Task ordering method and system based on task work order management

Abstract

The invention discloses a task sorting method and system based on task work order management, and the method comprises the steps: obtaining all work orders to be executed and being executed at present, and extracting work order tasks to be executed in the obtained work orders; establishing a sorting model of the work order tasks to be executed; according to the time limit of each work order, a constraint condition of the sorting model is established; and generating a final sorting result through the constraint of the constraint condition. The task scheduling accuracy and efficiency can be remarkably improved, resource waste is reduced, the reliability and stability of a work order management system are ensured, and complex production environments and variable service requirements are effectively met.

Description

Technical Field

[0001] The present invention relates to the technical field of computer platform load balancing, and in particular to a task sorting method and system based on task work order management. Background Art

[0002] In the modern industrial and information society, task management and work order scheduling have become key links in enterprise operations and production management. Enterprises usually rely on work order management systems to ensure that various tasks can be executed on time and efficiently, so as to achieve optimal resource allocation and improve production efficiency. However, with the expansion of enterprise scale and the increase in business complexity, traditional task sorting and scheduling methods face many challenges, such as task priority conflicts, low resource utilization, and inaccurate execution time. In response to these challenges, a task sorting method based on task work order management is proposed to optimize the execution order of work order tasks, ensure that tasks are completed within the specified deadline, and maximize resource utilization, which has become a technical problem that needs to be solved urgently. Summary of the invention

[0003] In view of the above-mentioned problems, the present invention is proposed.

[0004] Therefore, the technical problem solved by the present invention is that the existing task sequencing method has large time consumption and low resource utilization.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a task sorting method based on task work order management, comprising:

[0006] Obtain all currently pending and executing work orders, and extract pending work order tasks from the obtained work orders;

[0007] Establish a sorting model for pending work order tasks;

[0008] Establishing constraints of the sorting model according to the time limit of each work order;

[0009] The final sorting result is generated by the constraints of the constraints.

[0010] As a preferred solution of the task sorting method based on task work order management described in the present invention, wherein: the work order task is a subtask of the work order, and has a number that uniquely identifies the work order task;

[0011] Build a set of pending work order tasks GD in each work order g ={D g1 , D g2 , ..., D gq};

[0012] At the same time, a set of to-be-executed work order tasks in all work orders is constructed R = {R1, R2, ..., R U};

[0013] Among them, GD g represents the set of work order tasks to be executed in the g-th work order, gq represents the number of work order tasks to be executed in the g-th work order, and U represents the total number of work order tasks to be executed.

[0014] As a preferred solution of the task sorting method based on task work order management described in the present invention, the sorting model of the work order tasks to be executed includes generating a random arrangement of a set of integers from 1 to U, expressed as:

[0015] {σ(1),σ(2),…,σ(U)}

[0016] Where, each σ(u) is a unique element in {1,2,…,U} without duplication, u represents the index, and σ represents the random permutation function;

[0017] Output the tasks in a random order according to the generated index arrangement:

[0018] {R σ (1),R σ (2),…,R σ ( U )}

[0019] Suppose N random permutation sequences are generated, and the order of tasks under the same work order is checked in each sequence to see whether it is disrupted. Specifically:

[0020] In the generated sequence, the work order tasks in the same work order are screened, and the arrangement order of the screened work order tasks is checked with the task flow in the work order. If they match, the check is passed; if they do not match, the check fails, the task order is judged to be disrupted, and the entire sequence currently being checked is deleted;

[0021] After each work order is verified, n sequences that have not been deleted are obtained.

[0022] As a preferred solution of the task sorting method based on task work order management described in the present invention, the sorting model of the work order tasks to be executed also includes calculating the score of each sequence by the equipment resource utilization rate and the operation time of adjacent tasks:

[0023] F(Σ i ′)=α·U(Σ i ′)-β·T(Σ i ′)

[0024] Among them, Σ′ i represents the i-th task sequence; U(Σ′i ) represents the sequence Σ′ i Utilization of equipment resources in

[0025] T(Σ′ i ) represents the sequence Σ′ i The total operation time of all adjacent tasks in; α and β are the weight coefficients of resource utilization and operation time respectively, α+β=1;

