A task processing method and system for an industrial Internet platform

Abstract

The present invention relates to an industrial Internet platform, and specifically to a task processing method and system for an industrial Internet platform, which obtains multiple task requests initiated by multiple task initiation terminals, divides the multiple task requests into multiple similar task sets, uses an improved ant colony algorithm to solve task execution plans for each similar task set, optimizes the task execution plans for each similar task set in combination with each task execution terminal, and determines the best task execution plan for each task execution terminal; the technical solution provided by the present invention can effectively overcome the defects of the prior art that it cannot cope with the concurrent execution of multiple tasks and large-batch tasks, and cannot quickly optimize and schedule task requests.

Description

Technical Field

[0001] The present invention relates to an industrial Internet platform, and in particular to a task processing method and system for an industrial Internet platform. Background Art

[0002] The Industrial Internet is a new type of infrastructure, application model and industrial ecology that deeply integrates the new generation of information and communication technology with the industrial economy. Through the comprehensive connection of people, machines, objects and systems, a new manufacturing and service system covering the entire industrial chain and the entire value chain is built. The Industrial Internet is based on the network, the platform is the hub, data is the element, and security is the guarantee. It is not only the infrastructure for the digitalization, networking and intelligent transformation of the industry, but also an application model for the deep integration of the Internet, big data, artificial intelligence and the real economy. It is also a new format and new industry that will reshape the corporate form, supply chain and industrial chain.

[0003] The Industrial Internet platform is a service system based on the Industrial Internet for the collection, aggregation and analysis of massive data, which supports the ubiquitous connection, flexible supply and efficient configuration of manufacturing resources. The Industrial Internet platform applies advanced technologies such as cloud computing, big data, and the Internet of Things to connect massive industrial equipment, sensors, control systems, etc., to achieve interconnection between equipment, real-time data collection and analysis, and intelligent production management.

[0004] However, the existing Industrial Internet platform only realizes the intelligent interconnection and unified control of industrial equipment, and does not consider how to deal with the task processing problems of concurrent execution of multiple tasks and large-scale tasks after the platform is built. It cannot quickly optimize and schedule task requests, resulting in low task processing efficiency of the Industrial Internet platform and inability to meet actual usage needs. Summary of the invention

[0005] 1. Technical issues to be resolved

[0006] In view of the above-mentioned shortcomings of the prior art, the present invention provides a task processing method and system for an industrial Internet platform, which can effectively overcome the defects of the prior art that it is unable to cope with the concurrent execution of multiple tasks and large-batch tasks, and cannot quickly optimize the scheduling of task requests.

[0007] (II) Technical solution

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0009] A task processing method for an industrial Internet platform obtains multiple task requests initiated by multiple task initiating terminals, divides the multiple task requests into multiple similar task sets, uses an improved ant colony algorithm to solve task execution plans for each similar task set, optimizes the task execution plans for each similar task set in combination with each task execution terminal, and determines the best task execution plan for each task execution terminal;

[0010] Among them, in the improved ant colony algorithm, the greedy algorithm is called to preprocess the solution task, and the initial pheromone between each path node is determined according to the preprocessing results. In the process of ant colony movement, the central ant is selected, and the non-central ants move towards the central ant. At the same time, the central ant moves in its own neighborhood by performing local search through hill climbing method, and the pheromone between each path node is dynamically updated by using the adaptive negative feedback dynamic update mechanism.

[0011] Preferably, dividing the multiple task requests into multiple similar task sets includes:

[0012] Determine the task dimension corresponding to each task request according to the task parameter, divide the task requests belonging to the same task dimension into a similar task set, and obtain multiple similar task sets;

[0013] The task parameters include task content information, task time information, task type information and task initiating terminal information, and the task dimension includes one of a time dimension, a space dimension and an environment dimension.

[0014] Preferably, the step of using the improved ant colony algorithm to solve the task execution scheme of each similar task set includes:

[0015] S1. For any similar task set, a minimization function of the execution time and execution cost of multiple task requests in the similar task set is used as an objective function, and an execution constraint condition is determined according to the task parameters of the multiple task requests in the similar task set;

[0016] S2. Based on the objective function and execution constraints, the improved ant colony algorithm is used to solve the task execution plan of the similar task set.

