A method, device and equipment for generating a maintenance plan and a storage medium

Abstract

The application discloses a maintenance scheme generation method, device and equipment and a storage medium. The method comprises the following steps: obtaining target information of a target elevator, wherein the target information comprises maintenance information; obtaining a total set of maintenance items, wherein the total set of maintenance items comprises a plurality of maintenance items; determining matched maintenance items and unmatched maintenance items from the total set of maintenance items according to the maintenance information, determining a first state value corresponding to the matched maintenance items, and determining a second state value corresponding to the unmatched maintenance items; obtaining a maintenance information chain template, wherein the maintenance information chain template comprises maintenance item identifiers arranged in a first set order; determining state values corresponding to the maintenance item identifiers; writing the state values into the maintenance information chain template; generating a state sequence corresponding to the target elevator; inputting the state sequence into a fitness model and receiving a fitness; and generating a maintenance scheme according to the fitness, so as to realize quantification of the excellent degree of the maintenance scheme, and generate a personalized and excellent maintenance scheme for the target elevator.

Description

Technical Field

[0001] This invention belongs to the technical field of elevator maintenance management, and particularly relates to a method, apparatus, equipment and storage medium for generating maintenance plans. Background Technology

[0002] Currently, elevators are becoming increasingly common. The proper use of elevators can ensure the safety of users and their normal work and life. Therefore, the importance of elevator operation safety is self-evident, and elevator maintenance is the most effective way to ensure elevator operation safety.

[0003] As an independently operating system, the safety of an elevator is affected by various factors, such as its age, environment, and load. Currently, elevator maintenance plans simply follow established rules, applying fixed maintenance cycles and items across the board. This approach is clearly unsuitable for the diverse conditions of different elevators, potentially leading to insufficient maintenance and hidden safety hazards, or excessive maintenance and wasted resources. Summary of the Invention

[0004] This invention provides a method, apparatus, equipment, and storage medium for generating maintenance plans, in order to solve the problem that the current practice of formulating maintenance plans according to the same rules for elevators with different situations leads to over-maintenance or under-maintenance. By formulating maintenance plans according to the specific situation of different elevators, the optimal balance between maintenance costs and maintenance effects can be achieved.

[0005] According to one aspect of the present invention, a method for generating a maintenance plan is provided, the method comprising:

[0006] Obtain target information of the target elevator that has maintenance needs, the target information including the maintenance information of the target elevator;

[0007] Obtain a pre-defined set of maintenance items, which includes multiple maintenance items;

[0008] Based on the maintenance information, matching maintenance items and non-matching maintenance items are determined from the total set of maintenance items, and a first state value corresponding to the matching maintenance item and a second state value corresponding to the non-matching maintenance item are determined.

[0009] Obtain a maintenance information chain template, wherein the maintenance information chain template includes the identifiers of each maintenance item arranged in a first preset order;

[0010] Determine the status value corresponding to each maintenance project identifier, and write the status value into the maintenance information chain template to generate the status sequence corresponding to the target elevator, wherein the status value is the first status value or the second status value;

[0011] The state sequence is input into a pre-trained fitness model, and the fitness output by the fitness model is received.

[0012] Based on the fitness level, a maintenance plan for the target elevator is generated.

[0013] According to one aspect of the present invention, a maintenance plan generation apparatus is provided, the apparatus comprising:

[0014] The target information acquisition module is used to acquire target information of target elevators that have maintenance needs, and the target information includes the maintenance information of the target elevator;

[0015] The maintenance project set acquisition module is used to acquire a pre-set maintenance project set, which includes multiple maintenance projects;

[0016] The matching module is used to determine matching maintenance items and non-matching maintenance items from the total set of maintenance items based on the maintenance information, and to determine a first state value corresponding to the matching maintenance item and a second state value corresponding to the non-matching maintenance item.

[0017] The template acquisition module is used to acquire a maintenance information chain template, wherein the maintenance information chain template includes the identifiers of each maintenance item arranged in a first preset order;

[0018] The state sequence generation module is used to determine the state value corresponding to each maintenance project identifier, write the state value into the maintenance information chain template, and generate the state sequence corresponding to the target elevator, wherein the state value is the first state value or the second state value.

[0019] The fitness receiving module is used to input the state sequence into a pre-trained fitness model and receive the fitness output by the fitness model.

[0020] A generation module is used to generate a maintenance plan for the target elevator based on the fitness level.

[0021] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0022] At least one processor; and

[0023] A memory communicatively connected to the at least one processor; wherein,

[0024] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to execute a maintenance scheme generation method according to any embodiment of the present invention.

[0025] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions, the computer instructions being configured to cause a processor to execute and implement a maintenance scheme generation method as described in any embodiment of the present invention.

[0026] The technical solution of this invention provides a method for generating a maintenance plan. The method includes: acquiring target information of a target elevator with maintenance needs, the target information including maintenance information; acquiring a pre-set set of maintenance items, the set including multiple maintenance items; determining matching and non-matching maintenance items from the set of maintenance items based on the maintenance information, and determining a first state value corresponding to the matching maintenance item and a second state value corresponding to the non-matching maintenance item; acquiring a maintenance information chain template, the maintenance information chain template including maintenance item identifiers arranged in a first pre-set order; determining the state value corresponding to each maintenance item identifier; and writing the state value into the maintenance plan. In the information chain template, a state sequence corresponding to the target elevator is generated. The state value is either the first state value or the second state value. The state values ​​in the obtained state sequence include both state values ​​determined by the individual situation of the target elevator and state values ​​not determined by the individual situation. It can quickly determine the maintenance plan that meets the target elevator from many maintenance projects. The state sequence is input into a pre-trained fitness model and the fitness output of the fitness model is received. Based on the fitness, a maintenance plan for the target elevator is generated, realizing the generation of a personalized maintenance plan for the target elevator. Based on the fitness, the excellence of the maintenance plan can be quantified to determine whether the generated maintenance plan is good enough.