[0026] Define a neural network MLP (T k ,T k+1 ), let k be the index of the task, identify task T k and T k+1 Whether they belong to the same process flow, the output is 1 for the same process, and the output is 0 for different processes;

[0027]

[0028] The resource utilization is calculated based on the number of identical process task pairs. The resource utilization can be expressed as the ratio of identical process task pairs:

[0029]

[0030] Use the neural network model to estimate the operation time of adjacent tasks, assuming t k,k+1 Represents task T k and Task T k+1 The operation time between:

[0031]

[0032] t k,k+1 The trained neural network model is estimated based on the historical records, expressed as:

[0033] t k,k+1 =TCN(T k ,T k+1 )

[0034] Among them, TCN (T k ,T k+1 ) represents the temporal convolutional network model according to the task T k and T k+1 The predicted operation time of the feature output, S represents the preset standard value;

[0035] Let the objective function F(Σ′ i ) maximizes the sequence:

[0036]

[0037] Among them, Σ opt represents the optimal sequence.

[0038] As a preferred solution of the task sorting method based on task work order management described in the present invention, wherein: the time limit of each work order is the deadline of each work order;

[0039] Through the temporal convolutional network, the execution time of each task in the sequence is predicted:

[0040] Get task j to synthesize a feature vector φ j ;

[0041] φ j =[τ j ,ρ j ,κ j ,δ j ]

[0042] Among them, τ j represents the task type, ρ j represents resource demand, κ j represents the type of the previous task, δ j represents the historical execution time; the task sequence is represented by Φ={φ1,φ2,…,φ n};

[0043] Input task sequence Φ = {φ1, φ2, …, φ n};

[0044] The model outputs the predicted execution time of each task. The predicted execution time of the jth task is expressed as

[0045] As a preferred solution of the task sorting method based on task work order management described in the present invention, wherein: the constraint condition includes that each work order is completed within the deadline;

[0046] For any work order, the following conditions must be met:

[0047]

[0048] Where L represents the index number of the last task of the current work order in the task sequence; JZ p Indicates the deadline of the current work order.

[0049] As a preferred solution of the task sorting method based on task work order management described in the present invention, wherein: the final sorting result includes: using the constraint condition to sort the objective function F(Σ′ i ) is constrained, and the sequence obtained by solving the objective function is used as the final sorting result.

[0050] A task sorting system based on task work order management using the method described in the present invention is characterized by:

[0051] The task generation unit obtains each work order task of the complete work order and the status of the work order task;

[0052] A prediction unit, which evaluates the current progress of the work order task and predicts the schedule of other tasks that have not yet been executed;

[0053] The output unit combines the evaluated current task progress with the other task schedules to generate a prediction for the completion of the work order.

[0054] A computer device comprises: a memory and a processor; the memory stores a computer program, wherein the processor implements the steps of any one of the methods of the present invention when executing the computer program.

[0055] A computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of any one of the methods of the present invention.

[0056] Beneficial effects of the present invention: The task sorting method based on task work order management provided by the present invention obtains the work order tasks to be executed and establishes a sorting model to ensure that the tasks are completed within the specified deadline. The temporal convolutional network (TCN) is introduced to predict the task execution time, and the multi-layer perceptron (MLP) is used to identify the tasks with the same process flow, which effectively improves the utilization rate of equipment resources. By setting reasonable constraints, the execution order of adjacent tasks is optimized and the switching time between tasks is reduced. The method of the present invention can significantly improve the accuracy and efficiency of task scheduling, reduce resource waste, ensure the reliability and stability of the work order management system, and effectively cope with complex production environments and changing business needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0058] Figure 1 An overall flow chart of a task sorting method based on task work order management provided in the first embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the above-mentioned purposes, 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 drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0060] Example 1, reference Figure 1 , is an embodiment of the present invention, and provides a task sorting method based on task work order management, comprising:

[0061] S1: Obtain all currently pending and executing work orders, and extract pending work order tasks from the obtained work orders.

[0062] The work order task is a subtask of the work order and has a number that uniquely identifies the work order task.

[0063] Build a set of pending work order tasks GD in each work order g ={D g1 , D g2 , ..., D gq}.

[0064] At the same time, a set of to-be-executed work order tasks in all work orders is constructed R = {R1, R2, ..., R U}.