[0017] Preferably, in S2, based on the objective function and the execution constraints, an improved ant colony algorithm is used to solve the task execution plan of the similar task set, including:

[0018] S21, calling the greedy algorithm to preprocess the solution task, determining the initial pheromone between each path node according to the preprocessing result, and initializing the parameters of the improved ant colony algorithm;

[0019] S22, determining a heuristic function for path selection according to task parameters of multiple task requests in a similar task set;

[0020] S23, during the movement of the ant colony, the central ant is selected, and the non-central ants move toward the central ant. Meanwhile, the central ant moves by performing local search in its own neighborhood by means of hill climbing method;

[0021] S24, determining the current optimal task execution plan based on the objective function and the execution constraints, and dynamically updating the pheromones between the path nodes using an adaptive negative feedback dynamic update mechanism;

[0022] S25, repeat S23 to S24 until the maximum number of iterations is reached, and use the current optimal task execution plan as the task execution plan for the similar task set;

[0023] Among them, the parameters of the improved ant colony algorithm include the number of ants, pheromone volatility factor, pheromone importance factor, heuristic function importance factor and the maximum number of iterations.

[0024] Preferably, in S21, a greedy algorithm is called to preprocess the solution task, and the initial pheromone between each path node is determined according to the preprocessing result, including:

[0025] S211, calling the greedy algorithm to pre-process the solution task to obtain multiple groups of feasible task execution solutions;

[0026] The path nodes in each task execution plan obtained by calling the greedy algorithm follow the principle of shortest distance and no duplication, that is, the selection of each path node only considers the nearest path node that meets the requirements, and a path node will not appear in more than two task execution plans;

[0027] S212, calculating the total path distance of each group of task execution plans, and obtaining the target task execution plan with the shortest total path distance;

[0028] S213, using the following formula to calculate the initial pheromone between each path node in the target task execution plan:

[0029]

[0030] Among them, R ab is the initial pheromone between any two preceding and following path nodes a and b in the target task execution plan, R0 is the preset pheromone between all adjacent path nodes, and d ab is the path distance between any two forward and backward path nodes a and b in the target task execution plan;

[0031] S214, using the following formula to calculate the initial pheromone between each path node in the non-target task execution plan:

[0032]

[0033] in, is the initial pheromone between any two nodes m and n on the path before and after the i-th non-target task execution plan, is the path distance between any two preceding and succeeding path nodes m and n in the i-th non-target task execution plan, and ε is the total path distance L of the target task execution plan and the total path distance L of the i-th non-target task execution plan. i The ratio between them, that is, ε=L / L i .

[0034] Preferably, in S23, a central ant is selected during the movement of the ant colony, and the non-central ants move toward the central ant, and the central ant moves by performing a local search in its own neighborhood by a hill climbing method, including:

[0035] S231. Construct the fitness function F of the ant individual based on the objective function:

[0036]

[0037] Where f1'(j) is the execution time function of the first j path nodes in the ant individual moving path to execute the task request in the similar task set, f1 is the unoptimized execution time of the task request in the corresponding similar task set, λ1 is the weight coefficient of the execution time, and s is the number of path nodes in the ant individual moving path;

[0038] f2'(j) is the execution cost function of the first j path nodes in the ant individual moving path to execute the task request in the similar task set, f2 is the unoptimized execution cost of the task request in the corresponding similar task set, and λ2 is the weight coefficient of the execution cost;

[0039] S232, the first central ant is selected according to the fitness function value, the second central ant is selected according to the pheromone, and the movement probability between individual ants is calculated using the following formula:

[0040]

[0041] Among them, p xy is the probability of ant individual x moving to ant individual y, d xy is the distance between ant individual x and ant individual y, F(x) is the fitness function value of ant individual x, F(y) is the fitness function value of ant individual y, and k is a constant;

[0042] S233. Based on the movement principle that non-central ants move toward central ants and central ants move by performing local search in their own neighborhoods by hill climbing, the ant colony performs search movement.

[0043] Preferably, in S24, the pheromone between each path node is dynamically updated by using an adaptive negative feedback dynamic update mechanism, including:

[0044] S241, sorting the multiple task execution schemes obtained by the ant colony search and movement solution according to the total execution time, and sorting the execution time of the same task request in each group of task execution schemes to obtain multiple groups of execution time sequences;

[0045] S242, arbitrarily obtain two groups of execution time series, perform difference processing on each execution time series by subtracting the previous item from the latter item, and obtain two corresponding groups of time difference sequences;

[0046] S243, extract the common sequence in the two groups of time difference sequences, record the position and length of the common sequence, and calculate the similarity S between the two groups of time difference sequences using the following formula:

[0047] S = p·l;

[0048] Among them, p is the position of the common sequence, l is the length of the common sequence, and the two parameters are dimensionless;

[0049] S244, repeat S242 to S243, calculate the similarity between any two groups of execution time sequences in the multiple groups of execution time sequences, and obtain the maximum similarity value S max , similarity minimum value S min , normalize the similarity S, and use the following formula to calculate the normalized similarity S':

[0050]

[0051] S245, calculating the negative feedback dynamic update coefficient, dynamically updating the pheromone between each path node based on the negative feedback dynamic update coefficient and the fitness function F of the ant individual, and calculating the negative feedback dynamic update coefficient λ using the following formula:

[0052] λ=1-S′.

