A genetic algorithm scheduling task constraint method for order scheduling

Abstract

The present invention relates to the technical field of genetic algorithm scheduling, and discloses a genetic algorithm scheduling task constraint method for order scheduling, comprising setting a multi-task model and defining a resource model, establishing a first screening function, unifying a set of constraint conditions, determining a constraint condition evaluation function, evaluating the set of constraint conditions, setting a second screening function, inputting the output value of the constraint condition evaluation function into the second screening function, completing task locking of genetic algorithm scheduling, optimizing the constraint processing flow through an optimization function, generating a scheduling calculation monitoring visualization diagram, particularly considering the problem of task priority, and ensuring that locked tasks are given priority through modifying fitness functions and gene operations.

Description

Technical Field

[0001] The invention relates to the technical field of genetic algorithm scheduling and discloses a genetic algorithm scheduling task constraint method for order scheduling. Background Art

[0002] Although genetic algorithms have significant advantages and applications in the development of task scheduling, they also have some shortcomings.

[0003] For large-scale task scheduling problems, genetic algorithms converge slowly and require a lot of computing resources and time. As the number of tasks and constraints increase, the efficiency of the algorithm will further decrease, which limits its application in real-time or high-frequency task scheduling. In actual task scheduling, there are often complex constraints between tasks, such as time windows, dependencies, resource allocation, etc. When dealing with multiple constraints, genetic algorithms may require additional complexity evaluation and processing mechanisms, otherwise it may not be possible to effectively meet all constraints. In actual task scheduling, there are often complex constraints between tasks, such as time windows, dependencies, resource allocation, etc. When dealing with multiple constraints, genetic algorithms may require additional complexity evaluation and processing mechanisms, otherwise it may not be possible to effectively meet all constraints. Now there is an urgent need for a scheduling task constraint method to overcome the above shortcomings.

[0004] For example, the Chinese patent with the existing publication number CN113139710A provides a high-level planning and scheduling method for multi-resource parallel tasks based on genetic algorithms. This method is based on the dynamic determination of process paths and the dynamic generation of manufacturing capabilities, with genetic algorithms as the core. This method uses multi-resource combination constraints, adjacency matrix construction and analysis, and efficient iteration of genetic algorithms to achieve functions such as flexible matching of enterprise resource needs, automatic combing of process paths, and automatic scheduling of multi-order tasks. This invention realizes a method of accurately outputting the process paths of different products into the operation rules with a universal operation logic. When the product process path changes, the corresponding process execution order can be output without changing the program code, thereby improving the universality and efficiency of the scheduling results.

[0005] The above invention is not flexible enough in handling hard constraints and soft constraints. Genetic algorithms are usually easier to handle hard constraints, but are not flexible and accurate enough in handling soft constraints. Secondly, in actual scheduling problems, a large number of tasks and resources are usually involved, which leads to very complex and diverse constraints. When dealing with high-dimensional and complex constraints, genetic algorithms are prone to fall into local optimal solutions or problems of excessive search space, resulting in inefficiency or inability to find suitable solutions. The dynamic constraints are not adaptable enough. In actual scheduling, the constraints may be dynamically adjusted with time or changes in the environment. Summary of the invention

[0006] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0007] In order to solve the above technical problems, the main purpose of the present invention is to provide a genetic algorithm scheduling task constraint method for order scheduling, comprising:

[0008] S1, setting a multi-task model and defining a resource model, and establishing a first screening function;

[0009] S2. Unify the set of constraint conditions, determine the constraint condition evaluation function, and evaluate the set of constraint conditions;

[0010] S3, setting a second screening function, and inputting the output value of the constraint condition evaluation function into the second screening function;

[0011] S4. If the constraint evaluation value is greater than the first priority threshold, the task priority is high; if the constraint evaluation value is less than or equal to the first priority threshold and greater than or equal to the second priority threshold, the task priority is medium; if the constraint evaluation value is less than the second priority threshold, the task priority is low;

[0012] S5, completing the task locking of genetic algorithm scheduling, and optimizing the constraint processing flow through the optimization function;

[0013] S6. Generate a visual graph of scheduling calculation monitoring;

[0014] S7. Sort and monitor order scheduling through scheduling calculation and monitoring visual graphs.