[0065] Among them, GD g represents the set of work order tasks to be executed in the g-th work order, gq represents the number of work order tasks to be executed in the g-th work order, and U represents the total number of work order tasks to be executed.

[0066] By assigning a unique identification number to each work order task, each task can be accurately tracked and managed to prevent task confusion or omission. This method ensures that each task has a clear identity in the system, making it easier to manage and query. Organizing the tasks to be performed in each work order into sets can clearly understand the task composition of each work order. This set design facilitates batch operations on tasks, such as sorting, priority assignment, and progress tracking, effectively simplifying the task management process.

[0067] The purpose of building a collection of pending tasks in all work orders is to achieve global optimization. By integrating all pending tasks, resource allocation, task sorting, and scheduling can be effectively optimized to ensure efficient resource utilization and timely completion of tasks. This global view helps balance resource requirements between different work orders and avoid resource conflicts and task delays. This structured task collection design provides a basic data structure for subsequent sorting models and optimization algorithms, allowing the algorithm to more effectively prioritize and optimize tasks, ensure tasks are completed on time, and improve the overall efficiency of work order management.

[0068] S2: Establish a sorting model for work order tasks to be executed.

[0069] Furthermore, the sorting model of the work order tasks to be executed includes generating a random permutation of a set of integers from 1 to U, expressed as:

[0070] {σ(1),σ(2),…,σ(U)}

[0071] Here, each σ(u) is a unique element in {1,2,…,U} without duplication, u represents the index, and σ represents the random permutation function.

[0072] Output the tasks in a random order according to the generated index arrangement:

[0073] {R σ(1) ,R σ(2) ,…,R σ(U)}

[0074] Suppose N random permutation sequences are generated, and the order of tasks under the same work order is checked in each sequence to see whether it is disrupted. Specifically:

[0075] In the generated sequence, the work order tasks in the same work order are screened, and the order of the screened work order tasks is checked with the task flow in the work order. If they match, the check is passed; if not, the check fails, the task order is disrupted, and the entire sequence currently being checked is deleted. After each work order is checked, n sequences that have not been deleted are obtained.

[0076] What needs to be said is that by generating a random sequence and checking the order of tasks in each work order, the execution order of tasks in the same work order is ensured not to be disrupted. This method ensures the logic and continuity of tasks and avoids execution errors or work order process confusion caused by disordered tasks.

[0077] Furthermore, the score of each sequence is calculated by the device resource utilization and the adjacent task operation time:

[0078] F(Σ i ′)=α·U(Σ i′)-β·T(Σ i ′)

[0079] Among them, Σ′ i represents the i-th task sequence; U(Σ′ i ) represents the sequence Σ′ i The utilization rate of equipment resources in T(Σ′ i ) represents the sequence Σ′ i is the total operation time of all adjacent tasks in ; α and β are the weight coefficients of resource utilization and operation time respectively, α+β=1.

[0080] Define a neural network MLP (T k ,T k+1 ), let k be the index of the task, identify task T k and T k+1 Whether they belong to the same process flow, the output is 1 for the same process, and the output is 0 for different processes:

[0081]

[0082] The resource utilization is calculated based on the number of identical process task pairs. The resource utilization can be expressed as the ratio of identical process task pairs:

[0083]

[0084] Use the neural network model to estimate the operation time of adjacent tasks, assuming t k,k+1 Represents task T k and Task T k+1 The operation time between:

[0085]

[0086] t k,k+1 The trained neural network model is estimated based on the historical records, expressed as:

[0087] t k,k+1 =TCN(T k ,T k+1 )

[0088] Among them, TCN (T k ,T k+1 ) represents the temporal convolutional network model according to the task T k and T k+1 The predicted operation time of the characteristic output, S represents the preset standard value. Let the objective function F(Σ i ) maximizes the sequence:

[0089]

[0090] Among them, Σ opt represents the optimal sequence.

[0091] It is important to know that the calculation of equipment resource utilization is introduced, and a neural network (MLP) is used to identify whether the tasks belong to the same process flow. Task pairs of the same process flow can reduce the switching time of the equipment and improve resource utilization. By calculating the proportion of task pairs with the same process, the efficiency of resource utilization can be quantified, thereby guiding the optimization of task sorting. The time convolutional network (TCN) model is used to estimate the operation time of adjacent tasks. These time estimates reflect the overhead when switching tasks. Including operation time in the optimization target can help reduce unnecessary task switching, thereby improving the overall efficiency of task execution.