[0053] Preferably, the step of optimizing the task execution schemes of each similar task set in combination with each task execution terminal to determine the best task execution scheme of each task execution terminal includes:

[0054] The corresponding task dimension is determined according to the characteristics of each task execution terminal, the task execution plan of each similar task set is optimized according to the task dimension of each task execution terminal, and the best task execution plan of each task execution terminal is determined.

[0055] A task processing system for an industrial Internet platform, comprising a task request acquisition module, a task request division module, an improved ant colony algorithm construction module, an improved ant colony algorithm execution module and a task execution scheme optimization module;

[0056] A task request acquisition module is used to acquire multiple task requests initiated by multiple task initiating terminals;

[0057] A task request partitioning module partitions multiple task requests into multiple similar task sets;

[0058] Improve the building module of the ant colony algorithm. On the basis of the traditional ant colony algorithm, the greedy algorithm is called to preprocess the solution task. The initial pheromone between each path node is determined according to the preprocessing result. The central ant is selected in the process of ant colony movement, and the non-central ants move towards the central ant. At the same time, the central ant moves in its own neighborhood by local search through hill climbing method, and the pheromone between each path node is dynamically updated by using the adaptive negative feedback dynamic update mechanism.

[0059] Improve the ant colony algorithm execution module, and use the improved ant colony algorithm to solve the task execution plan of each similar task set;

[0060] The task execution plan optimization module optimizes the task execution plans of each similar task set in combination with each task execution terminal, and determines the best task execution plan for each task execution terminal.

[0061] Preferably, the improved ant colony algorithm execution module uses the improved ant colony algorithm to solve the task execution plan of each similar task set, including:

[0062] S1. For any similar task set, a minimization function of the execution time and execution cost of multiple task requests in the similar task set is used as an objective function, and an execution constraint condition is determined according to the task parameters of the multiple task requests in the similar task set;

[0063] S2. Based on the objective function and execution constraints, the improved ant colony algorithm is used to solve the task execution plan of the similar task set, which specifically includes:

[0064] S21, calling the greedy algorithm to preprocess the solution task, determining the initial pheromone between each path node according to the preprocessing result, and initializing the parameters of the improved ant colony algorithm;

[0065] S22, determining a heuristic function for path selection according to task parameters of multiple task requests in a similar task set;

[0066] S23, during the movement of the ant colony, the central ant is selected, and the non-central ants move toward the central ant. Meanwhile, the central ant moves by performing local search in its own neighborhood by means of hill climbing method;

[0067] S24, determining the current optimal task execution plan based on the objective function and the execution constraints, and dynamically updating the pheromones between the path nodes using an adaptive negative feedback dynamic update mechanism;

[0068] S25, repeat S23 to S24 until the maximum number of iterations is reached, and use the current optimal task execution plan as the task execution plan for the similar task set;

[0069] Among them, the parameters of the improved ant colony algorithm include the number of ants, pheromone volatility factor, pheromone importance factor, heuristic function importance factor and the maximum number of iterations.

[0070] (III) Beneficial effects

[0071] Compared with the prior art, the task processing method and system of the industrial Internet platform provided by the present invention has the following beneficial effects:

[0072] 1. Obtain multiple task requests initiated by multiple task initiation terminals, and divide the multiple task requests into multiple similar task sets. Use the improved ant colony algorithm to solve the task execution plan of each similar task set, and optimize the task execution plan of each similar task set in combination with each task execution terminal to determine the best task execution plan for each task execution terminal, so as to achieve effective optimization and scheduling of task requests, improve the ability of the industrial Internet platform to cope with the concurrent execution of multiple tasks and large-scale tasks, and effectively improve the task processing efficiency of the industrial Internet platform;

[0073] 2. In the technical solution of this application, the traditional ant colony algorithm is improved in the following three aspects:

[0074] 1) Call the greedy algorithm to preprocess the solution task, and determine the initial pheromone between each path node according to the preprocessing result;

[0075] 2) During the movement of the ant colony, the central ant is selected and the non-central ants move towards the central ant. At the same time, the central ant moves by performing local search in its own neighborhood through the hill climbing method;

[0076] 3) Using adaptive negative feedback dynamic update mechanism to dynamically update the pheromones between each path node;

