, so as to determine the target correspondence between the edge computing task and the target edge server according to the task information, the edge server information and the preset decision variables using the nonlinear optimization model and the preset hybrid genetic taboo search algorithm.

5. The task scheduling method for industrial Internet devices according to claim 1 is characterized in that: The determining the target correspondence between the edge computing task and the target edge server by using the nonlinear optimization model and a preset hybrid genetic taboo search algorithm according to the task information, the edge server information, and the preset decision variables includes: Determine the target decision variable value corresponding to the edge computing task by a preset decision variable determination rule, and determine the target correspondence between the edge computing task and the target edge server by using the nonlinear optimization model and the preset hybrid genetic taboo search algorithm according to the task information, the edge server information, and the target decision variable value; The step of determining the target decision variable value corresponding to the edge computing task by using a preset decision variable determination rule includes: If the edge computing task is assigned to the target node of the target edge server, determining the preset decision variable as the first decision variable value; If the edge computing task is not assigned to the target node of the target edge server, the preset decision variable is determined as the second decision variable value.

6. The task scheduling method for industrial Internet devices according to any one of claims 1 to 5, characterized in that: The using the target edge server to process the edge computing task according to the task scheduling decision includes: According to the preset hybrid genetic taboo search algorithm and the task scheduling decision, a task processing server corresponding to the edge computing task is determined from a number of target edge servers, so that the task processing server can be used to process the corresponding edge computing task; the edge computing task corresponds one-to-one to the industrial Internet device in the preset industrial Internet.

7. The task scheduling method for industrial Internet devices according to claim 6 is characterized in that: The using the target edge server to process the edge computing task according to the task scheduling decision includes: Using the task processing server to receive corresponding edge computing tasks in sequence; Determine a task receiving order of the plurality of edge computing tasks, and process the edge computing tasks in sequence according to the task receiving order using the task processing server and the task scheduling decision.

8. A task scheduling device for industrial Internet equipment, characterized in that: include: A task information determination module is used to obtain edge computing tasks of each industrial Internet device preset in the industrial Internet within a target time period, and to group the edge computing tasks into corresponding target task sets, and to determine the task information of the edge computing tasks in the target task set; The target time period is a time period determined according to the initial time of scheduling the edge computing task; A server information acquisition module, used to determine a target edge server from the preset industrial Internet, group the target edge servers into a corresponding edge server set, and obtain edge server information of the target edge server in the edge server set; A model building module, used to build a nonlinear optimization model according to preset decision variables, and use the nonlinear optimization model to determine the target correspondence between the edge computing task and the target edge server according to the task information, the edge server information, the preset decision variables and a preset hybrid genetic taboo search algorithm; The preset hybrid genetic taboo search algorithm is an algorithm constructed according to a preset genetic algorithm and a preset taboo search algorithm; The task processing module is used to construct a task scheduling decision corresponding to the edge computing task according to the target corresponding relationship, so as to use the target edge server to process the edge computing task according to the task scheduling decision.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, used to execute the computer program to implement the task scheduling method for industrial Internet devices as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store a computer program, which, when executed by a processor, implements the task scheduling method for industrial Internet devices as described in any one of claims 1 to 7.

Images