Maintenance task scheduling method and system

Abstract

The invention relates to the technical field of electric data processing, in particular to a maintenance task scheduling method and system.The maintenance task scheduling method comprises the steps that chromosomes composed of a to-be-scheduled maintenance task set are obtained, and all genes in the chromosomes represent scheduling schemes corresponding to the genes; and for any chromosome, calculating a mean value of initial fitness functions of all genes in the chromosome as a final fitness function of the chromosome, and carrying out iteration through screening, crossing, variation and population updating in a genetic algorithm to achieve a set algebra or fitness convergence so as to obtain an optimal chromosome. According to the method, individuals capable of performing crossover and mutation operation are screened according to the final fitness function of the corresponding chromosome, so that generation of illegal offspring chromosomes is reduced fundamentally, the search efficiency of a genetic algorithm under the background of maintenance task scheduling is improved, and the quality of a final output scheduling scheme is improved.

Description

Technical Field

[0001] This invention relates to the field of electrical data processing technology. More specifically, this invention relates to a maintenance task scheduling method and system. Background Technology

[0002] In the field of large-scale equipment operation and maintenance (such as power systems, industrial production lines, and communication base stations), the scheduling of maintenance tasks is a core management activity. Its goal is to efficiently allocate limited maintenance resources, such as maintenance technicians and necessary tools and equipment, to a series of pending maintenance tasks. The efficiency of maintenance task scheduling directly affects equipment availability and operation and maintenance costs. For example, generator set or power transmission and transformation equipment failures can lead to regional power outages, causing significant economic losses and social impacts, thus requiring intelligent scheduling optimization support.

[0003] Genetic algorithms are global optimization search algorithms based on biological evolution and genetic mechanisms. Their core idea is to iteratively search for the optimal solution in the solution space by simulating the processes of natural selection (survival of the fittest), heredity (gene transmission), and variation (gene mutation).

[0004] In the context of maintenance task scheduling, the crossover and mutation operations in genetic algorithms are performed in a "blind" manner. Standard crossover generates offspring by randomly exchanging gene segments from two parent chromosomes, while mutation randomly alters the attributes of a gene. These operations completely disregard the spatiotemporal uniqueness constraints of resources (engineers and equipment). Therefore, they are highly prone to producing "illegal" offspring chromosomes, such as assigning the same engineer or the same piece of equipment to two different tasks within the same timeframe, causing logical resource conflicts. This leads to conflicts in maintenance task scheduling schemes, reducing the efficiency of maintenance task completion and causing incalculable losses. Summary of the Invention

[0005] This invention provides a maintenance task scheduling method and system, aiming to solve the problem in related technologies that the execution of cross and variation operations is "blind," completely disregarding the spatiotemporal uniqueness constraints of resources (engineers and equipment), causing logical resource conflicts, leading to conflicts in maintenance task scheduling schemes, and thus reducing the efficiency of completing maintenance tasks.

[0006] In a first aspect, the present invention provides a maintenance task scheduling method, comprising: obtaining a chromosome composed of a set of maintenance tasks to be scheduled, wherein each gene in the chromosome represents a scheduling scheme corresponding to that gene; for any chromosome, calculating the mean of the initial fitness functions of all genes within the chromosome as the final fitness function of the chromosome, and iterating through screening, crossover, mutation, and population update in a genetic algorithm until convergence of a set number of generations or fitness is achieved, thereby obtaining an optimal chromosome, wherein the optimal chromosome is the optimal scheduling scheme; wherein the initial fitness function of the gene is positively correlated with the time efficiency index and risk control index of the scheduling scheme corresponding to the gene, the time efficiency index is negatively correlated with the idle waiting time of the engineer in the scheduling scheme corresponding to the gene, and the risk control index is positively correlated with the buffer time of the engineer after completing the maintenance task in the scheduling scheme corresponding to the gene. Individuals capable of crossover and mutation operations are selected based on the final fitness evaluation results of the corresponding chromosome, thereby reducing the generation of "illegal" offspring chromosomes at the source, improving the search efficiency of the genetic algorithm in the context of maintenance task scheduling, and improving the quality of the final output scheduling scheme.

