Machine tool equipment cloud service dynamic optimization configuration method with idle time window

Abstract

The invention discloses a machine tool equipment cloud service dynamic optimization configuration method with an idle time window, and relates to the technical field of cloud manufacturing. According to the invention, by analyzing the focus of a machine tool equipment resource supplier in cloud service cooperation, a multi-target optimization configuration model giving consideration to service benefits and service risks under the constraint of a machine tool idle time window is constructed; according to the machine tool equipment cloud service dynamic optimization configuration method with the idle time window, the problem of comprehensive quantification and collaborative optimization of cloud manufacturing task benefits and risks in minute-level fragmentary time periods is solved, the matching process of cloud manufacturing service resources is divided into a combination stage and an optimization stage, and in the combination stage, the cloud manufacturing task benefits and risks are optimized in a collaborative mode. Generating all feasible candidate subtask combination schemes based on an idle time window of the machine tool; in the optimization stage, an improved GOOSE-ESC algorithm is used for solving the feasible schemes, and a global optimal task scheduling scheme with the maximum service benefit and the minimum service risk is obtained.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cloud manufacturing, in particular to a machine tool equipment cloud service dynamic optimization configuration method with idle time window. BACKGROUND

[0002] Cloud manufacturing (CMfg) is a new networked manufacturing mode, adhering to the core concept of "centralized use of distributed resources and services", and has been widely applied in many fields. For example, in the mechanical processing industry, cloud manufacturing platforms can realize the sharing and collaboration of idle machine tool resources, and provide personalized manufacturing services for demand enterprises. Through the cloud manufacturing mode, small and medium-sized enterprises can make full use of idle high-precision machine tool equipment resources, improve manufacturing capacity and market competitiveness.

[0003] With the in-depth and parallel development of various production process cloud services, there is a possibility of sharing an MCS (Manufacturing Control System) among multiple cloud services, which makes the combination of SDNs (Software-defined Networking) form a larger CDN (Complex Dynamic Network), that is, the transformation of SDN to CDN. The nodes of the CDN network are composed of machine tool equipment resources participating in multiple cloud service tasks SDN. Starting from a single machine tool equipment resource, the profile of the machine tool equipment resource CDN network under the cloud manufacturing environment is the ordering combination of cloud service tasks based on time series. When the service provider optimizes the cloud service tasks for ordering combination, the service revenue, service risk and other optimization objectives need to be considered, and the service task time limit and the idle time limit and ordering combination limit of the machine tool equipment resource also need to be considered. It is a typical strong NP-hard problem (Non-deterministic Polynomial).

[0004] However, in the current cloud manufacturing environment, the collaborative sharing effect of machine tool equipment is not ideal. The existing cloud manufacturing resource sharing mainly focuses on the utilization of idle resources, and the utilization rate of machine tool equipment in the spare time period after occupation has not been fully utilized.

[0005] Therefore, a new solution is needed to solve the above problems. SUMMARY

[0006] The purpose of the present application is to provide a machine tool equipment cloud service dynamic optimization configuration method with idle time window, which breaks through the bottleneck of easy falling into local optimum and low optimization efficiency of the prior art in solving such complex scheduling problems, thereby improving the utilization rate of machine tool equipment resources in the idle time period, to solve the technical problems proposed in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for dynamic optimization configuration of machine tool equipment cloud services with idle time windows, comprising at least the following steps:

[0008] S1: Processing orders from cloud manufacturing demanders. Machine tool services from cloud manufacturing resource providers The information is input into the service demand task pool and machine tool resource pool of the cloud manufacturing platform respectively. The focus of machine tool equipment resource suppliers in choosing to cooperate with cloud services is analyzed, and the cloud manufacturing service resources are matched according to the idle time window into two stages: combination and optimization.

[0009] S2: Combination Phase: Utilizing the idle time window of the machine tool equipment Matching candidate cloud manufacturing subtasks The tasks are sorted and combined to form a set of multi-subtask combination schemes for all machine tool services. ;

[0010] S3: Optimization Phase: Based on the availability benefits of machine tool equipment in providing cloud services and the risks associated with executing cloud manufacturing tasks, a multi-objective optimization configuration model is constructed, where availability benefits refer to service revenue. The risks constituted are service risks. For multi-subtask combination schemes The GOOSE-ESC algorithm is used to solve the multi-objective optimization configuration model in order to maximize service benefits and minimize service risks, and select the optimal configuration scheme with the greatest global benefits.