[0077] Based on the traditional ant colony algorithm, an improved ant colony algorithm with faster convergence speed is constructed by combining the greedy algorithm, hill climbing method and adaptive negative feedback dynamic update mechanism. It can realize rapid optimization scheduling of task requests and further enhance the ability of the industrial Internet platform to cope with the concurrent execution of multiple tasks and large-scale tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0079] Figure 1 It is a schematic diagram of the process of the present invention;

[0080] Figure 2 It is a detailed schematic diagram of the process of the present invention;

[0081] Figure 3 It is a system schematic diagram of the present invention. DETAILED DESCRIPTION

[0082] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0083] A task processing method for an industrial Internet platform, such as Figure 1 and Figure 2 As shown, multiple task requests initiated by multiple task initiating terminals are obtained, and the multiple task requests are divided into multiple similar task sets, and the task execution plan of each similar task set is solved by using the improved ant colony algorithm, and the task execution plan of each similar task set is optimized in combination with each task execution terminal to determine the best task execution plan of each task execution terminal;

[0084] Among them, in the improved ant colony algorithm, the greedy algorithm is called to preprocess the solution task, and the initial pheromone between each path node is determined according to the preprocessing results. In the process of ant colony movement, the central ant is selected, and the non-central ants move towards the central ant. At the same time, the central ant moves in its own neighborhood by performing local search through hill climbing method, and the pheromone between each path node is dynamically updated by using the adaptive negative feedback dynamic update mechanism.

[0085] I. Divide multiple task requests into multiple similar task sets, including:

[0086] Determine the task dimension corresponding to each task request according to the task parameter, divide the task requests belonging to the same task dimension into a similar task set, and obtain multiple similar task sets;

[0087] The task parameters include task content information, task time information, task type information and task initiating terminal information, and the task dimension includes one of a time dimension, a space dimension and an environment dimension.

[0088] II. Using the improved ant colony algorithm to solve the task execution plan for each similar task set, including:

[0089] S1. For any similar task set, a minimization function of the execution time and execution cost of multiple task requests in the similar task set is used as an objective function, and an execution constraint condition is determined according to the task parameters of the multiple task requests in the similar task set;

[0090] S2. Based on the objective function and execution constraints, the improved ant colony algorithm is used to solve the task execution plan of the similar task set.

[0091] Specifically, based on the objective function and execution constraints, the improved ant colony algorithm is used to solve the task execution plan of the similar task set, including:

[0092] S21, calling the greedy algorithm to preprocess the solution task, determining the initial pheromone between each path node according to the preprocessing result, and initializing the parameters of the improved ant colony algorithm;

[0093] S22, determining a heuristic function for path selection according to task parameters of multiple task requests in a similar task set;

[0094] S23, during the movement of the ant colony, the central ant is selected, and the non-central ants move toward the central ant. Meanwhile, the central ant moves by performing local search in its own neighborhood by means of hill climbing method;

[0095] S24, determining the current optimal task execution plan based on the objective function and the execution constraints, and dynamically updating the pheromones between the path nodes using an adaptive negative feedback dynamic update mechanism;

[0096] S25, repeat S23 to S24 until the maximum number of iterations is reached, and use the current optimal task execution plan as the task execution plan for the similar task set;

[0097] Among them, the parameters of the improved ant colony algorithm include the number of ants, pheromone volatility factor, pheromone importance factor, heuristic function importance factor and the maximum number of iterations.

[0098] 1) In S21, the greedy algorithm is called to preprocess the solution task, and the initial pheromone between each path node is determined according to the preprocessing result, including:

[0099] S211, calling the greedy algorithm to pre-process the solution task to obtain multiple groups of feasible task execution solutions;

[0100] The path nodes in each task execution plan obtained by calling the greedy algorithm follow the principle of shortest distance and no duplication, that is, the selection of each path node only considers the nearest path node that meets the requirements, and a path node will not appear in more than two task execution plans;

[0101] S212, calculating the total path distance of each group of task execution plans, and obtaining the target task execution plan with the shortest total path distance;

[0102] S213, using the following formula to calculate the initial pheromone between each path node in the target task execution plan:

[0103]

[0104] Among them, R ab is the initial pheromone between any two preceding and following path nodes a and b in the target task execution plan, R0 is the preset pheromone between all adjacent path nodes, and d ab is the path distance between any two forward and backward path nodes a and b in the target task execution plan;

[0105] S214, using the following formula to calculate the initial pheromone between each path node in the non-target task execution plan:

[0106]

[0107] in, is the initial pheromone between any two nodes m and n on the path before and after the i-th non-target task execution plan, is the path distance between any two preceding and succeeding path nodes m and n in the i-th non-target task execution plan, and ε is the total path distance L of the target task execution plan and the total path distance L of the i-th non-target task execution plan. i The ratio between them, that is, ε=L / L i .