[0027] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used 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 those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 This is a flowchart of a maintenance scheme generation method according to Embodiment 1 of the present invention;

[0030] Figure 2 This is a schematic diagram of a maintenance information chain template provided in Embodiment 1 of the present invention;

[0031] Figure 3 This is a flowchart of a maintenance scheme generation method according to Embodiment 2 of the present invention;

[0032] Figure 4 This is a schematic diagram of a state sequence crossover provided in Embodiment 2 of the present invention;

[0033] Figure 5 This is a schematic diagram of a neural network structure provided in Embodiment 2 of the present invention;

[0034] Figure 6 This is a schematic diagram of a maintenance scheme generation device according to Embodiment 3 of the present invention;

[0035] Figure 7 This is a schematic diagram of the structure of an electronic device that implements a maintenance scheme generation method according to an embodiment of the present invention. Detailed Implementation

[0036] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0037] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0038] Example 1

[0039] Figure 1 This is a flowchart of a maintenance scheme generation method provided in Embodiment 1 of the present invention.

[0040] This method can be executed by a maintenance plan generation device, which can be implemented in hardware and / or software.

[0041] S110, Obtain target information of the target elevator that has maintenance needs, including the maintenance information of the target elevator.

[0042] The target elevator refers to the individual elevator for which a maintenance plan needs to be generated. Maintenance information can include the elevator's maintenance cycle, maintenance requirements, maintenance precautions, maintenance contract details, etc.

[0043] The target information can be obtained through manual input or by pre-establishing a database of elevators within the management scope, pre-entering the target information corresponding to all elevators within the management scope, and then directly retrieving the target information from the database by simply selecting the corresponding elevator.

[0044] S120: Obtain a pre-defined set of maintenance items, which includes multiple maintenance items.

[0045] The pre-defined maintenance item set can be based on all maintenance items that have occurred in the historical maintenance process of all elevators of the same type as the target elevator. For example, the maintenance item set can be formed by the union of all maintenance items recorded in the maintenance contracts for all elevator types, in order to deal with elevators of different types and elevators with different maintenance items recorded in the maintenance contracts.

[0046] S130, based on the maintenance information, determine matching maintenance items and non-matching maintenance items from the overall set of maintenance items, and determine the first state value corresponding to the matching maintenance items and the second state value corresponding to the non-matching maintenance items.

[0047] In the maintenance information, a portion of the maintenance items can be designated as matching maintenance items. Matching maintenance items can be those that need to be completed for the target elevator, or those that do not need to be completed. The maintenance information can directly indicate the corresponding matching maintenance items, or it can be determined through information analysis based on the maintenance information. For example, if the target elevator is only 3 months old, maintenance items related to elevator aging can be selected as matching maintenance items from the overall maintenance item set. From the maintenance items in the overall maintenance item set other than the matching maintenance items, a portion of the maintenance items can be randomly selected as non-matching maintenance items. These non-matching maintenance items can be either required or not required in the generated maintenance plan. In other words, a portion of the maintenance items other than the matching maintenance items can be randomly selected as non-matching maintenance items. These non-matching maintenance items can be assigned a second status value to indicate whether they need to be completed or not.

[0048] The first status value corresponding to a matched maintenance project can be a status value indicating whether the matched maintenance project has been determined to be completed or not, while the status value of a non-matched maintenance project can be a status value indicating whether it has been randomly determined to be completed or not.

[0049] In one embodiment, the first state value and the second state value include 0 and 1. When the state value is 1, it indicates that the corresponding maintenance item needs to be completed. When the state value is 0, it indicates that the corresponding maintenance item does not need to be completed. The maintenance information includes a first maintenance item that needs to be completed and a second maintenance item that does not need to be completed. S130 includes the following steps:

[0050] S130-1, Match the first maintenance project and the second maintenance project from the total maintenance project set, and use them as the matching maintenance projects;

[0051] S130-2, the maintenance items in the total maintenance item set, excluding the matching maintenance items, are identified as non-matching maintenance items;

[0052] S130-3, the first state value of the first maintenance item is determined to be 1, and the first state value of the second maintenance item is determined to be 0;

[0053] S130-4, randomly determine the second state value of the non-matching maintenance item as 0 or 1.

[0054] In the maintenance plan for the target elevator, there are only two possibilities for each maintenance item: to perform maintenance or not to perform maintenance. Therefore, binary encoding can be used to represent the status value of whether each maintenance item needs to be completed or not in the maintenance plan. When the status value is 1, it means that the corresponding maintenance item needs to be completed, and when the status value is 0, it means that the corresponding maintenance item does not need to be completed.

[0055] The maintenance information can include a first maintenance item that needs to be completed and a second maintenance item that does not need to be completed. The first maintenance item and the second maintenance item can be directly determined from the total set of maintenance items as matching maintenance items. At this time, the first status value corresponding to the first maintenance item is 1, and the first status value corresponding to the second maintenance item is 0.

[0056] Non-matching maintenance items can be all maintenance items in the total set of maintenance items except for matching maintenance items. For non-matching maintenance items, in this maintenance plan, they can be maintenance items that need to be completed or maintenance items that do not need to be completed. The second state value corresponding to the non-matching maintenance item can be randomly determined to be 0 or 1.

[0057] S140, Obtain the maintenance information chain template, which includes the identifiers of each maintenance item arranged in a first set order.

[0058] The maintenance information chain template can be formed by arranging the maintenance item identifiers corresponding to all maintenance items in the maintenance item master set in a first predetermined order. After arranging all maintenance item identifiers in the first predetermined order, a chain-like structure can be formed. Since elevators of the same type can correspond to the same maintenance item master set, the maintenance information chain formed by arranging the maintenance item identifiers corresponding to all maintenance items in the maintenance item master set in the first predetermined order can serve as a maintenance information chain template applicable to the corresponding type of elevator.

[0059] S150, determine the status value corresponding to each maintenance project identifier, write the status value into the maintenance information chain template, and generate the status sequence corresponding to the target elevator, with the status value being either the first status value or the second status value.

[0060] After obtaining the maintenance information chain template, the first status value of the matching maintenance project and the second status value of the non-matching maintenance project can be used as the status value corresponding to each maintenance project identifier and written into the maintenance information chain template.

[0061] refer to Figure 2 A schematic diagram of a maintenance information chain template. Figure 2 The AG can be viewed as a list of maintenance item identifiers arranged in a predefined order. It can match the first status value corresponding to a maintenance item and write the second status value corresponding to an unmatched maintenance item into the maintenance information chain template, such as... Figure 2 As shown, the state sequence corresponding to the target elevator is 1011101.