[0092] By calculating the score of each sequence, taking into account resource utilization and task switching time, a weight coefficient is set to balance these two factors. The goal is to find an optimal task sorting sequence that maximizes resource utilization and minimizes task switching time. Such a design can improve production efficiency and reduce operating costs. By generating multiple random permutation sequences and screening qualified sequences, the design can flexibly adapt to different task requirements and work order structures. This method can handle a large number of work order tasks, adapt to a variety of task combinations, and provide a general sorting optimization solution.

[0093] S3: Establish constraint conditions of the sorting model according to the time limit of each work order.

[0094] It should be noted that the time limit for each work order is the deadline for each work order.

[0095] Through the temporal convolutional network, the execution time of each task in the sequence is predicted:

[0096] Get task j to synthesize a feature vector φ j .

[0097] φ j =[τ j ,ρ j ,κ j ,δ j ]

[0098] Among them, τ j represents the task type, ρ j represents resource demand, κ j represents the type of the previous task, δ j represents the historical execution time; the task sequence is represented by Φ={φ1,φ2,…,φ n}.

[0099] Input task sequence Φ = {φ1, φ2, …, φ nThe model outputs the predicted execution time of each task. The predicted execution time of the jth task is expressed as

[0100] The constraints include that each work order must be completed within the deadline. For any work order, the following conditions must be met:

[0101]

[0102] Where L represents the index number of the last task of the current work order in the task sequence; JZ p Indicates the deadline of the current work order.

[0103] Each work order has a clear deadline, and constraints are set to ensure that all tasks are completed before the deadline. This approach ensures the time management of tasks and avoids delays caused by task timeouts, thereby improving the response efficiency and reliability of the entire system. The execution time of each task is predicted through a temporal convolutional network (TCN). TCN can capture the time dependencies in the task sequence and generate accurate execution time predictions for each task. These predictions provide data support for task scheduling, making task sorting and resource allocation more reasonable. Obtain the relevant features of the task (such as task type, resource requirements, previous task type, historical execution time, etc.), and synthesize these features into a feature vector. This vectorized processing method allows the model to better understand and learn the relationship between tasks and improve the accuracy of predictions.

[0104] A constraint condition is introduced into the sorting model, that is, the last task of each work order must be completed before the deadline. This constraint can effectively prevent the work order timeout problem caused by task sorting. The existence of the constraint condition makes the sorting model not only focus on the efficiency of task execution, but also consider the feasibility of time to ensure that the result is realistic and operational. In a multi-task environment, different tasks may have different execution times and time limits. By predicting the execution time of each task and setting time constraints, it is possible to flexibly respond to complex and changing task scenarios and ensure that tasks are completed on time without affecting overall efficiency.

[0105] S4: Generate a final sorting result through the constraints of the constraint conditions.

[0106] Specifically, the objective function F(Σ′ i ) is constrained, and the sequence obtained by solving the objective function is used as the final sorting result.

[0107] On the other hand, the present embodiment further provides a task sorting system based on task work order management, which includes: a task generation unit, which obtains each work order task of a complete work order and simultaneously obtains the status of the work order task.

[0108] The prediction unit evaluates the current progress of the work order task and predicts the schedule of other tasks that have not yet been executed.

[0109] The output unit combines the evaluated current task progress with the other task schedules to generate a prediction for the completion of the work order.

[0110] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0111] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0112] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0113] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0114] Example 2 is an embodiment of the present invention, which provides a task sorting method based on task work order management. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0115] A scheduling scenario of a manufacturing enterprise production line was selected for simulation experiment. The production line involves multiple work orders, each of which contains several subtasks, and the task types cover equipment debugging, parts assembly, quality inspection, etc. The experiment aims to optimize the task sorting through the method of the present invention, so that each task can be completed within the deadline, while improving the utilization of equipment resources and reducing the time of task switching. Part of the experimental data is recorded in Table 1.