[0015] As a preferred solution of a genetic algorithm scheduling task constraint method for order scheduling of the present invention, wherein:

[0016] The multi-task model is used to parse out multiple objectives in the order and allocate resources to the multiple objectives of the task through a multi-objective fitness function;

[0017] The resource model is used to schedule the multi-objective allocation of the task;

[0018] The first screening function is used to dynamically screen the multi-objective schedule of the task, and the schedule is sorted and screened by the first screening function;

[0019] If the multi-objective allocated resources of the task are greater than the maximum resources carried by the resource model, they are screened out through the first screening function. If the multi-objective allocated resources of the task are not greater than the maximum resources carried by the resource model, the resource occupancy is sorted through the first screening function, and the multi-objectives of the sorted tasks are scheduled.

[0020] As a preferred solution of a genetic algorithm scheduling task constraint method for order scheduling of the present invention, wherein:

[0021] The multi-task model analyzes the tasks by establishing a multi-objective set. The multi-objective set calculation expression is as follows:

[0022]

[0023] in, is the task data, is the first target data, is the second target data, For the third target data, is the nth target data, n is any positive integer;

[0024] The first screening function optimizes the multi-objective scheduling of tasks whose multi-objective allocated resources are not greater than the maximum amount of resources carried by the resource model. The calculation expression of the first screening function is as follows:

[0025]

[0026]

[0027] in, For the task with the minimum resource allocation in multi-objective data, Allocate resources for the first target data, Allocate resources for the second target data, Allocate resources for the nth target data, is the amount of resources corresponding to the first target data, is the minimum scheduling quantity, is the maximum scheduling amount, is the linear equality constraint for the ith target schedule, is the nonlinear constraint for the ith target schedule, is the linear constraint constant, is the linear constraint result quantity, A is the linear inequality constant, and b is the linear inequality result quantity.

[0028] As a preferred solution of a genetic algorithm scheduling task constraint method for order scheduling of the present invention, wherein:

[0029] The first screening function optimizes the multi-objective scheduling of tasks whose multi-objective allocated resources are not greater than the maximum amount of resources carried by the resource model. If the multi-objective scheduling of the task is output as a linear equation through the constraint condition of the first screening function, the constraint condition of the multi-objective scheduling of the task is evaluated and the multi-objective scheduling of the task is output. If the multi-objective scheduling of the task is output as nonlinear through the constraint condition of the first screening function, the multi-objective scheduling of the nonlinear task is filtered;

[0030] The constraint evaluation function is used to evaluate the constraints in the first screening function. If the multi-objective scheduling screening of the task is nonlinear, the constraints are dynamically optimized. If the multi-objective scheduling screening of the task is normal, the constraints are cached and scheduled cleared.

[0031] As a preferred solution of a genetic algorithm scheduling task constraint method for order scheduling of the present invention, wherein:

[0032] The constraint conditions are evaluated by an evaluation function. The evaluation function evaluates the constraint conditions by evaluating the scheduling level through the multi-objective scheduling gain of the fixed task. The calculation expression of the evaluation function is as follows:

[0033]

[0034] in, is the evaluation value of the constraint condition, is the multi-objective scheduling level of the i-th task, is the average level of multi-objective scheduling of the i-th task, is the actual scheduling level of the multi-objective scheduling mapping of the i-th task, and max is the maximum value function;

[0035] Since the actual scheduling level of the multi-objective scheduling mapping of the i-th task will be affected by the scheduled tasks and cache scheduling, the iteration of the difference between the average level of the multi-objective scheduling of the i-th task and the actual scheduling level of the multi-objective scheduling mapping of the i-th task will also differ due to different tasks and the different actual scheduling levels of the multi-objective scheduling mapping of the i-th task.