[0062] S160, input the state sequence into the pre-trained fitness model, and receive the fitness output by the fitness model.

[0063] The pre-trained fitness model can be obtained by training a large amount of sample data and the labeled data after processing the sample data on an artificial neural network model. It is used to evaluate whether the combination of executed and unexecuted maintenance items corresponding to the state sequence can achieve the optimal balance between maintenance cost and maintenance effect.

[0064] In the fitness model, maintenance item identifiers corresponding to each point in the state sequence can be pre-defined. This means that even when only the state sequence is obtained, the fitness model can determine the state value corresponding to each maintenance item based on the pre-defined maintenance item identifiers. Compared to directly generating a combination of executed and non-executed maintenance items and then inputting it into the neural network model for analysis, the method described in this invention significantly reduces the computational burden on the fitness model, making its processing faster and smoother.

[0065] Additionally, some elevator attributes of the target elevator can be pre-input into the fitness model, allowing the fitness model to determine the fitness based on these attributes. For example, elevator attributes could include maintenance data from previous maintenance, the application scenario, and the maintenance cycle.

[0066] S170 generates a maintenance plan for the target elevator based on fitness.

[0067] Fitness can be the ratio between the maintenance costs incurred and the maintenance results obtained. It can generate maintenance plans by determining the relationship between maintenance costs and maintenance results, and thus quantify maintenance plans.

[0068] Fitness can be used to determine whether a greater maintenance effect can be achieved at a reasonable maintenance cost under a combination of maintenance items that need to be completed and maintenance items that do not need to be completed.

[0069] When the obtained fitness meets the pre-set maintenance plan conditions, such as when the ratio between maintenance cost and maintenance effect is greater than the set threshold, the maintenance items that need to be completed and those that do not need to be completed corresponding to the state sequence can be combined with the elevator attribute information of the target elevator to generate a maintenance plan for the target elevator.

[0070] If the obtained fitness does not meet the pre-set maintenance plan conditions, S130 can be re-executed to redetermine the second state values ​​corresponding to each non-matching maintenance item identifier. In another implementation, the second state values ​​corresponding to non-matching maintenance item identifiers can be determined from the state sequence that does not meet the pre-set maintenance plan conditions, and some of the second state values ​​can be randomly changed to form a new state sequence. This new state sequence is then input into the fitness model until a state sequence that meets the maintenance plan conditions is obtained, at which point the maintenance plan is generated.

[0071] This invention proposes a method for generating a maintenance plan. The method includes: acquiring target information of a target elevator requiring maintenance, the target information including maintenance information; acquiring a pre-defined set of maintenance items, the set including multiple maintenance items; determining matching and non-matching maintenance items from the set based on the maintenance information, and determining a first state value corresponding to the matching maintenance item and a second state value corresponding to the non-matching maintenance item; acquiring a maintenance information chain template, the template including maintenance item identifiers arranged in a first predetermined order; determining the state value corresponding to each maintenance item identifier; and writing the state value into the maintenance information chain. The template generates a state sequence corresponding to the target elevator, with each state value being either the first or second state value. The state values ​​in the resulting state sequence include both those determined by the individual circumstances of the target elevator and those not determined by the individual circumstances. This allows for the rapid identification of a suitable maintenance plan for the target elevator from numerous maintenance projects. The state sequence is then input into a pre-trained fitness model, and the fitness output from the fitness model is received. Based on the fitness, a maintenance plan for the target elevator is generated, enabling the generation of a personalized maintenance plan for the target elevator. Furthermore, the fitness level can be used to quantify the excellence of the maintenance plan to determine whether the generated maintenance plan is sufficiently good.

[0072] Example 2

[0073] According to a specific embodiment of the present invention, in combination with Figure 3 A flowchart illustrating a method for generating a maintenance plan is provided below, which details a method for generating a maintenance plan according to the present invention.

[0074] This invention provides a method for generating a maintenance plan, comprising the following steps:

[0075] S310: Obtain the target information of the target elevator that has maintenance needs. The target information includes the maintenance information of the target elevator.

[0076] S320: Obtain a pre-defined set of maintenance items, which includes multiple maintenance items.

[0077] S330, based on the maintenance information, determine the matching maintenance items and non-matching maintenance items from the overall maintenance item set, and determine the first state value corresponding to the matching maintenance items and the second state value corresponding to the non-matching maintenance items.

[0078] S340, Obtain the maintenance information chain template, which includes the identifiers of each maintenance item arranged in a first set order.

[0079] Steps S310-S340 can be referred to the description of S110-S140 in Embodiment 1.

[0080] In one embodiment, the target information further includes elevator attributes and a third state value corresponding to the elevator attributes. The maintenance information chain template also includes various elevator attribute identifiers. The maintenance project identifiers and elevator attribute identifiers are arranged in a second predetermined order. The structure of the maintenance information chain template is a chromosome structure. Each maintenance project identifier and each elevator attribute identifier is a gene on the chromosome. The first state value, the second state value, and the third state value are the gene values ​​corresponding to the genes.

[0081] In one implementation, a genetic algorithm can be used to determine the maintenance plan. Genetic algorithms draw upon phenomena from evolutionary biology, including heredity, mutation, natural selection, and hybridization. For an optimization problem, a population of abstract representations (called chromosomes) of a certain number of candidate solutions (individuals) can evolve towards better solutions. Individuals can be represented in binary (i.e., strings of 0s and 1s). This embodiment of the invention can determine the optimal maintenance plan for the target elevator based on the characteristics of the genetic algorithm.

[0082] In a chromosome, each gene location corresponds one-to-one with an input layer neuron in the fitness model. To streamline the processing of state sequences input into the fitness model and avoid the need for frequent additional splicing of elevator attribute-related information each time the fitness model processes a state sequence, elevator attribute identifiers can also be treated as genes and arranged together with maintenance project identifiers in a second predetermined order to obtain a complete chromosome. Elevator attributes can be determined based on the actual situation of the elevator. For example, when the target elevator is a specific elevator, the elevator attributes could be the building's age, the elevator's age, elevator type, elevator speed, etc.