[0116] Data collection: Obtain the characteristic data of relevant tasks from the enterprise production line, including task type, resource requirements, historical execution time, previous task type, etc. A total of 10 work orders were collected, marked as W1 to W10, each work order contains 2 to 5 subtasks, totaling 30 tasks. The deadline and resource requirements of each task are recorded.

[0117] Feature vector construction: Generate feature vectors for each task. Features include task type (such as equipment debugging, parts assembly, quality inspection), required resource type (such as robotic arms, laser welding machines, test instruments, etc.), previous task type, and historical execution time (in hours). These feature vectors will be used as input to the temporal convolutional network (TCN) to predict task execution time.

[0118] Temporal Convolutional Network Training: Use the collected historical execution data to train a temporal convolutional network model to predict the execution time of tasks. The training dataset includes the task feature vector and the actual execution time. The model continuously adjusts parameters to optimize the prediction accuracy.

[0119] According to the invented method, 30! random permutations are generated from a set of integers from 1 to 30. Where! represents factorial.

[0120] Sequence screening: Check the generated 30 sequences to see if the order of tasks under the same work order in each sequence is disrupted. After screening, 50 sequences are retained to meet the requirement that the order of tasks within the same work order is not disrupted.

[0121] Equipment resource utilization and operation time evaluation: Define a multi-layer perceptron (MLP) neural network to identify whether the task pair belongs to the same process flow. Calculate resource utilization through the process identification results output by the MLP. At the same time, use the trained temporal convolutional network to predict the operation time of adjacent tasks. For each screened sequence, calculate the equipment resource utilization and task switching time, and calculate the sequence score based on this.

[0122] Objective function optimization: Set the weight coefficients to 0.5 so that the sum of resource utilization and operation time weights is 1. Based on the calculated scores, select the sequence with the highest score as the optimal sorting solution.

[0123] Table 1 Experimental data record table

[0124]

[0125] The analysis of the table data clearly shows the innovation and advantages of the present invention in task sequencing and work order management.

[0126] First, the execution time of each task is predicted by the temporal convolutional network (TCN) and compared with the historical execution time, showing a high prediction accuracy. For example, the historical execution time of W1-Task1 is 90 minutes, while the execution time predicted by TCN is 88 minutes; the historical execution time of W2-Task2 is 150 minutes, and the predicted time is 148 minutes. These data show that TCN can effectively capture the time dependency of tasks and give a relatively accurate prediction of execution time. This high-precision time prediction can improve the rationality of task sorting in actual task scheduling and reduce task delays caused by inaccurate time estimation.

[0127] Secondly, the data shows that the task sorting performed by the method of the present invention can improve the utilization of equipment resources. By identifying task pairs of the same process flow through a multi-layer perceptron (MLP) neural network, the switching and waiting time of the equipment can be reduced. Tasks W1-Task2 and W3-Task2 in the table both use assembly machine B, and the predicted execution times are 115 minutes and 130 minutes, respectively. Through the same process identification, MLP can identify that these tasks belong to the same process flow, so as to arrange them to be executed in close time periods, thereby improving the utilization efficiency of assembly machine B. Compared with random sorting, this optimization can reduce equipment idle time and repeated setup time, and improve resource utilization by an average of 20%.

[0128] In addition, the method of the present invention also shows significant advantages in terms of task switching time. With the optimized sorting sequence, the task switching time between tasks W1-Task2 and W1-Task3 is reduced from the original average of 10 minutes to 5 minutes. Through the accurate estimation of the operation time of adjacent tasks by the temporal convolutional network, the task switching time is effectively controlled, the connection between tasks is closer, and the execution efficiency of the entire work order is improved.

[0129] 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 task sorting method based on task work order management, characterized in that: include: Obtain all currently pending and executing work orders, and extract pending work order tasks from the obtained work orders; Establish a sorting model for pending work order tasks; Establishing constraints of the sorting model according to the time limit of each work order; The final sorting result is generated by the constraints of the constraints.

2. The task sorting method based on task work order management according to claim 1, characterized in that: The work order task is a subtask of the work order and has a number that uniquely identifies the work order task; Build a set of pending work order tasks GD in each work order g ={D g1 , D g2 , ..., D gq }; At the same time, a set of to-be-executed work order tasks in all work orders is constructed R = {R1, R2, ..., R U }; Among them, GD g represents the set of work order tasks to be executed in the g-th work order, gq represents the number of work order tasks to be executed in the g-th work order, and U represents the total number of work order tasks to be executed.