[0007] Furthermore, a scheduling scheme includes an engineer ID, a device ID, a planned start time, and a planned end time, wherein the planned end time is the sum of the planned start time and the estimated working duration.

[0008] Furthermore, the method for obtaining the estimated working time includes: obtaining a set of all historical maintenance tasks with the same engineer ID and equipment ID as the current maintenance task; and using the average actual maintenance time of the historical maintenance tasks as the estimated working time. By filtering historical maintenance records of the same engineer and equipment as the current task and using the average of their actual maintenance times as the estimated time for the current task, the experience level of individual engineers and the maintenance characteristics of the equipment can be fully combined, thereby improving the accuracy of the estimated time and avoiding deviations caused by individual differences or different equipment complexity.

[0009] Furthermore, the initial fitness function of the gene is positively correlated with the resource utilization index of the corresponding scheduling scheme (the greater the excess, the lower the index, indicating unreasonable resource utilization and overutilization of talent). The resource utilization index is negatively correlated with the extent to which the engineer's maintenance skill level exceeds the equipment's required skill level. By measuring whether the engineer's skill level is significantly higher than the equipment's maintenance requirements, the resource waste of "highly skilled engineers doing low-skill work" is avoided, encouraging optimal matching of skill resources while meeting maintenance needs, thereby improving the overall scheduling economy and system efficiency.

[0010] Furthermore, the initial fitness function of the gene is positively correlated with the travel cost index of the corresponding scheduling scheme, while the travel cost index is negatively correlated with the distance between the current equipment location and the engineer's previous maintenance task location. By minimizing the engineer's travel distance to reduce travel costs, scheduling consistency and overall maintenance efficiency are effectively improved.

[0011] Furthermore, the method for obtaining the buffer time after an engineer completes a maintenance task in the corresponding scheduling scheme includes: if the maintenance task is the engineer's last maintenance task, the buffer time is a set value; if the maintenance task is not the engineer's last maintenance task, the buffer time is the difference between the planned start time and the planned completion time of the engineer's next maintenance task. By dynamically calculating the buffer time, the rationality of task connection can be accurately measured, improving the utilization rate of scheduling time and the feasibility of engineer scheduling.

[0012] Furthermore, the idle waiting time of the engineer in the scheduling scheme corresponding to this gene is obtained as follows: if the engineer has no other maintenance tasks before executing the maintenance task, the space waiting time is 0; if the engineer has other maintenance tasks before executing the maintenance task, the latest planned end time among all the maintenance tasks executed by the engineer is obtained, and the difference between the planned start time of the maintenance task and the latest time is used as the space waiting time.

[0013] Furthermore, the genetic algorithm involves screening, including using roulette wheel selection to select parent individuals.

[0014] Furthermore, the set value is 8 hours. In a second aspect, the present invention also provides a maintenance task scheduling system, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the maintenance task scheduling method described in any of the above claims.

[0015] Beneficial effects: By considering multiple influencing factors before executing genetic operations, an initial fitness function for each scheduling scheme is obtained. Based on the initial fitness evaluation results of each scheduling scheme, the final fitness function of the corresponding chromosome is obtained. Individuals capable of crossover and mutation operations are selected based on the final fitness function of the corresponding chromosome, thereby reducing the generation of "illegal" offspring chromosomes at the source, improving the search efficiency of the genetic algorithm in the context of maintenance task scheduling, and enhancing the quality of the final output scheduling scheme. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the process of obtaining the optimal scheduling scheme according to an embodiment of the present invention; Figure 2This is a flowchart illustrating a computation time efficiency index according to an embodiment of the present invention. Detailed Implementation

[0017] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0018] like Figure 1 As shown, S101: Constructing chromosomes.

[0019] In one embodiment, a set of maintenance tasks to be scheduled is obtained, where each maintenance task corresponds to a scheduling scheme. Specifically, the scheduling scheme includes an engineer ID, an equipment ID, a planned start time, and a planned end time. Each scheduling scheme in the maintenance task set is treated as a gene, thereby obtaining the chromosome composed of all scheduling schemes in the maintenance task set.