[0083] When the first and second state values ​​are both 0 or 1, which can be represented by specific numbers to form a state sequence, the third state value of the elevator attribute identifier can also be expressed numerically. For example, the building's age, the elevator's age, and the elevator speed can be unitless integer values, while the elevator type can be a type value. Regarding genes on chromosomes, several maintenance cycle genes can be pre-defined, such as 15 days, 30 days, 45 days, and 60 days. The gene values ​​for maintenance cycle genes can be mutually exclusive; that is, only one maintenance cycle gene on the same chromosome is allowed to have a value of 1, and not all values ​​can be 0.

[0084] S350, determine the gene value corresponding to each gene.

[0085] Genes on chromosomes are arranged in a predefined order, and each gene has its own specific meaning. For example, such as... Figure 2In the AG, gene A represents the gantry board fault code detection. When the gene value of gene A is 0, it means that the gantry board fault code detection is not performed, and when it is 1, it means that the gantry board fault code detection is performed. Gene B represents the traction wheel and guide wheel groove wear inspection. When the gene value of gene B is 0, it means that the traction wheel and guide wheel groove wear inspection is not performed, and when it is 1, it means that the traction wheel and guide wheel groove wear inspection is performed, and so on.

[0086] The gene value corresponding to the gene identified by the maintenance project can be determined according to the first state value and the second state value. The gene value corresponding to the gene identified by the elevator attribute can be determined according to the actual situation of the elevator attribute.

[0087] In one implementation, each gene can also have its own gene structure, which can be "gene identifier -- gene -- variable type -- exclusive type -- gene value range". Here, the gene identifier is the specific maintenance project identifier and elevator attribute identifier; the gene value refers to the current value of the gene, which must be within the gene value range. For example, the gene value of the maintenance project identifier must be 0 or 1; the variable type indicates whether the gene value can be changed, with 1 indicating it can be changed and 0 indicating it cannot be changed, meaning the gene value cannot be changed in subsequent crossover and mutation operations; the exclusive type indicates that the gene... The value indicates whether the gene excludes other genes of the same type. A value of 1 indicates that there is exactly one site with a value of 1 for the same type of gene. For example, different maintenance cycles of the same type are exclusive, meaning that there cannot be two or more maintenance cycles in the same maintenance plan. The gene value range limits the range of values ​​for gene expression. For example, elevators have attributes such as elevator speed. The speed of an elevator will change under normal operation over a certain period of time. For example, there is acceleration and deceleration during normal operation. The elevator speed is not a fixed value and there is a normal speed range, which is the range of values ​​for gene expression.

[0088] S360 writes the gene values ​​into the chromosome, generating the gene value sequence of the chromosome, which serves as the state sequence corresponding to the target elevator.

[0089] Since the genes on the chromosome are arranged in the second preset order, after the gene value corresponding to each gene is determined, the gene value is written into the chromosome. At this time, the gene value will also be arranged in the second preset order to form a gene value sequence, which is the state sequence corresponding to the target elevator.

[0090] The gene values ​​of each gene on the chromosome can be determined based on national or industry standards, contract requirements, management requirements, and timeliness requirements. The initial gene value and whether it can be changed can be determined. For variable types, 0 means that it is determined to be an immutable part. If it is a maintenance item that must be done, the variable type is set to 1. If it is a maintenance item that does not need to be done, the variable type is set to 0. For the changeable part (such as the maintenance item that can be selected in this maintenance plan), the variable type is set to 1.

[0091] In one embodiment, there are multiple chromosomes, and S350 includes the following steps:

[0092] The first state value is used as the gene value corresponding to the gene that indicates the matching maintenance project.

[0093] The third state value is used as the gene value corresponding to the gene that indicates the elevator's attributes;

[0094] Determine multiple possible combinations of the second state value corresponding to the non-matching maintenance project;

[0095] The second state value in each possible combination is used as the gene value corresponding to the gene indicating the non-matching maintenance project.

[0096] Since non-matching maintenance items are optional, meaning they can be performed or not, different second-state values ​​for multiple non-matching maintenance items can form multiple possible combinations. A maximum allowed number of possible combinations can be pre-defined. When forming multiple possible combinations based on different second-state values, the total number of possible combinations may not exceed the maximum allowed number; in this case, the multiple possible combinations represent all possible combinations. Alternatively, the total number of possible combinations may exceed the maximum allowed number. In this case, different second-state values ​​can be randomly combined to form the same number of possible combinations as the maximum allowed number. This method can evolve the optimal combination using a limited number of chromosomes, effectively avoiding the conventional exhaustive search approach to determine all possible combinations and significantly improving processing speed.

[0097] For the maintenance plan to be generated this time, based on the randomness of these non-matching maintenance projects, we can determine multiple chromosomes, and then obtain the most suitable chromosome from multiple chromosomes, thereby determining the most suitable maintenance plan.

[0098] In one embodiment, step S360 includes the following steps:

[0099] The second state value in each possible combination is written into different chromosomes to generate gene value sequences of multiple chromosomes. The number of chromosomes is determined by the length of the gene value sequence and is no more than 500.

[0100] The first and third state values ​​are written into all chromosomes to generate gene value sequences for multiple chromosomes.

[0101] Each gene on each chromosome corresponds to an identifier. Multiple chromosomes represent multiple maintenance schemes. In the case of non-matching maintenance projects with multiple possible combinations of different second state groups, chromosomes representing multiple maintenance schemes can be generated.

[0102] For each possible combination of second-state values, they are written into the corresponding non-matching maintenance project identifier in the chromosome, with each possible combination corresponding to one chromosome. Since the first and third-state values ​​are determined in advance and there is no randomness, the genes for matching maintenance project identifiers and elevator attribute identifiers are consistent across each chromosome, meaning the corresponding gene values ​​are the same. It is important to note that for chromosomes with the same maintenance plan, the order of each gene is fixed and identical; only the gene values ​​in the gene structure of some genes within the chromosome differ.

[0103] To ensure processing speed and reduce the amount of computation in each round, the optimal result can be continuously approximated through iteration. The total number of chromosomes in each round can be limited. Typically, the number of chromosomes is 5 times the length of the gene sequence, and the maximum allowed number is no more than 500.

[0104] S370 inputs the state sequence into the pre-trained fitness model and receives the fitness output by the fitness model.