[0108] 2) In S23, the central ant is selected during the movement of the ant colony, and the non-central ants move toward the central ant. At the same time, the central ant moves by performing local search in its own neighborhood by hill climbing, including:

[0109] S231. Construct the fitness function F of the ant individual based on the objective function:

[0110]

[0111] Where f1'(j) is the execution time function of the first j path nodes in the ant individual moving path to execute the task request in the similar task set, f1 is the unoptimized execution time of the task request in the corresponding similar task set, λ1 is the weight coefficient of the execution time, and s is the number of path nodes in the ant individual moving path;

[0112] f2'(j) is the execution cost function of the first j path nodes in the ant individual moving path to execute the task request in the similar task set, f2 is the unoptimized execution cost of the task request in the corresponding similar task set, and λ2 is the weight coefficient of the execution cost;

[0113] S232, the first central ant is selected according to the fitness function value, the second central ant is selected according to the pheromone, and the movement probability between individual ants is calculated using the following formula:

[0114]

[0115] Among them, p xy is the probability of ant individual x moving to ant individual y, d xy is the distance between ant individual x and ant individual .y, F(x) is the fitness function value of ant individual x, F(y) is the fitness function value of ant individual .y, and k is a constant;

[0116] S233. Based on the movement principle that non-central ants move toward central ants and central ants move by performing local search in their own neighborhoods by hill climbing, the ant colony performs search movement.

[0117] 3) In S24, the pheromone between each path node is dynamically updated by using an adaptive negative feedback dynamic update mechanism, including:

[0118] S241, sorting the multiple task execution schemes obtained by the ant colony search and movement solution according to the total execution time, and sorting the execution time of the same task request in each group of task execution schemes to obtain multiple groups of execution time sequences;

[0119] S242, arbitrarily obtain two groups of execution time series, perform difference processing on each execution time series by subtracting the previous item from the latter item, and obtain two corresponding groups of time difference sequences;

[0120] S243, extract the common sequence in the two groups of time difference sequences, record the position and length of the common sequence, and calculate the similarity S between the two groups of time difference sequences using the following formula:

[0121] S = p·l;

[0122] Among them, p is the position of the common sequence, l is the length of the common sequence, and the two parameters are dimensionless;

[0123] S244, repeat S242 to S243, calculate the similarity between any two groups of execution time sequences in the multiple groups of execution time sequences, and obtain the maximum similarity value S max , similarity minimum value S min , normalize the similarity S, and use the following formula to calculate the normalized similarity S':

[0124]

[0125] S245, calculating the negative feedback dynamic update coefficient, dynamically updating the pheromone between each path node based on the negative feedback dynamic update coefficient and the fitness function F of the ant individual, and calculating the negative feedback dynamic update coefficient λ using the following formula:

[0126] λ=1-S′.

[0127] III. Optimize the task execution plan of each similar task set in combination with each task execution terminal, and determine the best task execution plan for each task execution terminal, including:

[0128] The corresponding task dimension is determined according to the characteristics of each task execution terminal, the task execution plan of each similar task set is optimized according to the task dimension of each task execution terminal, and the best task execution plan of each task execution terminal is determined.

[0129] In the technical solution of the present application, multiple task requests initiated by multiple task initiation terminals are obtained, and the multiple task requests are divided into multiple similar task sets. The task execution plan of each similar task set is solved by using an improved ant colony algorithm. The task execution plan of each similar task set is optimized in combination with each task execution terminal, and the optimal task execution plan of each task execution terminal is determined, thereby achieving effective optimization and scheduling of task requests, improving the ability of the industrial Internet platform to cope with the concurrent execution of multiple tasks and large-scale tasks, and effectively improving the task processing efficiency of the industrial Internet platform.

[0130] In the technical solution of this application, the traditional ant colony algorithm is improved in the following three aspects:

[0131] 1) Call the greedy algorithm to preprocess the solution task, and determine the initial pheromone between each path node according to the preprocessing result;

[0132] 2) During the movement of the ant colony, the central ant is selected and the non-central ants move towards the central ant. At the same time, the central ant moves by performing local search in its own neighborhood through the hill climbing method;

[0133] 3) Using adaptive negative feedback dynamic update mechanism to dynamically update the pheromones between each path node;

[0134] Based on the traditional ant colony algorithm, an improved ant colony algorithm with faster convergence speed is constructed by combining the greedy algorithm, hill climbing method and adaptive negative feedback dynamic update mechanism. It can realize rapid optimization scheduling of task requests and further enhance the ability of the industrial Internet platform to cope with the concurrent execution of multiple tasks and large-scale tasks.