[0036] As a preferred solution of a genetic algorithm scheduling task constraint method for order scheduling of the present invention, wherein:

[0037] The second screening function is used to further filter candidate solutions that meet the initial constraint conditions;

[0038] The evaluation value of the evaluation constraint condition is input into the second screening function, and the constraint condition is further screened by the second screening function.

[0039] As a preferred solution of a genetic algorithm scheduling task constraint method for order scheduling of the present invention, wherein:

[0040] The optimization function includes hard optimization function and soft optimization function;

[0041] The hard optimization function is used to optimize the multi-objective scheduling of tasks that must meet the constraints;

[0042] The soft optimization function is used to optimize the multi-objective scheduling of tasks that do not necessarily meet the constraint conditions;

[0043] The hard optimization function includes decoding and regeneration;

[0044] The soft optimization function includes a constraint handling function.

[0045] As a preferred solution of a genetic algorithm scheduling task constraint method for order scheduling of the present invention, wherein:

[0046] S401, if the task priority is high, modify the population generation function, fitness function, crossover point selection function and mutation point selection function in sequence;

[0047] S402, modifying the population generation function to include the locked task in the initial population, and determining the population generation order by the task priority when generating the population, and locking the task in the process of generating the population;

[0048] S403, the fitness function is used to evaluate the pros and cons of each individual, and a higher fitness score is assigned to the locked task by modifying the fitness function;

[0049] S404, modifying the intersection selection function to avoid selecting a position of a locked task when selecting an intersection;

[0050] S405. Modify the mutation point selection function to avoid selecting the position of the locked task when selecting the mutation point, so as to ensure that the position of the locked task will not be accidentally changed.

[0051] As a preferred solution of a genetic algorithm scheduling task constraint method for order scheduling of the present invention, wherein:

[0052] The modified population generation function prioritizes the multi-objective scheduling of tasks, and generates the size of the population based on the multi-objective scheduling of tasks, and the multi-objective scheduling of the locked tasks obtained by the task locking is included in the population;

[0053] The fitness function is used to sort the multi-objective scheduling of tasks and calculate the task locking fitness scores of multiple tasks;

[0054] The modified intersection selection function is used to avoid the position of the locked task when selecting the intersection;

[0055] The modification of the mutation point selection function includes random selection of the location of the non-locked task and weight-based selection;

[0056] The randomly selecting the position of the non-locked task includes randomly selecting a position from all possible positions of the chromosome when selecting the mutation point to determine the position of the locked task;

[0057] The weight-based selection is that each position is assigned a weight and the variation points are selected according to these weights.

[0058] As a preferred solution of a genetic algorithm scheduling task constraint method for order scheduling of the present invention, wherein:

[0059] The generated scheduling calculation monitoring visualization diagram is used to monitor information and visualize information during the multi-objective analysis and scheduling process of the task;

[0060] The monitoring information includes task status and resource utilization;

[0061] Visualize information through visualization tools for task scheduling and monitoring;

[0062] Display monitoring information and visualization information through a user-friendly monitoring interface.

[0063] Beneficial effects of the present invention:

[0064] This method embeds the task constraint method in the genetic algorithm for scheduling. Compared with traditional scheduling methods or existing technologies, this method has higher flexibility and customizability, and takes into account the issue of task priority. By modifying the fitness function and genetic operations, it ensures that the locked tasks are given priority. This is different from traditional scheduling methods or existing technologies, and can better meet the requirements of task priority in business needs. It can not only obtain the production schedule of production orders and the production path of production orders in different processes, but also calculate the optimal start time and end time of each process of each order. After completing all the production scheduling calculations, it will generate production schedules, production Gantt charts, fitness curves and other visual charts to prove the advancement and effectiveness of the production schedule, which can greatly increase customer trust. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0066] Figure 1 A flow chart of a genetic algorithm scheduling task constraint method for order scheduling according to the present invention;

[0067] Figure 2 This is a processing flow chart for the constraint condition with a high evaluation value in the present invention. DETAILED DESCRIPTION

[0068] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0069] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0070] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0071] Embodiment 1

[0072] like Figure 1 As shown, a genetic algorithm scheduling task constraint method for order scheduling includes:

[0073] S1. Setting a multi-task model and defining a resource model, and establishing a first screening function (for determining the priority of tasks);

[0074] Among them, the order contains multiple tasks;

[0075] The multi-task model is used to optimize multiple objectives of a task simultaneously, and allocate resources to multiple objectives of the task through a multi-objective fitness function;

[0076] Multi-objective allocation of resources for scheduling tasks through resource models;

[0077] Furthermore, the multi-task model includes:

[0078] Task ID: unique identifier; Task name: descriptive name; Task duration: how long it takes to complete the task; Earliest start time: the earliest time a task can start; Latest end time: the latest time a task must be completed; Priority: how important or urgent a task is; Predecessors: other tasks that must be completed before this task; Resource requirements: the type and amount of resources required to complete the task; Task type: a classification of tasks that may affect scheduling decisions; Flexibility: the degree to which a task can be adjusted in time;

[0079] The resource model includes: Resource ID: unique identifier; Resource name: descriptive name; Resource type: classification of resources (such as manpower; equipment; materials, etc.); Capacity: the available quantity of resources or the degree of simultaneous availability; Available time window: the available time period of resources; Skills / abilities: the types of tasks that resources can perform; Efficiency factor: the efficiency of resources in performing tasks; Cost: the cost of using resources; Substitutability: other resources that can replace this resource; Maintenance time: the time period when resources need regular maintenance;

[0080] The first screening function is used to dynamically screen the multi-objective schedule of the task, and the schedule is sorted and screened by the first screening function;

[0081] If the multi-objective allocated resources of the task are greater than the maximum resources carried by the resource model, they are screened out through the first screening function. If the multi-objective allocated resources of the task are not greater than the maximum resources carried by the resource model, the resource occupancy is sorted through the first screening function, and the multi-objectives of the sorted tasks are scheduled.

[0082] The multi-task model analyzes the tasks by establishing a multi-objective set. The calculation expression of the multi-objective set is as follows:

[0083]

[0084] in, is the task data, is the first target data, is the second target data, For the third target data, is the nth target data, n is any positive integer;

[0085] The resource model is used to schedule the resources for multi-objective allocation of tasks;

[0086] The first screening function optimizes the multi-objective scheduling of tasks whose multi-objective allocated resources are not greater than the maximum amount of resources carried by the resource model. The calculation expression of the first screening function is as follows:

[0087]

[0088]

[0089] in, For the task with the minimum resource allocation in multi-objective data, Allocate resources for the first target data, Allocate resources for the second target data, Allocate resources for the nth target data, is the amount of resources corresponding to the first target data, is the minimum scheduling quantity, is the maximum scheduling amount, is the linear equality constraint for the ith target schedule, is the nonlinear constraint for the ith target schedule, is the linear constraint constant, is the linear constraint result quantity, A is the linear inequality constant, and b is the linear inequality result quantity;

[0090] The amount of resources allocated to the first target data is a set with labels, The resource amount corresponding to the first target data is quantity data;

[0091] When the linear programming constraints have When the condition is met, use Aeq and beq to ensure that the equality sign holds, that is, Then the Aeq position is a constant, the others are 0, and beq has several values. Just take a few ;

[0092] The first screening function optimizes the multi-objective scheduling of tasks whose multi-objective allocation resources are not greater than the maximum amount of resources carried by the resource model. If the multi-objective scheduling of the task is output as a linear equation through the constraint condition of the first screening function, the constraint condition of the multi-objective scheduling of the task is evaluated and the multi-objective scheduling of the task is output. If the multi-objective scheduling of the task is output as nonlinear through the constraint condition of the first screening function, the multi-objective scheduling of the nonlinear task is filtered;

[0093] If the multi-objective scheduling of the task is linear, it means that the multi-objective scheduling processing of the task will not have garbled characters or task resource allocation conflicts and other faults. If the multi-objective scheduling of the task is non-linear, it means that the multi-objective scheduling resources or task processing of the task will fail and conflict with the constraints. Therefore, the multi-objective scheduling of the task is filtered out. If the multi-objective scheduling of the task is linear, it is screened through the constraint set, and the constraints are evaluated and dynamically updated in real time. If the multi-objective scheduling of the task is non-linear, it is re-analyzed, resource allocated and screened through the feedback task, and re-screened through the dynamically updated constraints.