[0105] Since information related to elevator attributes also exists as genes in chromosomes, the fitness model only needs to pre-set the points according to the second preset order to determine the information represented by each value in the state sequence.

[0106] S380, based on fitness, determines candidate state sequences from multiple state sequences.

[0107] S390, randomly select multiple state sequences from the candidate state sequences according to the set crossover probability, pair them up in pairs, and randomly determine the split point of the paired state sequences.

[0108] S3010: Split the paired state sequences at the split points and cross them to generate two new state sequences.

[0109] S3011, the new state sequence and the candidate state sequences that have not been selected for pairing are used as candidate variant state sequences to form a candidate variant state sequence set.

[0110] S3012, randomly select multiple state sequences from the candidate mutated state sequence set according to a pre-set mutation probability and perform mutation operations to generate multiple mutated state sequences.

[0111] S3013, based on the mutated state sequence and the candidate mutated state sequences in the candidate mutated state sequence set that have not been selected for mutation operation, continue to input the state sequence into the pre-trained fitness model and receive the fitness output by the fitness model until the first preset condition is met.

[0112] Since the first fitness received from the fitness model may not be able to generate a suitable maintenance plan, in steps S380-S3013, the multiple state sequences generated can be cross-combined and mutated to obtain a richer set of state sequences that meet the basic requirements of the target elevator maintenance plan. Then, a better maintenance plan can be determined from these more state sequences.

[0113] All chromosomes corresponding to each state sequence before pairing can be considered as forming the initial maintenance scheme gene population. The fitness of each chromosome's corresponding state sequence can be obtained from the fitness. The higher the fitness, the better the chromosome corresponding to that state sequence, that is, the better the maintenance scheme corresponding to that chromosome.

[0114] After evaluating the fitness of the state sequences corresponding to each chromosome in the initial maintenance scheme gene population using a fitness model, before pairing, each chromosome in the gene population can be selected, and then pairing, splitting, and crossing over can be performed on the state sequences corresponding to the selected chromosomes, followed by random mutation.

[0115] When determining candidate state sequences from multiple state sequences, a roulette wheel selection algorithm can be used. Roulette wheel selection is a probabilistic selection algorithm; the higher the selection probability, the more likely the chromosome will be selected. The selection probability can be determined based on fitness. To determine the probability for each chromosome, the fitness of all state sequences is summed to obtain the total fitness. The fitness of each chromosome's corresponding state sequence divided by the total fitness gives the selection probability of that chromosome. Therefore, the higher the fitness of a state sequence, the higher the probability of its corresponding chromosome being selected. Unselected chromosomes can be eliminated, and the selected chromosomes are retained, thus determining the candidate state sequences. These are then used in steps S380-S3013 for pairing, splitting, crossover, and mutation to generate a richer set of state sequences.

[0116] refer to Figure 4This diagram illustrates a state sequence crossover. The splitting points of paired state sequences from different groups can be randomly determined, while the splitting points of paired state sequences within the same group must be at the same gene locus to ensure that structurally consistent state sequences can be reassembled after crossover. The crossover probability can be a pre-set probability.

[0117] like Figure 4 As shown, the split point of the paired two state sequences is after the third gene. After the split, each state sequence will divide into a first half and a second half, at which point crossing over occurs. (See reference...) Figure 4 The two newly generated state sequences are spliced ​​together, which are the first half and the second half of the original different state sequences.

[0118] For each round of candidate state sequences, the state sequence with the highest fitness can be automatically assigned to the next round of candidate state sequences, until a state sequence with higher fitness appears in the candidate state sequences, at which point it can be replaced.

[0119] After the crossover operation, the genetic population of the maintenance scheme stores chromosomes corresponding to the crossover state sequence. These chromosomes can be mutated to make the genetic population more diverse and prevent premature maturation.

[0120] In the candidate state sequence, after some state sequences undergo crossover, the new state sequence generated by the crossover and the candidate state sequences that were not selected for pairing together form a candidate mutated state sequence set. Alternatively, multiple state sequences can be randomly selected from the candidate mutated state sequence set according to a certain mutation probability for mutation operation. The mutation probability can be a pre-set probability.

[0121] For example, in a mutation operation, a chromosome can be randomly selected, and then the gene value of that chromosome can be randomly selected for inversion, i.e., 0 becomes 1, and 1 becomes 0. During the mutation operation, when randomly selecting genes, only genes with a variable type of 1 can be selected for gene value inversion.

[0122] After obtaining the state sequence after the mutation operation, along with the new state sequence that was not selected for mutation, we can continue to execute the steps of inputting the state sequence into the pre-trained fitness model and receiving the fitness output by the fitness model, so as to perform multiple eliminations and updates of the state sequence and determine the fitness of the newly obtained state sequence in each round.

[0123] In one embodiment, the first preset condition is: the fitness corresponding to the state sequence satisfies the second preset condition, or the step of inputting the state sequence into the pre-trained fitness model and receiving the fitness output by the fitness model reaches a specified number of times.

[0124] After each iteration of the fitness model algorithm, it is necessary to determine whether the chromosome meets the requirements to exit the process of inputting the state sequence into the pre-trained fitness model and receiving the fitness output from the fitness model. The first preset condition for determining whether to continue execution can be reaching the maximum number of algorithm iterations. One algorithm iteration can refer to completing the process of selecting chromosomes, crossover, mutation, determining fitness, and determining the optimal state based on fitness. Alternatively, the condition can be the number of times the step of inputting the state sequence into the pre-trained fitness model and receiving the fitness output from the fitness model has been executed. Another condition is that among the obtained fitness values, there exists a state sequence with a fitness value greater than a fitness threshold. If either of these conditions is met, the process of inputting the state sequence into the pre-trained fitness model and receiving the fitness output from the fitness model can be exited, and then the maintenance plan for the target elevator can be generated based on the fitness value.

[0125] In addition, if the conditions for ending the execution, inputting the state sequence into the pre-trained fitness model, and receiving the fitness output by the fitness model have not yet been met, this step needs to be continued to perform algorithm iteration. In order to speed up the convergence of the genetic algorithm, the best gene population after each fitness is confirmed can be selected. This can be done by calculating the average fitness of all state sequences, taking out the gene sequence with the highest fitness, and then replacing the state sequence with a fitness lower than the average fitness with the state sequence with the highest fitness, and then proceeding to the next round of algorithm iteration.