[0135] In the technical solution of this application, a task processing system of an industrial Internet platform is also disclosed, such as Figure 3 As shown, it includes a task request acquisition module, a task request division module, an improved ant colony algorithm construction module, an improved ant colony algorithm execution module and a task execution plan optimization module;

[0136] A task request acquisition module is used to acquire multiple task requests initiated by multiple task initiating terminals;

[0137] A task request partitioning module partitions multiple task requests into multiple similar task sets;

[0138] Improve the building module of the ant colony algorithm. On the basis of the traditional ant colony algorithm, the greedy algorithm is called to preprocess the solution task. The initial pheromone between each path node is determined according to the preprocessing result. The central ant is selected in the process of ant colony movement, and the non-central ants move towards the central ant. At the same time, the central ant moves in its own neighborhood by local search through hill climbing method, and the pheromone between each path node is dynamically updated by using the adaptive negative feedback dynamic update mechanism.

[0139] Improve the ant colony algorithm execution module, and use the improved ant colony algorithm to solve the task execution plan of each similar task set;

[0140] The task execution plan optimization module optimizes the task execution plans of each similar task set in combination with each task execution terminal, and determines the best task execution plan for each task execution terminal.

[0141] Specifically, the improved ant colony algorithm execution module uses the improved ant colony algorithm to solve the task execution plan of each similar task set, including:

[0142] S1. For any similar task set, a minimization function of the execution time and execution cost of multiple task requests in the similar task set is used as an objective function, and an execution constraint condition is determined according to the task parameters of the multiple task requests in the similar task set;

[0143] S2. Based on the objective function and execution constraints, the improved ant colony algorithm is used to solve the task execution plan of the similar task set, which specifically includes:

[0144] S21, calling the greedy algorithm to preprocess the solution task, determining the initial pheromone between each path node according to the preprocessing result, and initializing the parameters of the improved ant colony algorithm;

[0145] S22, determining a heuristic function for path selection according to task parameters of multiple task requests in a similar task set;

[0146] S23, during the movement of the ant colony, the central ant is selected, and the non-central ants move toward the central ant. Meanwhile, the central ant moves by performing local search in its own neighborhood by means of hill climbing method;

[0147] S24, determining the current optimal task execution plan based on the objective function and the execution constraints, and dynamically updating the pheromones between the path nodes using an adaptive negative feedback dynamic update mechanism;

[0148] S25, repeat S23 to S24 until the maximum number of iterations is reached, and use the current optimal task execution plan as the task execution plan for the similar task set;

[0149] Among them, the parameters of the improved ant colony algorithm include the number of ants, pheromone volatility factor, pheromone importance factor, heuristic function importance factor and the maximum number of iterations.

[0150] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A task processing method for an industrial Internet platform, characterized in that: Acquire multiple task requests initiated by multiple task initiating terminals, divide the multiple task requests into multiple similar task sets, use the improved ant colony algorithm to solve the task execution plan of each similar task set, optimize the task execution plan of each similar task set in combination with each task execution terminal, and determine the best task execution plan of each task execution terminal; Among them, in the improved ant colony algorithm, the greedy algorithm is called to preprocess the solution task, and the initial pheromone between each path node is determined according to the preprocessing result. In the process of ant colony movement, the central ant is selected, and the non-central ants move towards the central ant. At the same time, the central ant moves in its own neighborhood by local search through the hill climbing method, and the pheromone between each path node is dynamically updated by using the adaptive negative feedback dynamic update mechanism. The process of selecting a central ant during the movement of the ant colony, and the non-central ants moving toward the central ant, while the central ant moving by performing a local search in its own neighborhood by means of a hill climbing method, includes: S231. Construct the fitness function F of the ant individual based on the objective function: Where f1'(j) is the execution time function of the first j path nodes in the ant individual moving path to execute the task request in the similar task set, f1 is the unoptimized execution time of the task request in the corresponding similar task set, λ1 is the weight coefficient of the execution time, and s is the number of path nodes in the ant individual moving path; f2'(j) is the execution cost function of the first j path nodes in the ant individual moving path to execute the task request in the similar task set, f2 is the unoptimized execution cost of the task request in the corresponding similar task set, and λ2 is the weight coefficient of the execution cost; S232, the first central ant is selected according to the fitness function value, the second central ant is selected according to the pheromone, and the movement probability between individual ants is calculated using the following formula: Among them, p xy is the probability of ant individual x moving to ant individual y, d xy is the distance between ant individual x and ant individual y, F(x) is the fitness function value of ant individual x, F(y) is the fitness function value of ant individual y, and k is a constant; S233. Based on the movement principle that non-central ants move toward central ants and central ants move by performing local search in their own neighborhoods by hill climbing, the ant colony performs search movement.