[0094] S2. Unify the set of constraint conditions, determine the constraint condition evaluation function, and evaluate the set of constraint conditions;

[0095] The constraint evaluation function is used to evaluate the constraint in the first screening function. If the multi-objective scheduling screening of the task is nonlinear, the constraint is dynamically optimized. If the multi-objective scheduling screening of the task is normal, the constraint is cached and scheduled cleared.

[0096] The constraint conditions are evaluated by an evaluation function. The evaluation function evaluates the constraint conditions by evaluating the scheduling level through the multi-objective scheduling gain of the fixed task. The calculation expression of the evaluation function is as follows:

[0097]

[0098] in, is the evaluation value of the constraint condition, is the multi-objective scheduling level of the i-th task, is the average level of multi-objective scheduling of the i-th task, is the actual scheduling level of the multi-objective scheduling mapping of the i-th task, and max is the maximum value function;

[0099] Since the actual scheduling level of the multi-objective scheduling mapping of the i-th task will be affected by the scheduled task and the cache schedule, the iteration of the difference between the average level of the multi-objective scheduling of the i-th task and the actual scheduling level of the multi-objective scheduling mapping of the i-th task will also be different due to the difference of the tasks and the actual scheduling level of the multi-objective scheduling mapping of the i-th task;

[0100] The actual scheduling level of the multi-objective scheduling mapping of the i-th task is combined with factors such as the task priority, duration, and resource requirements to reflect the multi-objective scheduling level of the task. The multi-objective scheduling of the task is mapped out by the mapping function:

[0101] Furthermore, the constraints are dynamically optimized by adjusting the average level of multi-objective scheduling of tasks and the actual level of multi-objective scheduling of tasks mapped by the mapping function;

[0102] Furthermore, the mapping function searches for the actual level of the multi-objective scheduling of the task through the multi-objective scheduling level table of the task preset by the system, and finally performs mapping;

[0103] S3, setting a second screening function, and inputting the output value of the constraint condition evaluation function into the second screening function;

[0104] The second screening function is used to further filter candidate solutions that meet the initial constraints to ensure that the final selected solution meets more stringent or specific requirements. This function can be regarded as a more sophisticated filter after the initial constraint evaluation to further improve the quality of the solution or meet specific requirements;

[0105] The evaluation value of the evaluation constraint is input into the second screening function, and the evaluation value of the evaluation constraint may be a certain rating, score or satisfaction metric, which reflects the performance of the solution under the initial constraint;

[0106] The output is determined by the design of the second screening function, which may be a Boolean value (pass / fail), a weighted score, a new constraint score, etc.

[0107] S4. If the constraint evaluation value is high, the task priority is high; if the constraint evaluation value is less than or equal to the first priority threshold and greater than or equal to the second priority threshold, the task priority is medium; if the constraint evaluation value is low and less than the second priority threshold, the task priority is low;

[0108] S5, completing the task locking of genetic algorithm scheduling, and optimizing the constraint processing flow through the optimization function;

[0109] The optimization function includes hard optimization function and soft optimization function;

[0110] Hard optimization functions are used to optimize multi-objective scheduling of tasks that must satisfy constraints;

[0111] The soft optimization function is used to optimize the multi-objective scheduling of tasks that do not necessarily meet the constraints.

[0112] Hard optimization functions include decoding and regeneration;

[0113] The soft optimization function includes constraint handling functions;

[0114] Furthermore, a special constraint processing function is designed to ensure that the generated solution does not violate the constraint conditions during the operation of the genetic algorithm.