[0011] S4: Obtain the processing order task and the machine tool service from the cloud manufacturing resource provider. Optimized configuration scheme;

[0012] S5: Implement the optimized configuration scheme in the machine tool equipment in the machining workshop to make full use of the idle time window of the machine tool equipment. This will improve resource utilization and corporate profits.

[0013] Furthermore, S2 includes at least the following steps:

[0014] S201: Obtain all candidate cloud manufacturing tasks Next task Processing volume Single piece processing time Earliest start time and latest end time and update all machine tool services. idle time window ;

[0015] S202: Find all machine service feasible solution domain under the idle time window ; ;

[0016] S203: For each machine service , randomly select a subtask from the feasible solution domain each time form a combination scheme, and update the idle time window ;

[0017] S204: Match the subtasks with the machine service one by one, and classify the combination of subtasks into each service to form a possible task scheduling scheme, i.e. repeat steps S201 to S203 until the feasible solution domain of the machine service is empty, i.e. for the machine service , , ;

[0018] S205: After selecting the subtasks one by one, obtain the combination of subtasks that match the candidate solutions for each machine service idle time window .

[0019] Further, the derivation process of S205 includes at least the following steps:

[0020] All matching cloud manufacturing subtasks of the machine service should be within the idle time window , in addition, when the subtask matches the machine service , it belongs to the feasible solution domain of the machine service under the idle time window , denoted as: ( );

[0021] When the latest start time and the earliest end time of the subtask should satisfy and , the calculation formula is:

[0022]

[0023]

[0024] ​

[0025]

[0026] in, Indicates the end time of the idle time window for machine tool services; For subtasks Time requirements; It is a coefficient; The planned start time; The planned completion time; and These are machine tool services The start and end times of the idle time window;

[0027] Ultimately, the idle time window for each machine tool service is obtained. The combined subtask set of matching candidate solutions .

[0028] Furthermore, the multi-objective optimization configuration model in S3 is expressed as follows:

[0029]

[0030] in:

[0031]

[0032]

[0033] Where, expression , The objective function is denoted as .

[0034] Furthermore, the application of the GOOSE-ESC algorithm in S3 includes at least the following steps:

[0035] S301: Input all candidate cloud manufacturing tasks Next task The machine tool service Information and the combined subtask set ;

[0036] S302: Machine tool service as described in the enterprise workshop The parameters of the GOOSE-ESC algorithm are initialized by state initialization, and the parameters of the GOOSE-ESC algorithm include population size. Total number of iterations Number of machine tools Number of tasks Current iteration number ;

[0037] S303: initialize the population with the low-discrepancy sequence and based on the machine service Initialize the initialized population with dynamic coding;

[0038] S304: Calculate the fitness value of the initialized population;

[0039] S305: and according to the dominance level and crowding distance, the goose population is non-dominated sorted;

[0040] S306: select the development stage and exploration stage for iteration by judging the size of the random probability value, when less than 0.9, enter S307, and greater than 0.9, enter S308;

[0041] S307: in the development stage, by judging the weight of the randomly generated stone , select different iteration methods in different individuals of geese;

[0042] S308: in the exploration stage, by recalculating the time required for the stone to fall , further find the position of the best guard goose;

[0043] S309: introduce attack strategy, through random behavior coefficient, simulate the escape behavior of goose group after the goose group is awakened by the guard goose, and further explore the feasible solution in the space, the escape behavior divides the goose group into follow-the-crowd group and panic group;

[0044] S310: use elite reverse learning strategy to perform reverse search on the elite individuals after iteration, the elite individuals are the top 10% of geese in escape speed;

[0045] S311: according to the dominance level and crowding distance of the individual, the goose population is non-dominated sorted and the best solution in the non-dominated sorting is outputted;

[0046] S312: set , repeat steps S306 to S311 until the termination condition is met, at this time the obtained Pareto front is the candidate solution set of the machine tool equipment cloud service multi-objective optimization configuration model based on the idle time window;

[0047] S313: output the optimal candidate solution set;

[0048] S314: end.