3. The task sorting method based on task work order management as claimed in claim 2, characterized in that: The sorting model of the work order tasks to be executed includes generating a random arrangement of a set of integers from 1 to U, expressed as: {σ(1),σ(2),…,σ(U)} in, Each σ(u) is a unique element in {1,2,…,U} without duplication, u represents the index, and σ represents the random permutation function; Output the tasks in a random order according to the generated index arrangement: {R σ(1) ,R σ(2) ,…,R σ(U) } Suppose N random permutation sequences are generated, and the order of tasks under the same work order is checked in each sequence to see whether it is disrupted. Specifically: In the generated sequence, the work order tasks in the same work order are screened, and the arrangement order of the screened work order tasks is checked with the task flow in the work order. If they match, the check is passed; if they do not match, the check fails, the task order is judged to be disrupted, and the entire sequence currently being checked is deleted; After each work order is verified, n sequences that have not been deleted are obtained.

4. The task sorting method based on task work order management as claimed in claim 3, characterized in that: The sorting model of the work order tasks to be executed also includes calculating the score of each sequence by the equipment resource utilization and the operation time of adjacent tasks: F(∑ i ′)=α·(∑ i ′)-β·T(∑ i ′) Among them, Σ′ i represents the i-th task sequence; U(Σ′ i ) represents the sequence Σ′ i The utilization rate of equipment resources in T(Σ′ i ) represents the sequence Σ′ i The total operation time of all adjacent tasks in; α and β are the weight coefficients of resource utilization and operation time respectively, α+β=1; Define a neural network MLP (T k ,T k+1 ), let k be the index of the task, identify task T k and T k+1 Whether they belong to the same process flow, the output is 1 for the same process, and the output is 0 for different processes; The resource utilization is calculated based on the number of identical process task pairs. The resource utilization can be expressed as the ratio of identical process task pairs: Use the neural network model to estimate the operation time of adjacent tasks, assuming t k,k+1 Represents task T k and Task T k+1 The operation time between: t k,k+1 The trained neural network model is estimated based on the historical records, expressed as: t k,k+1 =BC(T) k ,T k+1 ) Among them, TCN (T k ,T k+1 ) represents the temporal convolutional network model according to the task T k and T k+1 The predicted operation time of the feature output, S represents the preset standard value; Let the objective function F(Σ′ i ) maximizes the sequence: Among them, Σ opt represents the optimal sequence.

5. The task sorting method based on task work order management as claimed in claim 4, characterized in that: The time limit of each work order is the deadline of each work order; Through the temporal convolutional network, the execution time of each task in the sequence is predicted: Get task j to synthesize a feature vector φ j ; f j =[τ j ,r j ,k j ,d j ] Among them, τ j represents the task type, ρ j represents resource demand, κ j represents the type of the previous task, δ j represents the historical execution time; the task sequence is represented by Φ={φ1,φ2,…,φ n }; Input task sequence Φ = {φ1, φ2, …, φ n }; The model outputs the predicted execution time of each task. The predicted execution time of the jth task is expressed as 6. The task sorting method based on task work order management according to claim 5, characterized in that: The constraints include that each work order is completed within the deadline; For any work order, the following conditions must be met: Where L represents the index number of the last task of the current work order in the task sequence; JZ p Indicates the deadline of the current work order.

7. The task sorting method based on task work order management according to claim 6, characterized in that: The final sorting result includes: using the constraint condition to sort the objective function F(Σ′ i ) is constrained, and the sequence obtained by solving the objective function is used as the final sorting result.

8. A task sorting system based on task work order management using the method according to any one of claims 1 to 7, characterized in that: The task generation unit obtains each work order task of the complete work order and the status of the work order task; A prediction unit, which evaluates the current progress of the work order task and predicts the schedule of other tasks that have not yet been executed; The output unit combines the evaluated current task progress with the other task schedules to generate a prediction for the completion of the work order.

9. A computer device comprising: Memory and processor; The memory stores a computer program, characterized in that: when the processor executes the computer program, the steps of the task sorting method based on task work order management are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the task sorting method based on task work order management are implemented.

Images