[0115] S6. Generate a visual graph of scheduling calculation monitoring;

[0116] Generate a scheduling calculation monitoring visual graph including the multi-objective analysis of tasks and the order scheduling process, and monitor the multi-objective analysis information and order scheduling visualization information through the scheduling calculation monitoring visual graph;

[0117] Monitoring information includes task status and resource utilization;

[0118] Visualize information through visualization tools for task scheduling and monitoring;

[0119] Display monitoring information and visualize information through a user-friendly monitoring interface;

[0120] S7. Sort and monitor order scheduling through scheduling calculation and monitoring visual graphs.

[0121] Embodiment 2

[0122] like Figure 2 As shown, the processing flow for the constraint condition with high evaluation value;

[0123] S4. If the constraint evaluation value is greater than the first priority threshold, the task priority is high; if the constraint evaluation value is less than or equal to the first priority threshold and greater than or equal to the second priority threshold, the task priority is medium; if the constraint evaluation value is less than the second priority threshold, the task priority is low;

[0124] Among them, the constraint condition evaluation value threshold is set according to the task to classify the constraint condition evaluation value;

[0125] S401, if the task priority is high, modify the population generation function, fitness function, crossover point selection function and mutation point selection function;

[0126] S402, modifying the population generation function to ensure that the locked task is included in the initial population. When generating the population, the generation order of the population is determined by the task priority to ensure that the locked task is taken into consideration in the process of generating the population.

[0127] Modify the population generation function to prioritize the multi-objective scheduling of tasks, ensure that high-priority tasks are processed first, determine the size of the generated population and other constraints (such as the maximum number of tasks in each population), and generate the size of the population based on the multi-objective scheduling of tasks, and obtain the multi-objective scheduling of locked tasks in the population through the task locking;

[0128] For example, if certain tasks must appear in the initial population, you can explicitly add these tasks in the function that generates the population, making sure that the order in which the tasks are added matches their priority to ensure that high-priority tasks are processed first.

[0129] S403, the fitness function is used to evaluate the quality of each individual. By modifying the fitness function, a higher fitness score is assigned to the locked tasks to ensure that they are more likely to be selected and retained in the population;

[0130] The fitness function is used to sort the multi-objective scheduling of tasks and calculate the task locking fitness scores of multiple tasks, and modify the fitness function so that the locked tasks score higher when calculating the fitness;

[0131] Assume that the fitness function can be expressed as: [ \text{Fitness}(T_i) = w_1 \cdot \text{CompletionTime}(T_i)+ w_2 \cdot \text{ResourceUtilization}(T_i) + w_3 \cdot\text{Cost}(T_i)];

[0132] Among them, (T_i) represents the i-th task, (\text{CompletionTime}(T_i)) is the completion time, (\text{ResourceUtilization}(T_i)) is the resource utilization, (\text{Cost}(T_i)) is the cost, and (w_1,w_2,w_3) are the corresponding weights, indicating the importance of different goals;

[0133] S404, modifying the intersection selection function, by avoiding selecting the position of the locked task when selecting the intersection, to ensure that the position of the locked task will not be crossed. For example, when selecting the intersection, the position of the locked task can be excluded, or the position of the locked task can be fixed on both sides of the intersection.

[0134] When performing a crossover operation, two parent individuals are usually selected and the point where they cross over (the crossover point) is determined. Modify the crossover point selection function so that it avoids the position of the locked task when selecting the crossover point.

[0135] For example, if there is location information of a locked task on the chromosome, randomly select the location point of a non-locked task: When selecting a crossover point, randomly select a position from the chromosome, but make sure it is not the position of a locked task, assign a weight to each position on the chromosome, and the weight can reflect whether it is the position of a locked task. Then, when selecting a crossover point, randomly select a position based on the weight, so that the position of the locked task has a lower probability of selection.

[0136] S405. Modify the mutation point selection function to avoid selecting the position of the locked task when selecting the mutation point, thereby ensuring that the position of the locked task will not be accidentally changed.

[0137] Among them, modifying the mutation point selection function includes randomly selecting the location of non-locked tasks and weight-based selection;

[0138] Furthermore, randomly selecting a position of a non-locked task includes randomly selecting a position from all possible positions of the chromosome or solution when selecting a mutation point, but ensuring that it is not a position of a locked task.