[0049] Further, the GOOSE-ESC algorithm adopts a low-discrepancy sequence initialization strategy, generates an initial population using Halton sequence, and maps the initial point from the unit hypercube to the actual search space to improve the uniform distribution of the initial solution in space;

[0050] The low-difference sequence initialization strategy comprises at least the following steps:

[0051] First, Halton sequence is used to generate initial points with spatial uniformity;

[0052] Then, the points are mapped from the unit hypercube to the actual search space;

[0053] Finally, the mapped lattice points are used as the initial population to provide a set of uniformly distributed initial solutions for the optimization algorithm.

[0054] Further, the GOOSE-ESC algorithm adopts a dynamic encoding mechanism based on machine tool cloud service, which comprises at least the following steps:

[0055] First, the population individuals record the task number, processing procedure and machine tool information through an N-dimensional matrix;

[0056] Then, in the decoding process, it is judged whether the remaining idle time of a machine tool is sufficient to complete the assigned task;

[0057] Finally, when the idle time is insufficient, a dynamic recombination mechanism is triggered to reassign the manufacturing task that cannot be completed to other machine tools with sufficient excess idle time.

[0058] Further, the GOOSE-ESC algorithm adopts a two-stage collaborative optimization update strategy:

[0059] The first stage is the GOOSE development and exploration stage: according to the random probability judgment, if the condition is met, the search based on the goose alert behavior is executed, and the moving distance and direction of the individual are calculated;

[0060] The second stage is the ESC escape stage: if the first stage condition is not met, an attack mechanism is introduced to simulate the crowd evacuation behavior, and the position is updated in the panic group and the follow-the-crowd group to increase the population diversity.

[0061] Further, the GOOSE-ESC algorithm includes an elite reverse learning strategy;

[0062] After each iteration is completed, the elite individual, i.e. the best alert goose, is selected to generate its reverse solution;

[0063] If the fitness of the reverse solution is better than that of the original solution, the original elite solution is replaced to jump out of the local optimum.

[0064] Compared with the prior art, the present application has the following advantages:

[0065] 1. The application constructs a multi-objective optimization configuration model under the constraint of machine tool idle time window, which takes into account the service benefit and service risk, by analyzing the concerns of machine tool equipment resource suppliers in cloud service cooperation, solves the problem of comprehensive quantification and collaborative optimization of cloud manufacturing task benefit and risk in the minute-level fragmented time period.

[0066] 2. The machine tool equipment cloud service dynamic optimization configuration method with idle time window proposed in the application divides the matching process of cloud manufacturing service resources into two stages of combination and optimization. In the combination stage, all feasible candidate sub-task combination schemes are generated based on the idle time window of the machine tool. In the optimization stage, an improved GOOSE-ESC algorithm is used to solve from these feasible schemes to obtain the global optimal task scheduling scheme with maximum service benefit and minimum service risk, thereby realizing the fine utilization of machine tool fragmented idle time and effectively improving the resource utilization and enterprise income. BRIEF DESCRIPTION OF DRAWINGS

[0067] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed for the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0068] Figure 1 The flowchart of the cloud manufacturing service combination stage of the application;

[0069] Figure 2 The machine tool equipment cloud service combination flowchart with idle time window of the application;

[0070] Figure 3 The flowchart of the GOOSE-ESC algorithm applied in the application;

[0071] Figure 4 The low difference sequence initialization strategy applied in the application;

[0072] Figure 5 The initialization dynamic coding based on machine tool service MTRj applied in the application. DETAILED DESCRIPTION

[0073] The technical solutions in the embodiments of the application will be described clearly and completely in the following with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only some embodiments of the application, not all embodiments.

[0074] This invention addresses the problem of insufficient utilization of idle time in machine tool equipment from the perspective of resource suppliers. First, it constructs a multi-objective optimization configuration model for machine tool equipment cloud services based on idle time windows. This model comprehensively considers both service benefits and service risks of machine tool equipment cloud services. Second, it proposes an improved Goose Escape Algorithm-ESC (GOOSE-ESC) to solve the model. This algorithm uses low-discrepancy sequences to generate a uniformly distributed initial population, designs a dynamic encoding mechanism based on machine tool equipment cloud services to adapt to dynamically changing idle time windows, and integrates the global search capability of the GOOSE algorithm with the fast convergence characteristics of the ESC algorithm, introducing an elite back-learning strategy to escape local optima. This invention can effectively utilize the minute-level fragmented time of machine tool equipment, improving enterprise revenue and resource utilization.