[0126] S3014, based on fitness, generates a maintenance plan for the target elevator.

[0127] The suitability of an elevator needs to be measured by a definite standard. For elevator maintenance, the purpose of maintenance is to reduce the elevator failure rate and ensure the normal operation of the elevator. However, maintenance has costs, and different maintenance items and maintenance cycles will result in different cost inputs. Therefore, the suitability can reflect the result of a comprehensive consideration of cost and failure rate.

[0128] The system can select the state sequence with the highest fitness from multiple state sequences, and generate the optimal maintenance plan for the target elevator based on the information expressed by the chromosome corresponding to the state sequence, such as maintenance cycle, maintenance items to be completed, and maintenance items not to be completed.

[0129] In one embodiment, the fitness model is generated as follows:

[0130] Multiple sample state sequences are obtained, and the historical maintenance data corresponding to each sample state sequence is obtained. The historical maintenance data includes the elevator downtime, maintenance interval, project cost and fixed cost of each maintenance item to be completed for the sample elevator corresponding to the sample state sequence.

[0131] The ratio of elevator downtime due to malfunction to maintenance interval is defined as the failure rate.

[0132] The sum of project costs and fixed costs is determined as input costs;

[0133] The difference between 1 and the failure rate is used as the maintenance effectiveness.

[0134] The ratio of maintenance effectiveness to input cost is used as the sample fitness, and the sample fitness is used as the label value of the sample state sequence.

[0135] An fitness model is generated based on the sample state sequence and the corresponding label value.

[0136] Normally, fitness is calculated directly by substituting the gene values ​​of chromosomes into a formula. However, for elevators, there is no explicit formula to express the relationship between maintenance plans and failure rates, making it impossible to directly calculate the fitness of a maintenance plan using a formula. Therefore, a multi-layer artificial neural network can be used to calculate fitness. The input layer consists of gene values, the middle layers are hidden layers, and the output layer is the fitness. (See reference...) Figure 5 A schematic diagram of a neural network structure.

[0137] The Sigmoid function can be used in the fitness model, the loss function for training the fitness model can be MSELoss, and the SGD optimizer can be used.

[0138] Training the fitness model requires samples, which can be generated from historical maintenance data along with their labels. The labels on the sample state sequences and their order must be consistent with those used in the actual application of the fitness model.

[0139] Each sample state sequence corresponds to a label value, which is the fitness of that sample state sequence. This quantifies the excellence of the maintenance plan corresponding to the sample state sequence. Specifically, when calculating sample fitness, the ratio of maintenance effectiveness to maintenance cost can be used as the sample fitness. The formula for calculating sample fitness is as follows:

[0140]

[0141] Where F refers to fitness, f(e) refers to maintenance effect, f(c) refers to input cost, Ee refers to the failure rate of the sample elevator after maintenance, the failure rate is expressed as the ratio between the downtime of non-human-caused failures and the total time between two maintenance, that is, the maintenance interval time, Ts is the downtime of non-human-caused failures, Tt is the maintenance interval time, Ct is the input cost, Cn is the project cost corresponding to each maintenance item to be completed, N is the total number of maintenance items to be completed, M is the number of times the elevator stops due to non-human causes within the maintenance interval time, and C is the fixed cost.

[0142] After generating a large number of sample state sequences and label values, the fitness model can be trained. The specific training process is a common method for training models and will not be discussed here.

[0143] Once the model is trained, it can be used to evaluate the fitness of the state sequence corresponding to each chromosome in the maintenance scheme gene population, and obtain the fitness of the chromosome.

[0144] The training of the fitness model is an independent process and is not performed during the input of non-sample state sequences. For the same type of elevator, the fitness model does not need to be trained every time a maintenance plan is generated after training is completed. It can be reused. If there are more and newer samples, it can be retrained to optimize the model.

[0145] This invention proposes a method for generating maintenance plans. It uses a genetic algorithm to solve the multivariate optimization problem and quantifies the excellence of the maintenance plan. It generates an optimal maintenance plan that is personalized for the target elevator and maps the maintenance items and elevator attributes in the maintenance plan to the maintenance effect and maintenance cost, so that the generated optimal maintenance plan achieves the best balance between cost and efficiency.

[0146] Example 3

[0147] Figure 6 This is a schematic diagram of a maintenance scheme generation device provided in Embodiment 3 of the present invention, as shown below. Figure 6 As shown, the device includes:

[0148] The target information acquisition module 610 is used to acquire target information of target elevators that have maintenance needs, and the target information includes the maintenance information of the target elevators;

[0149] The maintenance project set acquisition module 620 is used to acquire a pre-set maintenance project set, which includes multiple maintenance projects;

[0150] The matching module 630 is used to determine matching maintenance items and non-matching maintenance items from the total set of maintenance items based on the maintenance information, and to determine a first state value corresponding to the matching maintenance item and a second state value corresponding to the non-matching maintenance item.

[0151] The template acquisition module 640 is used to acquire a maintenance information chain template, wherein the maintenance information chain template includes the identifiers of each maintenance item arranged in a first preset order;

[0152] The state sequence generation module 650 is used to determine the state value corresponding to each maintenance project identifier, write the state value into the maintenance information chain template, and generate the state sequence corresponding to the target elevator, wherein the state value is the first state value or the second state value.

[0153] The fitness receiving module 660 is used to input the state sequence into a pre-trained fitness model and receive the fitness output by the fitness model.

[0154] The generation module 670 is used to generate a maintenance plan for the target elevator based on the fitness level.

[0155] In one embodiment, the first state value and the second state value include 0 and 1. When the state value is 1, it indicates that the corresponding maintenance project needs to be completed. When the state value is 0, it indicates that the corresponding maintenance project does not need to be completed. The maintenance information includes a first maintenance project that needs to be completed and a second maintenance project that does not need to be completed.

[0156] The matching module 630 includes the following sub-modules:

[0157] The matching submodule is used to match the first maintenance project and the second maintenance project from the total set of maintenance projects, and use them as the matched maintenance projects;

[0158] The non-matching maintenance project determination submodule is used to determine maintenance projects other than the matching maintenance projects in the total set of maintenance projects as non-matching maintenance projects;

[0159] The first determining submodule is used to determine the first state value of the first maintenance item as 1 and the first state value of the second maintenance item as 0.