[0020] For example, with the first Taking the 1st gene as an example, the 2nd gene... The scheduling scheme for each gene is as follows: .in, Indicates the first The scheduling scheme corresponding to the gene Indicates the first The engineer ID corresponding to the gene. Indicates the first The device ID is assigned based on the gene. After determining the engineer and device match, a list scheduling algorithm is used to calculate the time window without conflicts. : Indicates the first The start time of the plan corresponding to the gene, among which . Indicates the first The project's end time corresponding to the gene. , This indicates the estimated working time.

[0021] In one embodiment, the estimated working time is obtained by: acquiring a set of all historical maintenance tasks with the same engineer ID and device ID as the current maintenance task; and using the average of the actual maintenance times in the set of historical maintenance tasks as the estimated working time. The calculation formula is as follows: ;in, It is the first Estimated working time for each gene . Representing historical engineers The actual repair time is in hours. , indicating the current number in the historical record. The same engineer ID in the gene ( And the device IDs are the same. The collection of all historical tasks. Represents a set The size of this value represents the number of historical matching tasks. If there are no historical matching records, then... If so, the industry benchmark time for the device type or the average processing efficiency of the engineer is used as the default value.

[0022] Finally, an initial population is randomly generated, in which each chromosome satisfies basic hard constraints (such as skill matching between engineers and corresponding maintenance tasks, equipment model matching, and no initial time conflicts). Each chromosome corresponds to a complete scheduling scheme. chromosomes (i.e.) A complete scheduling scheme constitutes an initial population.

[0023] S102: Calculate the final fitness function for each chromosome.

[0024] In one embodiment, for any chromosome, the mean of the initial fitness functions of all genes within that chromosome is calculated, and this is used as the final fitness function of that chromosome. The initial fitness function of a gene is positively correlated with both the time efficiency index and the risk control index of the corresponding scheduling scheme. The time efficiency index is negatively correlated with the idle waiting time of engineers in the corresponding scheduling scheme, and the risk control index is positively correlated with the buffer time after an engineer completes the maintenance task in the corresponding scheduling scheme.

[0025] like Figure 2 As shown, S1021: Calculate the time efficiency index of the scheduling scheme corresponding to this gene.

[0026] In one embodiment, the idle waiting time of the engineer in the scheduling scheme corresponding to this gene is obtained as follows: if the engineer has no other maintenance tasks before executing the maintenance task, the space waiting time is 0; if the engineer has other maintenance tasks before executing the maintenance task, the latest planned end time among all maintenance tasks executed by the engineer is obtained, and the difference between the planned start time of the maintenance task and the latest time is used as the space waiting time. The formula is expressed as: ;in, He is an engineer Execute the The time of the last repair task before the repair task in the gene. This indicates that all engineers A set of maintenance tasks to be performed. They represent the first The planned start and end times of the repair task in the gene. Indicates the current number The planned start time for the repair task in the gene.

[0027] In one embodiment, a formula is provided for calculating the time efficiency index of the scheduling scheme corresponding to the gene. The formula is as follows: .in, It is the first The time efficiency metrics of the scheduling scheme corresponding to the gene. It is the attenuation coefficient. , Indicates the first The start time of the plan corresponding to the gene. He is an engineer Execute the The time of the last repair task preceding the repair task in the gene. This represents an exponential function with base e. If this is the engineer's first task, the idle time interval is 0, because the engineer can immediately begin the next task. Repair tasks will be completed; otherwise, waiting is required. The idle time is one hour. The longer the idle waiting time, the more... Maintenance task (i.e., the first) The smaller the time efficiency index of the scheduling scheme corresponding to the gene, the better. In other words, the lower the time efficiency index... This reflects the efficiency of the task in terms of time management and penalizes engineers for idle time. Encourage time-sensitive scheduling ( The decay function ensures The smaller, The larger (the larger).