2. The task processing method of the industrial Internet platform according to claim 1 is characterized in that: The dividing of the plurality of task requests into a plurality of similar task sets comprises: Determine the task dimension corresponding to each task request according to the task parameter, divide the task requests belonging to the same task dimension into a similar task set, and obtain multiple similar task sets; The task parameters include task content information, task time information, task type information and task initiating terminal information, and the task dimension includes one of a time dimension, a space dimension and an environment dimension.

3. The task processing method of the industrial Internet platform according to claim 1 is characterized in that: The improved ant colony algorithm is used to solve the task execution scheme of each similar task set, including: S1. For any similar task set, a minimization function of the execution time and execution cost of multiple task requests in the similar task set is used as an objective function, and an execution constraint condition is determined according to the task parameters of the multiple task requests in the similar task set; S2. Based on the objective function and execution constraints, the improved ant colony algorithm is used to solve the task execution plan of the similar task set.

4. The task processing method of the industrial Internet platform according to claim 3 is characterized in that: In S2, based on the objective function and execution constraints, the improved ant colony algorithm is used to solve the task execution plan of the similar task set, including: S21, calling the greedy algorithm to preprocess the solution task, determining the initial pheromone between each path node according to the preprocessing result, and initializing the parameters of the improved ant colony algorithm; S22, determining a heuristic function for path selection according to task parameters of multiple task requests in a similar task set; S23, during the movement of the ant colony, the central ant is selected, and the non-central ants move toward the central ant. Meanwhile, the central ant moves by performing local search in its own neighborhood by means of hill climbing method; S24, determining the current optimal task execution plan based on the objective function and the execution constraints, and dynamically updating the pheromones between the path nodes using an adaptive negative feedback dynamic update mechanism; S25, repeat S23 to S24 until the maximum number of iterations is reached, and use the current optimal task execution plan as the task execution plan for the similar task set; Among them, the parameters of the improved ant colony algorithm include the number of ants, pheromone volatility factor, pheromone importance factor, heuristic function importance factor and the maximum number of iterations.

5. The task processing method of the industrial Internet platform according to claim 4 is characterized in that: In S21, the greedy algorithm is called to preprocess the solution task, and the initial pheromone between each path node is determined according to the preprocessing result, including: S211, calling the greedy algorithm to pre-process the solution task to obtain multiple groups of feasible task execution solutions; The path nodes in each task execution plan obtained by calling the greedy algorithm follow the principle of shortest distance and no duplication, that is, the selection of each path node only considers the nearest path node that meets the requirements, and a path node will not appear in more than two task execution plans; S212, calculating the total path distance of each group of task execution plans, and obtaining the target task execution plan with the shortest total path distance; S213, using the following formula to calculate the initial pheromone between each path node in the target task execution plan: Among them, R ab is the initial pheromone between any two preceding and following path nodes a and b in the target task execution plan, R0 is the preset pheromone between all adjacent path nodes, and d ab is the path distance between any two forward and backward path nodes a and b in the target task execution plan; S214, using the following formula to calculate the initial pheromone between each path node in the non-target task execution plan: in, is the initial pheromone between any two nodes m and n on the path before and after the i-th non-target task execution plan, is the path distance between any two preceding and succeeding path nodes m and n in the i-th non-target task execution plan, and ε is the total path distance L of the target task execution plan and the total path distance L of the i-th non-target task execution plan. i The ratio between them, that is, ε=L / L i .

6. The task processing method of the industrial Internet platform according to claim 4 is characterized in that: In S24, the pheromone between each path node is dynamically updated by using an adaptive negative feedback dynamic update mechanism, including: S241, sorting the multiple task execution schemes obtained by the ant colony search and movement solution according to the total execution time, and sorting the execution time of the same task request in each group of task execution schemes to obtain multiple groups of execution time sequences; S242, arbitrarily obtain two groups of execution time series, perform difference processing on each execution time series by subtracting the previous item from the latter item, and obtain two corresponding groups of time difference sequences; S243, extract the common sequence in the two groups of time difference sequences, record the position and length of the common sequence, and calculate the similarity S between the two groups of time difference sequences using the following formula: S = p·l; Among them, p is the position of the common sequence, l is the length of the common sequence, and the two parameters are dimensionless; S244, repeat S242 to S243, calculate the similarity between any two groups of execution time sequences in the multiple groups of execution time sequences, and obtain the maximum similarity value S max , similarity minimum value S min , normalize the similarity S, and use the following formula to calculate the normalized similarity S': S245, calculating the negative feedback dynamic update coefficient, dynamically updating the pheromone between each path node based on the negative feedback dynamic update coefficient and the fitness function F of the ant individual, and calculating the negative feedback dynamic update coefficient λ using the following formula: λ=1-S'.