[0160] The second determining submodule is used to randomly determine the second state value corresponding to the non-matching maintenance item as 0 or 1.

[0161] In one embodiment, the target information further includes one or more elevator attributes and a third state value corresponding to each elevator attribute. The maintenance information chain template further includes elevator attribute identifiers. The maintenance project identifiers and elevator attribute identifiers are arranged in a second predetermined order. The structure of the maintenance information chain template is a chromosome structure. Each maintenance project identifier and each elevator attribute identifier is a gene on a chromosome. The first state value, the second state value, and the third state value are gene values ​​corresponding to the genes.

[0162] The state sequence generation module 650 includes the following sub-modules:

[0163] The gene value determination submodule is used to determine the gene value corresponding to each of the genes;

[0164] The state sequence determination submodule is used to write the gene values ​​into the chromosome to generate the gene value sequence of the chromosome, which serves as the state sequence corresponding to the target elevator.

[0165] In one embodiment, there are multiple chromosomes, and the gene value determination submodule includes the following units:

[0166] The first gene value determination unit is used to use the first state value as the gene value corresponding to the gene indicating the matching maintenance project.

[0167] The second gene value determination unit is used to use the third state value as the gene value corresponding to the gene indicating the elevator attribute;

[0168] A probability combination determination unit is used to determine multiple probability combinations of the second state value corresponding to the non-matching maintenance item;

[0169] The third gene value determination unit is used to take the second state value in each of the possible combinations as the gene value corresponding to the gene indicating the non-matching maintenance project.

[0170] The state sequence determination submodule is specifically used for:

[0171] The second state value in each of the possible combinations is written into different chromosomes to generate one or more gene value sequences of the chromosomes. The number of chromosomes is determined by the length of the gene value sequence and is not greater than 500.

[0172] The first state value and the third state value are written into all the chromosomes to generate gene value sequences of one or more chromosomes.

[0173] In one embodiment, the device is further configured to:

[0174] Based on the fitness, candidate state sequences are determined from the plurality of state sequences;

[0175] In the candidate state sequences, multiple state sequences are randomly selected according to a pre-set crossover probability and paired up in pairs, and the split point of the paired state sequences is randomly determined.

[0176] The paired state sequences are split at the split points and then cross-interchanged to generate two new state sequences.

[0177] The new state sequence and the candidate state sequences that were not selected for pairing are used as candidate variant state sequences to form a candidate variant state sequence set.

[0178] Multiple candidate mutant state sequences are randomly selected from the candidate mutant state sequence set according to a preset mutation probability and subjected to mutation operation to generate multiple mutant state sequences;

[0179] Based on the mutated state sequence and the candidate mutated state sequences in the candidate mutated state sequence set that have not been selected for mutation operation, the process of inputting the state sequence into the pre-trained fitness model and receiving the fitness output by the fitness model continues until the first preset condition is met.

[0180] In one embodiment, the first preset condition is: the fitness corresponding to the state sequence satisfies the second preset condition, or the step of inputting the state sequence into a pre-trained fitness model and receiving the fitness output by the fitness model reaches a specified number of times.

[0181] In one embodiment, the fitness model is characterized by being generated in the following manner:

[0182] Acquire historical maintenance data corresponding to multiple sample state sequences. The historical maintenance data includes the elevator downtime due to failure, maintenance interval, project cost and fixed cost of each maintenance item to be completed for the sample elevator corresponding to the sample state sequence.

[0183] The ratio of the elevator downtime due to malfunction to the maintenance interval is defined as the failure rate.

[0184] The sum of the project cost and the fixed cost is determined as the input cost;

[0185] The difference between 1 and the failure rate is determined as the maintenance effect;

[0186] The ratio of the maintenance effect to the input cost is used as the sample fitness, and the sample fitness is used as the label value of the sample state sequence.

[0187] The fitness model is generated based on the sample state sequence and the label value corresponding to the sample state sequence.

[0188] The maintenance scheme generation device provided in this embodiment of the invention can realize the maintenance scheme generation method provided in Embodiment 1 and Embodiment 2 of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0189] Example 4

[0190] Figure 7 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0191] like Figure 7 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0192] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0193] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as a method for generating a maintenance plan.

[0194] In some embodiments, a method for generating a maintenance plan can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the method for generating a maintenance plan described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform a method for generating a maintenance plan by any other suitable means (e.g., by means of firmware).

[0195] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0196] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0197] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0198] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0199] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0200] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0201] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0202] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for generating a maintenance plan, characterized in that, The method includes: Obtain target information of the target elevator that has maintenance needs, the target information including the maintenance information of the target elevator; Obtain a pre-defined set of maintenance items, which includes multiple maintenance items; Based on the maintenance information, matching maintenance items and non-matching maintenance items are determined from the total set of maintenance items. A first status value corresponding to the matching maintenance item and a second status value corresponding to the non-matching maintenance item are determined. The non-matching maintenance item is a maintenance item that needs to be completed or does not need to be completed in the maintenance plan generated this time. Obtain a maintenance information chain template, wherein the maintenance information chain template includes the identifiers of each maintenance item arranged in a first preset order; Determine the status value corresponding to each maintenance project identifier, and write the status value into the maintenance information chain template to generate the status sequence corresponding to the target elevator, wherein the status value is the first status value or the second status value; The state sequence is input into a pre-trained fitness model, and the fitness output by the fitness model is received. Based on the fitness, candidate state sequences are determined from the plurality of state sequences; In the candidate state sequences, multiple state sequences are randomly selected according to a pre-set crossover probability and paired up in pairs, and the split point of the paired state sequences is randomly determined. The paired state sequences are split at the split points and then cross-interchanged to generate two new state sequences. The new state sequence and the candidate state sequences that were not selected for pairing are used as candidate variant state sequences to form a candidate variant state sequence set. Multiple candidate mutant state sequences are randomly selected from the candidate mutant state sequence set according to a preset mutation probability and subjected to mutation operation to generate multiple mutant state sequences; Based on the mutated state sequence and the candidate mutated state sequences in the candidate mutated state sequence set that have not been selected for mutation operation, continue to execute the step of inputting the state sequence into the pre-trained fitness model and receiving the fitness output by the fitness model until the first preset condition is met. The first preset condition is: the fitness corresponding to the state sequence satisfies the second preset condition, or the step of inputting the state sequence into the pre-trained fitness model and receiving the fitness output by the fitness model reaches a specified number of times; Based on the fitness level, a maintenance plan for the target elevator is generated.