[0075] Specifically as follows:

[0076] See Figure 1 A method for dynamically optimizing the configuration of machine tool equipment cloud services with idle time windows includes at least the following steps:

[0077] S1: Processing orders from cloud manufacturing demanders. Machine tool services from cloud manufacturing resource providers The information is input into the service demand task pool and machine tool resource pool of the cloud manufacturing platform respectively. The focus of machine tool equipment resource suppliers in choosing to cooperate with cloud services is analyzed, and the cloud manufacturing service resources are matched according to the idle time window into two stages: combination and optimization.

[0078] S2: Combination Phase: Utilizing the idle time window of the machine tool equipment Matching candidate cloud manufacturing subtasks The tasks are sorted and combined to form a set of multi-subtask combination schemes for all machine tool services. ;

[0079] S3: Optimization Phase: Based on the availability benefits of machine tool equipment in providing cloud services and the risks associated with executing cloud manufacturing tasks, a multi-objective optimization configuration model is constructed, where availability benefits are equivalent to service revenue. The risks constituted are service risks. From all possible combinations of subtask sets The GOOSE-ESC algorithm is used to solve the multi-objective optimization configuration model in order to maximize service benefits and minimize service risks, and select the optimal configuration scheme with the greatest global benefits.

[0080] S4: Receive processing orders and machine tool services from cloud manufacturing resource providers. Optimized configuration scheme;

[0081] S5: execute the optimized configuration scheme in the machine tool equipment of the processing workshop to make full use of the idle time window of the machine tool equipment , improve resource utilization and enterprise revenue.

[0082] Referring to Figure 2 , S2 at least includes the following steps:

[0083] S201: obtain the processing amount , single-piece processing time , earliest start time , and latest end time of each subtask , and update the idle time window of all machine tool services ;

[0084] S202: find the feasible solution domain of all machine tool services in the idle time window ;

[0085] S203: for each machine tool service , randomly select a subtask from the feasible solution domain each time to form a combined scheme, and update the idle time window , for example, select a subtask from , form a combined scheme of , calculate the planned start time and planned completion time of the subtask , and replace the start time of the idle time window of with the planned completion time ;

[0086] S204: sequentially match the various combinations of subtasks and machine tool services , and classify and combine the subtasks into each service to form a possible task scheduling scheme, i.e., repeat steps S201 to S203 until the feasible solution domain of the machine tool service is empty, i.e., for the machine tool service , ;

[0087] S205: after sequentially selecting subtasks, obtain the idle time window of each machine tool service ​The set of combined subtasks of candidate solutions matching the idle time window of each machine service .

[0088] The derivation process of S205 includes at least the following steps:

[0089] Machine service All matching cloud manufacturing subtasks should be in the idle time window In addition, when the subtask Matches the machine service , it belongs to the feasible solution domain of the machine service In the idle time window , denoted as: ( );

[0090] When the latest start time And the earliest end time Of the subtask Should satisfy And , the calculation formula is:

[0091]

[0092]

[0093]

[0094]

[0095] Where, Indicates the end time of the idle time window of the machine service; The time requirement of the subtask ; Is a coefficient; Is the planned start time; Is the planned completion time; And The start time and end time of the idle time window of the machine service ;

[0096] Finally, the set of combined subtasks of candidate solutions matching each machine service idle time window . .

[0097] The multi-objective optimization configuration model in S3 is expressed as:

[0098]

[0099] Where:

[0100]

[0101]

[0102] wherein the expression , is the objective function.