[0139] Weight-based selection assigns a weight to each position, which can be adjusted based on whether the position is a task-locked position. Mutation points are then selected based on these weights, ensuring that task-locked positions have a low probability of selection.

[0140] Importantly, it should be noted that the construction and arrangement of the present application shown in a plurality of different exemplary embodiments are only exemplary. Although only two embodiments are described in detail in this disclosure, it should be readily understood by those who refer to this disclosure that many modifications are possible, for example, the size, scale, structure, shape and ratio of various elements, and parameter values ​​(e.g., temperature, pressure, etc.), installation arrangement, use of materials, color, directional changes, etc., without substantially departing from the novel teachings and advantages of the subject matter described in the application. For example, the element shown as integrally formed can be composed of multiple parts or elements, the position of the element can be inverted or otherwise changed, and the nature or number or position of the discrete element can be changed or changed. Therefore, all such modifications are intended to be included in the scope of the present invention. The order or sequence of any process or method steps can be changed or reordered according to alternative embodiments. Any "device plus function" clause is intended to cover the structure of the execution function described in this article, and is not only structurally equivalent but also equivalent structure. Without departing from the scope of the present invention, other replacements, modifications, changes and omissions can be made in the design, operating conditions and arrangement of the exemplary embodiment. Therefore, the invention is not limited to a specific embodiment, but extends to numerous modifications still falling within the scope of the appended claims.

[0141] Additionally, in order to provide a concise description of exemplary embodiments, all features of an actual embodiment may not be described (ie, those features that are not relevant to the best mode presently contemplated for carrying out the invention or those that are not relevant to implementing the invention).

[0142] It should be understood that in the development of any actual implementation, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort may be complex and time-consuming, but for those of ordinary skill having the benefit of this disclosure, the development effort will be a routine task of design, fabrication, and production without undue experimentation.

[0143] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A genetic algorithm scheduling task constraint method for order scheduling, characterized in that: include: S1, setting a multi-task model and defining a resource model, and establishing a first screening function; The multi-task model is used to parse out multiple objectives in the order and allocate resources to the multiple objectives of the task through a multi-objective fitness function; The resource model is used to schedule the multi-objective allocation of the task; The first screening function is used to dynamically screen the multi-objective schedule of tasks, and the schedule is sorted and screened by the first screening function; If the multi-objective allocation resources of the task are greater than the maximum resources carried by the resource model, the task is screened out through the first screening function. If the multi-objective allocation resources of the task are not greater than the maximum resources carried by the resource model, the resource occupation is sorted through the first screening function, and the multi-objectives of the sorted tasks are scheduled; The multi-task model analyzes the tasks by establishing a multi-objective set. The multi-objective set calculation expression is as follows: ； in, is the task data, is the first target data, is the second target data, For the third target data, is the nth target data, n is any positive integer; The first screening function optimizes the multi-objective scheduling of tasks whose multi-objective allocated resources are not greater than the maximum amount of resources carried by the resource model. The calculation expression of the first screening function is as follows: ； ； in, For the task with the minimum resource allocation in multi-objective data, Allocate resources for the first target data, Allocate resources for the second target data, Allocate resources for the nth target data, is the amount of resources corresponding to the first target data, is the minimum scheduling quantity, is the maximum scheduling amount, is the linear equality constraint for the ith target schedule, is the nonlinear constraint for the ith target schedule, is the linear constraint constant, is the linear constraint result quantity, A is the linear inequality constant, and b is the linear inequality result quantity; S2. Unify the set of constraint conditions, determine the constraint condition evaluation function, and evaluate the constraint condition set; The first screening function optimizes the multi-objective scheduling of tasks whose multi-objective allocated resources are not greater than the maximum amount of resources carried by the resource model. If the multi-objective scheduling of the task is output as a linear equation through the constraint condition of the first screening function, the constraint condition of the multi-objective scheduling of the task is evaluated and the multi-objective scheduling of the task is output. If the multi-objective scheduling of the task is output as nonlinear through the constraint condition of the first screening function, the multi-objective scheduling of the nonlinear task is filtered; The constraint evaluation function is used to evaluate the constraint in the first screening function. If the multi-objective scheduling screening of the task is nonlinear, the constraint is dynamically optimized. If the multi-objective scheduling screening of the task is normal, the constraint is cached and scheduled cleared. The constraint conditions are evaluated by an evaluation function. The evaluation function evaluates the constraint conditions by evaluating the scheduling level through the multi-objective scheduling gain of the fixed task. The calculation expression of the evaluation function is as follows: ； in, is the evaluation value of the constraint condition, is the multi-objective scheduling level of the i-th task, is the average level of multi-objective scheduling of the i-th task, is the actual scheduling level of the multi-objective scheduling mapping of the i-th task, and max is the maximum value function; S3, setting a second screening function, and inputting the output value of the constraint condition evaluation function into the second screening function; S4. If the constraint evaluation value is greater than the first priority threshold, the task priority is high; if the constraint evaluation value is less than or equal to the first priority threshold and greater than or equal to the second priority threshold, the task priority is medium; if the constraint evaluation value is less than the second priority threshold, the task priority is low; S5, completing the task locking of genetic algorithm scheduling, and optimizing the constraint processing flow through the optimization function; S6. Generate a visual graph of scheduling calculation monitoring; S7. Sort and monitor order scheduling through scheduling calculation and monitoring visual graphs.