[0028] S1022: Calculate the risk control index of the scheduling scheme corresponding to this gene.

[0029] In one embodiment, the method for obtaining the buffer time after an engineer completes a maintenance task in the gene-corresponding scheduling scheme includes: if the maintenance task is the engineer's last maintenance task, the buffer time is a set value, wherein the set value is 8 hours, and in other embodiments the set value can also be 10 hours; if the maintenance task is not the engineer's last maintenance task, the buffer time is the difference between the planned start time of the engineer's next maintenance task and the planned completion time of the maintenance task (the planned start time of the next maintenance task minus...). ). In one embodiment, a formula for calculating the risk control index of the scheduling scheme corresponding to the gene is also provided. The formula is as follows: .in, It is the first Maintenance task (i.e., the first) Risk control indicators for scheduling schemes corresponding to genes. For buffer time, This is a time constant. If this maintenance task is the engineer's last maintenance task, then the buffer time... The set value is 8 hours (meaning there is sufficient buffer time), otherwise, the buffer time is... Subtract the start time of the next task The buffer time for the last maintenance task should be dynamically set based on the total scheduling time; 8 hours here is an example value. The larger the buffer time, the more likely the last maintenance task will be completed. Maintenance task (i.e., the first) The higher the risk control index of the scheduling scheme corresponding to the gene, the better the risk control, and vice versa.

[0030] At this point, the time efficiency index and risk control index of the scheduling scheme corresponding to the gene can be calculated. Then, the average of the sum of the time efficiency index and the risk control index is used as the initial fitness function of the gene. Finally, the mean of all initial fitness functions corresponding to all genes in the chromosome is used as the final fitness function of the chromosome.

[0031] In another embodiment, the initial fitness function of the gene is positively correlated not only with the time efficiency index and the risk control index, but also with the resource utilization index and the route cost index of the scheduling scheme corresponding to the gene. Then, the average value of the sum of the resource utilization index, route cost index, time efficiency index, and risk control index of the scheduling scheme corresponding to the gene is calculated, and the average value is used as the initial fitness function of the gene.

[0032] In one embodiment, a formula for calculating resource utilization indicators is provided, and the formula is as follows: In the formula, It is the first Resource utilization metrics for the scheduling scheme corresponding to the gene. For the first The skill level of the engineer in the scheduling scheme corresponding to the gene. No. The skill level required for the maintenance task corresponding to the gene's scheduling scheme. (If the engineer's skill level...) Exactly equal to the skill level required for the task ,but If the engineer's skill level exceeds the limit, then (The more it exceeds the limit, the lower the indicator, indicating unreasonable resource utilization and underutilization of talent). If the engineer's skill level is lower than the task requirements, the hard constraint is not met, but since the initial population has already met the hard constraint, this will not occur.

[0033] In one embodiment, a formula for calculating the distance cost index is provided, and the formula is as follows: ;in, It is the first The path cost metric for the scheduling scheme corresponding to the gene. It is the distance between the location of the maintenance task in the scheduling scheme and the location of the engineer's previous maintenance task (if it is the first maintenance task, it is the distance from the engineer's initial location to the location of the maintenance task). In this embodiment, the distance is a constant. The value is 3 kilometers, and engineers within a 3-kilometer radius are given priority in assignment.

[0034] S103: Obtain the optimal chromosome based on the genetic algorithm, and then obtain the optimal scheduling scheme.

[0035] In one embodiment, for any chromosome, the mean of the initial fitness functions of all genes within that chromosome is calculated as the final fitness function of the chromosome. This final fitness function is then iterated through screening, crossover, mutation, and population updates in a genetic algorithm until a set number of generations or fitness convergence is achieved, resulting in the optimal chromosome. This optimal chromosome is the optimal scheduling scheme. The screening, crossover, mutation, and population updates in the genetic algorithm are existing technologies and will not be elaborated upon here.