2. The method according to claim 1, characterized in that, The first state value and the second state value include 0 and 1. When the state value is 1, it means that the corresponding maintenance project needs to be completed. When the state value is 0, it means that the corresponding maintenance project does not need to be completed. The maintenance information includes a first maintenance project that needs to be completed and a second maintenance project that does not need to be completed. The step of determining matching and non-matching maintenance items from the total set of maintenance items based on the maintenance information, and determining a first status value corresponding to the matching maintenance item and a second status value corresponding to the non-matching maintenance item, includes: The first maintenance project and the second maintenance project are matched from the total set of maintenance projects and used as the matched maintenance projects; The maintenance items in the total maintenance item set, excluding the matching maintenance items, are identified as non-matching maintenance items. The first status value of the first maintenance item is set to 1, and the first status value of the second maintenance item is set to 0. The second state value corresponding to the non-matching maintenance item is randomly determined to be 0 or 1.

3. The method according to claim 1 or 2, characterized in that, The target information also includes one or more elevator attributes, and a third state value corresponding to each elevator attribute. The maintenance information chain template also includes elevator attribute identifiers. The maintenance project identifiers and elevator attribute identifiers are arranged in a second preset order. The structure of the maintenance information chain template is a chromosome structure. Each maintenance project identifier and each elevator attribute identifier is a gene on the chromosome. The first state value, the second state value, and the third state value are gene values ​​corresponding to the genes. The step of determining the status value corresponding to each maintenance item identifier and writing the status value into the maintenance information chain template to generate the status sequence corresponding to the target elevator includes: Determine the gene value corresponding to each of the aforementioned genes; The gene values ​​are written into the chromosome to generate the gene value sequence of the chromosome, which serves as the state sequence corresponding to the target elevator.

4. The method according to claim 3, characterized in that, The chromosome has multiple lines, and determining the gene value corresponding to each gene includes: The first status value is used as the gene value corresponding to the gene indicating the matching maintenance project; The third state value is used as the gene value corresponding to the gene that indicates the elevator attribute; Determine multiple possible combinations of the second state value corresponding to the non-matching maintenance item; The second state value in each of the possible combinations is used as the gene value corresponding to the gene indicating the non-matching maintenance project. The step of writing the gene value into the chromosome to generate the gene value sequence of the chromosome includes: The second state value in each of the possible combinations is written into different chromosomes to generate gene value sequences of multiple chromosomes. The number of chromosomes is determined by the length of the gene value sequence and is not greater than 500. The first state value and the third state value are written into all the chromosomes to generate gene value sequences of multiple chromosomes.

5. The method according to claim 1, 2, or 4, characterized in that, The fitness model is generated in the following manner: Acquire historical maintenance data corresponding to multiple sample state sequences. The historical maintenance data includes the elevator downtime due to failure, maintenance interval, project cost and fixed cost of each maintenance item to be completed for the sample elevator corresponding to the sample state sequence. The ratio of the elevator downtime due to malfunction to the maintenance interval is defined as the failure rate. The sum of the project cost and the fixed cost is determined as the input cost; The difference between 1 and the failure rate is determined as the maintenance effect; The ratio of the maintenance effect to the input cost is used as the sample fitness, and the sample fitness is used as the label value of the sample state sequence. The fitness model is generated based on the sample state sequence and the label value corresponding to the sample state sequence.

6. A maintenance plan generation device, characterized in that, The device includes: The target information acquisition module is used to acquire target information of target elevators that have maintenance needs, and the target information includes the maintenance information of the target elevator; The maintenance project set acquisition module is used to acquire a pre-set maintenance project set, which includes multiple maintenance projects; The matching module is used to determine matching maintenance items and non-matching maintenance items from the total set of maintenance items based on the maintenance information, and to determine a first state value corresponding to the matching maintenance item and a second state value corresponding to the non-matching maintenance item; wherein, the non-matching maintenance item is a maintenance item that needs to be completed or does not need to be completed in the maintenance plan generated this time; The template acquisition module is used to acquire a maintenance information chain template, wherein the maintenance information chain template includes the identifiers of each maintenance item arranged in a first preset order; The state sequence generation module is used to determine the state value corresponding to each maintenance project identifier, write the state value into the maintenance information chain template, and generate the state sequence corresponding to the target elevator, wherein the state value is the first state value or the second state value. The fitness receiving module is used to input the state sequence into a pre-trained fitness model and receive the fitness output by the fitness model. Based on the fitness, candidate state sequences are determined from the plurality of state sequences; In the candidate state sequences, multiple state sequences are randomly selected according to a pre-set crossover probability and paired up in pairs, and the split point of the paired state sequences is randomly determined. The paired state sequences are split at the split points and then cross-interchanged to generate two new state sequences. The new state sequence and the candidate state sequences that were not selected for pairing are used as candidate variant state sequences to form a candidate variant state sequence set. Multiple candidate mutant state sequences are randomly selected from the candidate mutant state sequence set according to a preset mutation probability and subjected to mutation operation to generate multiple mutant state sequences; Based on the mutated state sequence and the candidate mutated state sequences in the candidate mutated state sequence set that have not been selected for mutation operation, continue to execute the step of inputting the state sequence into the pre-trained fitness model and receiving the fitness output by the fitness model until the first preset condition is met. The first preset condition is: the fitness corresponding to the state sequence satisfies the second preset condition, or the step of inputting the state sequence into the pre-trained fitness model and receiving the fitness output by the fitness model reaches a specified number of times; A generation module is used to generate a maintenance plan for the target elevator based on the fitness level.

7. An electronic device, characterized in that, The electronic device includes: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform a method for generating a maintenance scheme as described in claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that are used to cause a processor to execute a method for generating a maintenance scheme as described in claims 1-5.

Images