[0103] Referring to Figure 3 , the application of the GOOSE-ESC algorithm in S3 includes at least the following steps:

[0104] S301: input all candidate cloud manufacturing tasks sub-tasks , machine tool service information and combine the sub-task set ;

[0105] S302: initialize the parameters of the GOOSE-ESC algorithm according to the state of the enterprise workshop machine tool service , the parameters of the GOOSE-ESC algorithm including population size , total number of iterations , number of machine tools , number of tasks , current number of iterations ;

[0106] S303: initialize the population with low difference sequence and initialize the dynamic coding of the initialized population based on machine tool service ;

[0107] S304: calculate the fitness value of the initialized population;

[0108] S305: and sort the goose according to the dominance level and crowding distance;

[0109] S306: select the development stage and the exploration stage for iteration by judging the size of the random probability value, when is less than 0.9, enter S307, and when it is greater than 0.9, enter S308;

[0110] S307: in the development stage, by judging the weight of the randomly generated stone , different iteration methods are selected in different geese;

[0111] S308: in the exploration stage, by recalculating the time required for the stone to fall , further find the position of the best guard goose;

[0112] S309: Introduce the attack strategy, further explore the feasible solution in the space by randomly simulating the escape behavior of the goose group after the alert goose wakes up the goose group through the random behavior coefficient, and the escape behavior divides the goose group into a follow-the-crowd group and a panic group;

[0113] S310: The elite individual after iteration is searched in reverse using the elite reverse learning strategy, and the elite individual is the goose with the top 10% escape speed;

[0114] S311: The goose group is non-dominantly sorted according to the advantage level and crowded distance of the individual, and the best solution in the non-dominant sorting is output;

[0115] S312: Set , repeat steps S306 to S311 until the termination condition is met, at which time the obtained Pareto front is the candidate solution set of the machine tool equipment cloud service multi-objective configuration model based on the idle time window;

[0116] S313: Output the optimal candidate solution set;

[0117] S314: End.

[0118] Referring to Figure 4 , the GOOSE-ESC algorithm adopts a low-dispersion sequence initialization strategy, generates initial population using Halton sequence, and maps the initial points from the unit hypercube to the actual search space to improve the spatial distribution uniformity of the initial solution.

[0119] The low-dispersion sequence initialization strategy at least includes the following steps:

[0120] First, Halton sequence is used to generate initial points with spatial uniformity;

[0121] Then, the points are mapped from the unit hypercube to the actual search space;

[0122] Finally, the mapped lattice points are used as the initial population to provide a set of uniformly distributed initial solutions for the optimization algorithm.

[0123] Referring to Figure 5 , the GOOSE-ESC algorithm adopts a dynamic encoding mechanism based on machine tool equipment cloud service, and the setting of the dynamic encoding mechanism at least includes the following steps:

[0124] First, the population individuals record the task number, processing procedure and machine tool equipment information through an N-dimensional matrix;

[0125] Then, in the decoding process, it is judged whether the remaining idle time of a machine tool equipment is sufficient to complete the assigned task;

[0126] Finally, when there is insufficient idle time, a dynamic reorganization mechanism is triggered to reassign unfinished manufacturing tasks to other machine tools with sufficient spare time.

[0127] Specific applications include:

[0128] like Figure 5 As shown, the population of individuals consists of machine tool equipment occupancy information composed of manufacturing tasks and task order. The figure shows that there are 3 machines, completing a total of 3 tasks and 10 sub-tasks. Among them: (1) Each row displays the processing information and processing order of the machine tool equipment. (2) In the manufacturing task In the middle, each column displays the task allocation of each subtask, using 0 and 1 to indicate whether the device processes the subtask in this column. Where 0 indicates no processing and 1 indicates processing; (3) in the processing sequence In the process, the processing tasks selected by the machine tool are processed in order of increasing code value; (4) During the processing, if the idle time of a certain machine tool is insufficient to complete the manufacturing task assigned by the cloud service platform, the sub-tasks of this equipment will be recombined and the manufacturing tasks that cannot be completed will be assigned to other machine tools with extra idle time.

[0129] For example, machine tool equipment 3 The processing task is , , The processing order corresponding to these three sub-tasks The priorities are 4, 9, and 10 respectively, therefore the priority order of the manufacturing tasks is { , , }. But at this time The available free time is insufficient to complete the task. Therefore, it is necessary to... The subtasks are reorganized, and other machine tools with available idle time are sought. Complete the task .

[0130] The GOOSE-ESC algorithm employs a two-stage collaborative optimization and update strategy:

[0131] The first stage is the GOOSE development and exploration stage: based on random probability, if the conditions are met, a search based on the flock's vigilance behavior is executed to calculate the individual's movement distance and direction;

[0132] The second stage is the ESC escape stage: if the conditions of the first stage are not met, an attack mechanism is introduced to simulate crowd evacuation behavior, and the positions of the panic group and the follower group are updated to increase population diversity.