[0036] By employing the above steps, a fitness function is constructed by integrating four key indicators: time efficiency, resource utilization, travel cost, and risk control. This overcomes the limitations of traditional scheduling, which focuses only on a single objective, and generates a balanced and practical scheduling scheme. In the initial population generation and genetic operations (local repair after crossover / mutation), hard constraints such as "engineer-task skill matching," "equipment model matching," and "no time conflicts" are enforced, fundamentally avoiding the search for invalid solutions and accelerating the convergence process.

[0037] The present invention also provides a maintenance task scheduling system. The system includes a processor and a memory, the memory storing computer program instructions, which, when executed by the processor, implement a maintenance task scheduling method according to the first aspect of the present invention.

[0038] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and therefore will not be described in detail here.

[0039] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions stored or otherwise maintained on such a computer-readable medium.

[0040] The embodiments described above are merely examples of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A maintenance task scheduling method, characterized in that, include: Obtain a chromosome consisting of a set of maintenance tasks to be scheduled, wherein each gene in the chromosome represents the scheduling scheme corresponding to that gene; For any chromosome, the mean of the initial fitness function of all genes in the chromosome is calculated as the final fitness function of the chromosome. The chromosome is then iterated through screening, crossover, mutation and population update in the genetic algorithm until the set number of generations or fitness convergence is reached, and the optimal chromosome is obtained. The optimal chromosome is the optimal scheduling scheme. The initial fitness function of the gene is positively correlated with the time efficiency index and risk control index of the scheduling scheme corresponding to the gene. The time efficiency index is negatively correlated with the idle waiting time of the engineer in the scheduling scheme corresponding to the gene. The risk control index is positively correlated with the buffer time of the engineer after completing the maintenance task in the scheduling scheme corresponding to the gene.

2. The maintenance task scheduling method according to claim 1, characterized in that, A scheduling scheme includes an engineer ID, a device ID, a planned start time, and a planned end time, wherein the planned end time is the sum of the planned start time and the estimated working duration.

3. The maintenance task scheduling method according to claim 2, characterized in that, Methods for obtaining estimated working hours include: Obtain a set of all historical maintenance tasks that have the same engineer ID and device ID as the current maintenance task, and use the average of the actual maintenance time in the set of historical maintenance tasks as the estimated working time.

4. The maintenance task scheduling method according to claim 1, characterized in that, The initial fitness function of the gene is positively correlated with the resource utilization index of the scheduling scheme corresponding to the gene, and the resource utilization index is negatively correlated with the extent to which the engineer's maintenance skill level exceeds the skill level required by the equipment.

5. The maintenance task scheduling method according to claim 1 or 4, characterized in that, The initial fitness function of the gene is also positively correlated with the travel cost index of the scheduling scheme corresponding to the gene, and the travel cost index is negatively correlated with the distance between the current equipment location and the engineer's previous maintenance task location.

6. The maintenance task scheduling method according to claim 2, characterized in that, The method for obtaining the buffer time after the engineer completes the maintenance task in the corresponding scheduling scheme includes: If the maintenance task is the engineer's last maintenance task, the buffer time is set to a specific value. If the maintenance task is not the engineer's last maintenance task, the buffer time is the difference between the planned start time of the engineer's next maintenance task and the planned completion time of the maintenance task.

7. The maintenance task scheduling method according to claim 2, characterized in that, The method for obtaining the engineer's idle waiting time in the scheduling scheme corresponding to this gene is as follows: If the engineer has no other maintenance tasks before performing this maintenance task, the space wait time is 0. If the engineer has other maintenance tasks before performing this maintenance task, obtain the latest planned end time among all the maintenance tasks performed by the engineer, and use the difference between the planned start time of the maintenance task and the latest time as the space waiting time.

8. The maintenance task scheduling method according to claim 1, characterized in that, After screening using a genetic algorithm, including: The roulette wheel selection method is used to select parent individuals.

9. The maintenance task scheduling method according to claim 6, characterized in that, The setting is 8 hours.

10. A maintenance task scheduling system, comprising a processor and a memory, characterized in that, The memory stores a computer program, and the processor executes the computer program to implement the maintenance task scheduling method as described in any one of claims 1-9.

Images