7. The task processing method of the industrial Internet platform according to claim 1 is characterized in that: The step of optimizing the task execution schemes of each similar task set in combination with each task execution terminal to determine the best task execution scheme for each task execution terminal includes: The corresponding task dimension is determined according to the characteristics of each task execution terminal, the task execution plan of each similar task set is optimized according to the task dimension of each task execution terminal, and the best task execution plan of each task execution terminal is determined.

8. A task processing system for an industrial Internet platform, used to execute the task processing method for an industrial Internet platform according to claim 1, characterized in that: It includes a task request acquisition module, a task request division module, an improved ant colony algorithm construction module, an improved ant colony algorithm execution module and a task execution plan optimization module; A task request acquisition module is used to acquire multiple task requests initiated by multiple task initiating terminals; A task request partitioning module partitions multiple task requests into multiple similar task sets; Improve the building module of the ant colony algorithm. On the basis of the traditional ant colony algorithm, the greedy algorithm is called to preprocess the solution task. The initial pheromone between each path node is determined according to the preprocessing result. The central ant is selected in the process of ant colony movement, and the non-central ants move towards the central ant. At the same time, the central ant moves in its own neighborhood by local search through hill climbing method, and the pheromone between each path node is dynamically updated by using the adaptive negative feedback dynamic update mechanism. Improve the ant colony algorithm execution module, and use the improved ant colony algorithm to solve the task execution plan of each similar task set; The task execution plan optimization module optimizes the task execution plans of each similar task set in combination with each task execution terminal, and determines the best task execution plan for each task execution terminal; The process of selecting a central ant during the movement of the ant colony, and the non-central ants moving toward the central ant, while the central ant moving by performing a local search in its own neighborhood by means of a hill climbing method, includes: S231. Construct the fitness function F of the ant individual based on the objective function: Where f1'(j) is the execution time function of the first j path nodes in the ant individual moving path to execute the task request in the similar task set, f1 is the unoptimized execution time of the task request in the corresponding similar task set, λ1 is the weight coefficient of the execution time, and s is the number of path nodes in the ant individual moving path; f2'(j) is the execution cost function of the first j path nodes in the ant individual moving path to execute the task request in the similar task set, f2 is the unoptimized execution cost of the task request in the corresponding similar task set, and λ2 is the weight coefficient of the execution cost; S232, the first central ant is selected according to the fitness function value, the second central ant is selected according to the pheromone, and the movement probability between individual ants is calculated using the following formula: Among them, p xy is the probability of ant individual x moving to ant individual y, d xy is the distance between ant individual x and ant individual y, F(x) is the fitness function value of ant individual x, F(y) is the fitness function value of ant individual y, and k is a constant; S233. Based on the movement principle that non-central ants move toward central ants and central ants move by performing local search in their own neighborhoods by hill climbing, the ant colony performs search movement.

9. The task processing system of the industrial Internet platform according to claim 8, characterized in that: The improved ant colony algorithm execution module uses the improved ant colony algorithm to solve the task execution plan of each similar task set, including: S1. For any similar task set, a minimization function of the execution time and execution cost of multiple task requests in the similar task set is used as an objective function, and an execution constraint condition is determined according to the task parameters of the multiple task requests in the similar task set; S2. Based on the objective function and execution constraints, the improved ant colony algorithm is used to solve the task execution plan of the similar task set, which specifically includes: S21, calling the greedy algorithm to preprocess the solution task, determining the initial pheromone between each path node according to the preprocessing result, and initializing the parameters of the improved ant colony algorithm; S22, determining a heuristic function for path selection according to task parameters of multiple task requests in a similar task set; S23, during the movement of the ant colony, the central ant is selected, and the non-central ants move toward the central ant. Meanwhile, the central ant moves by performing local search in its own neighborhood by means of hill climbing method; S24, determining the current optimal task execution plan based on the objective function and the execution constraints, and dynamically updating the pheromones between the path nodes using an adaptive negative feedback dynamic update mechanism; S25, repeat S23 to S24 until the maximum number of iterations is reached, and use the current optimal task execution plan as the task execution plan for the similar task set; Among them, the parameters of the improved ant colony algorithm include the number of ants, pheromone volatility factor, pheromone importance factor, heuristic function importance factor and the maximum number of iterations.

Images