2. The genetic algorithm scheduling task constraint method for order scheduling according to claim 1, characterized in that: The second screening function is used to further filter candidate solutions that meet the initial constraint conditions; The evaluation value of the evaluation constraint condition is input into the second screening function, and the constraint condition is further screened by the second screening function.

3. The genetic algorithm scheduling task constraint method for order scheduling according to claim 2, characterized in that: The optimization function includes hard optimization function and soft optimization function; The hard optimization function is used to optimize the multi-objective scheduling of tasks that must meet the constraints; The soft optimization function is used to optimize the multi-objective scheduling of tasks that do not necessarily meet the constraint conditions; The hard optimization function includes decoding and regeneration; The soft optimization function includes a constraint handling function.

4. The genetic algorithm scheduling task constraint method for order scheduling according to claim 3 is characterized by: S401, if the task priority is high, modify the population generation function, fitness function, crossover point selection function and mutation point selection function in sequence; S402, modifying the population generation function to include the locked task in the initial population, and determining the generation order of the population by the task priority when generating the population; S403, the fitness function is used to evaluate the pros and cons of each individual, and a higher fitness score is assigned to the locked task by modifying the fitness function; S404, modifying the intersection selection function to avoid selecting a position of a locked task when selecting an intersection; S405. Modify the mutation point selection function to avoid selecting the position of the locked task when selecting the mutation point, so as to ensure that the position of the locked task will not be accidentally changed.

5. The genetic algorithm scheduling task constraint method for order scheduling according to claim 4 is characterized by: The modified population generation function prioritizes the multi-objective scheduling of tasks, and generates the size of the population based on the multi-objective scheduling of tasks, and the multi-objective scheduling of the locked tasks obtained by the task locking is included in the population; The fitness function is used to sort the multi-objective scheduling of tasks and calculate the task locking fitness scores of multiple tasks; The modified intersection selection function is used to avoid the position of the locked task when selecting the intersection; The modification of the mutation point selection function includes random selection of the location of the non-locked task and weight-based selection; The randomly selecting the position of the non-locked task includes randomly selecting a position from all possible positions of the chromosome when selecting the mutation point to determine the position of the locked task; The weight-based selection is that each position is assigned a weight and the variation points are selected according to these weights.

6. The genetic algorithm scheduling task constraint method for order scheduling according to claim 5, characterized in that: The generated scheduling calculation monitoring visualization diagram is used to monitor information and visualize information during the multi-objective analysis and scheduling process of the task; The monitoring information includes task status and resource utilization; Visualize information through visualization tools for task scheduling and monitoring; Display monitoring information and visualize information through a user-friendly monitoring interface.

Images