[0133] The GOOSE-ESC algorithm includes an elite reverse learning strategy;

[0134] After each iteration is completed, an elite individual, i.e., the best guard goose, is selected, and a reverse solution thereof is generated;

[0135] If the fitness of the reverse solution is better than that of the original solution, the original elite solution is replaced by the reverse solution, which is used to jump out of a local optimum.

[0136] It will be apparent to those skilled in the art that the application is not limited to the details of the foregoing exemplary embodiments, and that the application can be implemented in other particular forms without departing from the spirit or essential characteristics of the application. The foregoing embodiments are therefore to be considered in all respects as illustrative only, and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. No reference signs in the claims shall be construed as limiting the scope of the claims.

Claims

1. A method for dynamically optimizing the configuration of machine tool equipment cloud services with idle time windows, characterized in that: At least the following steps are included: S1: Processing orders from cloud manufacturing demanders. Machine tool services from cloud manufacturing resource providers The information is input into the service demand task pool and machine tool resource pool of the cloud manufacturing platform respectively. The focus of machine tool equipment resource suppliers in choosing to cooperate with cloud services is analyzed, and the cloud manufacturing service resources are matched according to the idle time window into two stages: combination and optimization. S2: Combination Phase: Utilizing the idle time window of the machine tool equipment Matching candidate cloud manufacturing subtasks The tasks are sorted and combined to form a set of multi-subtask combination schemes for all machine tool services. ; S3: Optimization Phase: Based on the availability benefits of machine tool equipment in providing cloud services and the risks associated with executing cloud manufacturing tasks, a multi-objective optimization configuration model is constructed, where availability benefits refer to service revenue. The risks constituted are service risks. For multi-subtask combination schemes The GOOSE-ESC algorithm is used to solve the multi-objective optimization configuration model in order to maximize service benefits and minimize service risks, and select the optimal configuration scheme with the greatest global benefits. S4: Obtain the processing order task and the machine tool service from the cloud manufacturing resource provider. Optimized configuration scheme; S5: Implement the optimized configuration scheme in the machine tool equipment in the machining workshop to make full use of the idle time window of the machine tool equipment. This will improve resource utilization and corporate profits.

2. The method for dynamic optimization configuration of machine tool equipment cloud services with idle time windows according to claim 1, characterized in that: S2 includes at least the following steps: S201: Obtain all candidate cloud manufacturing tasks Next task Processing volume Single piece processing time Earliest start time and latest end time and update all machine tool services. idle time window ; S202: Find all machine tool services During idle time windows Feasible solution domain ; S203: For each machine tool service Each time from the feasible solution domain Randomly select a subtask Formulate a combination plan and update the idle time window. ; S204: Pair Task machine tool services Multiple combinations are matched sequentially to identify subtasks. The tasks are categorized and combined into each service to form a task scheduling scheme, i.e., steps S201 to S203 are repeated until the feasibility domain of the machine tool service is completely empty, i.e., for the machine tool service... , ; S205: After selecting the sub-tasks in sequence, obtain the service details for each machine tool. Idle Time Window A set of combined subtasks for matching candidate solutions .

3. The method for dynamic optimization configuration of machine tool equipment cloud services with idle time windows according to claim 2, characterized in that: The derivation process of S205 includes at least the following steps: Machine tool service All matching cloud manufacturing subtasks should be performed within the idle time window. In addition, when subtasks machine tool services If a match is found, it belongs to the machine tool service. In the idle time window The feasible domain of solutions is denoted as: ( ); When subtask Latest start time and earliest end time Should meet and The calculation formula is: in, Indicates the end time of the idle time window for machine tool services; For subtasks Time requirements; It is a coefficient; The planned start time; The planned completion time; and These are machine tool services The start and end times of the idle time window; Ultimately, the idle time window for each machine tool service is obtained. The combined subtask set of matching candidate solutions .

4. The method for dynamic optimization configuration of machine tool equipment cloud services with idle time windows according to claim 1, characterized in that: The multi-objective optimization configuration model in S3 is described as follows: in: Where, expression , The objective function is denoted as .

5. The method for dynamic optimization configuration of machine tool equipment cloud services with idle time windows according to claim 4, characterized in that: The application of the GOOSE-ESC algorithm in S3 includes at least the following steps: S301: Input all candidate cloud manufacturing tasks Next task The machine tool service Information and the combined subtask set ; S302: Machine tool service as described in the enterprise workshop The parameters of the GOOSE-ESC algorithm are initialized by state initialization, and the parameters of the GOOSE-ESC algorithm include population size. Total number of iterations Number of machine tools Number of tasks Current iteration number ; S303: Initialize the population using the low-difference sequence and based on the machine tool service. Perform initial dynamic encoding on the initialized population; S304: Calculate the fitness value of the population after initialization; S305: And sort the flock of geese into non-dominant groups based on dominance level and crowding distance; S306: The selection of whether to enter the development or exploration phase is iterated by judging the magnitude of a random probability value. If the value is less than 0.9, proceed to S307; if the value is greater than 0.9, proceed to S308. S307: During the development phase, by adjusting the weight of randomly generated stones... Make a judgment and select different iteration methods for different individual geese; S308: During the exploration phase, the time required for the stone to fall is recalculated. To further search for the best location for the watchful geese; S309: Introduce an attack strategy, and use random behavioral coefficients to randomly simulate the escape behavior of the geese after the alert geese are awakened to further explore feasible solutions in the space. The escape behavior divides the geese into a follower group and a panic group. S310: Use an elite reverse learning strategy to perform a reverse search on the elite individuals after the iteration is completed. The elite individuals are the geese with the top 10% escape speed. S311: Perform a non-dominated sorting of the flock of geese based on the dominance level and crowding distance of the individuals, and output the best solution in the non-dominated sorting. S312: Let Repeat steps S306 to S311 until the termination condition is met. The Pareto front obtained at this time is the candidate solution set of the machine tool equipment cloud service multi-objective optimization configuration model based on idle time window. S313: Output the optimal candidate solution set; S314: End.

6. The method for dynamic optimization configuration of machine tool equipment cloud services with idle time windows according to claim 5, characterized in that: The GOOSE-ESC algorithm employs a low-difference sequence initialization strategy, using Halton sequences to generate an initial population and mapping the initial point from the unit hypercube to the actual search space, thereby improving the spatial uniformity of the initial solution. The low-difference sequence initialization strategy includes at least the following steps: First, Halton sequences are used to generate initial points with spatial uniformity; These points are then mapped from the unit hypercube to the actual search space; Finally, the mapped grid points are used as the initial population to provide a uniformly distributed initial solution for the optimization algorithm.

7. The method for dynamic optimization configuration of machine tool equipment cloud services with idle time windows according to claim 6, characterized in that: The GOOSE-ESC algorithm adopts a dynamic coding mechanism based on machine tool equipment cloud services. The setting of the dynamic coding mechanism includes at least the following steps: First, individuals in the population record task numbers, processing procedures, and machine tool equipment information using an N-dimensional matrix; Then, during the decoding process, it is determined whether the remaining idle time of a certain machine tool is sufficient to complete the assigned task; Finally, when there is insufficient idle time, a dynamic reorganization mechanism is triggered to reassign unfinished manufacturing tasks to other machine tools with sufficient spare time.

8. The method for dynamic optimization configuration of machine tool equipment cloud services with idle time windows according to claim 7, characterized in that: The GOOSE-ESC algorithm employs a two-stage collaborative optimization and update strategy: The first stage is the GOOSE development and exploration stage: based on random probability, if the conditions are met, a search based on the flock's vigilance behavior is executed to calculate the individual's movement distance and direction; The second stage is the ESC escape stage: if the conditions of the first stage are not met, an attack mechanism is introduced to simulate crowd evacuation behavior, and the positions of the panic group and the follower group are updated to increase population diversity.

9. A method for dynamic optimization configuration of machine tool equipment cloud services with idle time windows according to claim 8, characterized in that: The GOOSE-ESC algorithm includes an elite reverse learning strategy; After each iteration, the elite individual, i.e. the best vigilant goose, is selected, and its reverse solution is generated; If the fitness of the reverse solution is better than that of the original solution, then the original elite solution is replaced to escape local optima.

Images