Intelligent maintenance dispatching method and system for industrial production equipment

Abstract

The application relates to an intelligent maintenance order dispatching method and system for an industrial production device, and belongs to the technical field of intelligent order dispatching. A to-be-assigned repair work order is acquired, and the to-be-assigned repair work order at least includes the following information: a fault device, a fault position and a fault phenomenon; based on a pre-constructed knowledge graph, a plurality of suspected fault causes are determined according to the fault device, the fault position and the fault phenomenon; a work order associated with the plurality of suspected fault causes is selected from a pre-established historical repair work order set, and an associated work order set is constructed; the number of repair workers in the repair worker set and the total number of repair workers pre-acquired are used to determine the fault repair difficulty; and the corresponding order dispatching rule is used to determine the assigned repair worker of the to-be-assigned repair work order according to the fault repair difficulty. The application determines the fault grade of the current repair work order by the number of repair workers and the total number of repair workers, and determines the corresponding order dispatching rule according to the fault grade, so that differentiated order dispatching is realized.

Description

TECHNICAL FIELD

[0001] The application relates to an intelligent maintenance dispatching method and system for industrial production equipment, and belongs to the technical field of intelligent dispatching. BACKGROUND

[0002] In an industrial production environment, maintenance management is crucial for the continuous and efficient operation of production equipment, occupies a core and high-frequency use position, is not only a solid backing for guaranteeing production efficiency and quality, but also a key driving force for continuously optimizing equipment performance and effectively controlling costs. When equipment fails, an efficient maintenance management system can quickly mobilize resources, including professional maintenance teams, critical spare parts, and technical support, to quickly locate and repair faults, minimize downtime, and reduce direct economic losses caused by production interruptions.

[0003] In a traditional equipment management system, the allocation of repair work orders often highly depends on manual operation, especially the subjective judgment of the maintenance team leader, who mainly allocates new tasks based on the number of work orders in hand. This allocation mechanism is limited by individual subjective judgment and experience accumulation, and is easily affected by the team leader's immediate work (such as direct participation in maintenance), which may cause delays in repair work order processing and reduce overall maintenance efficiency.

[0004] Chinese authorized invention patent CN202011227098.6, a comprehensive dispatching method and system for distribution network maintenance work orders, discloses a comprehensive dispatching method for distribution network maintenance work orders, which obtains work order dispatching factors, classifies and calculates the weight of each factor for maintenance workers, and then obtains the comprehensive weight of each maintenance worker. The maximum weight value is selected as the optimal solution for dispatching, achieving accurate dispatching. However, this invention relies on a single non-differentiated dispatching rule, ignores the consideration of fault difficulty level, limits the flexibility and efficiency of the system, and lacks applicability in the field of industrial production equipment maintenance. SUMMARY

[0005] The purpose of the present application is to provide an intelligent maintenance dispatching method and system for industrial production equipment, which determines the fault level of the current repair work order by comparing the number of maintenance workers related to the current repair work order and the total number of maintenance workers, and determines the corresponding dispatching rule according to the fault level, to achieve the purpose of differentiated dispatching with better applicability.

[0006] To achieve the above-mentioned purpose, the present application adopts the following technical solutions.

[0007] On the one hand, the present application provides an intelligent maintenance dispatching method for industrial production equipment, which comprises:

[0008] obtaining a to-be-assigned repair work order, the to-be-assigned repair work order comprising at least the following information: fault equipment, fault position, and fault phenomenon;

[0009] Based on the pre-constructed knowledge graph, according to the fault equipment, the fault position and the fault phenomenon, a plurality of suspected fault causes of the to-be-assigned repair work order are determined;

[0010] All work orders associated with the suspected fault causes are selected from a pre-established historical repair work order set, and an associated work order set is constructed;

[0011] The maintenance workers corresponding to each work order in the associated work order set are obtained, and a maintenance worker set is constructed using all the obtained maintenance workers;

[0012] According to the number of maintenance workers in the maintenance worker set and the pre-obtained total number of maintenance workers, the difficulty of fault maintenance is determined;

[0013] According to the fault maintenance difficulty, a dispatching method is determined, and the assigned maintenance worker of the to-be-assigned repair work order is determined according to the dispatching method.

[0014] Optionally, the determination of the difficulty of fault maintenance according to the number of maintenance workers in the maintenance worker set and the pre-obtained total number of maintenance workers comprises:

[0015] According to the number of maintenance workers and the pre-obtained total number of maintenance workers, a fault maintenance coping rate is determined;

[0016] The calculation formula of the fault maintenance coping rate is: ; In the formula, is the fault maintenance coping rate; is the total number of maintenance workers, and m1 is the number of maintenance workers in the maintenance worker set;

[0017] If the fault maintenance coping rate belongs to a pre-set first range, the difficulty of fault maintenance is determined to be a first level;

[0018] If the fault maintenance coping rate belongs to a pre-set second range, the difficulty of fault maintenance is determined to be a second level;

[0019] If the fault maintenance coping rate belongs to a pre-set third range, the difficulty of fault maintenance is determined to be a third level;

[0020] The first range, the second range and the third range decrease in turn.

[0021] Optionally, when the difficulty of fault maintenance is the first level, the determination of the assigned maintenance worker of the to-be-assigned repair work order according to the dispatching method comprises:

[0022] The backlog work order duration of each maintenance worker in the maintenance worker set is obtained, the maintenance workers are arranged in descending order according to the backlog work order duration, and the first dispatchable maintenance worker in the sequence is determined as the assigned maintenance worker.

[0023] Optionally, when the fault repair difficulty is level two, the assigned repairman for the work order to be assigned, determined according to the dispatching method, includes:

[0024] Each associated work order in the associated work order set is evaluated sequentially as a work order to be evaluated.

[0025] Obtain the evaluation indicators for the work order to be evaluated, including positive and negative indicators;

[0026] The positive indicators include: the probability of the suspected fault cause corresponding to the work order occurring in the work orders to be assigned for repair, the recall weight of the faulty equipment corresponding to the work order, the first-time repair pass rate of the work order obtained in advance, and the comprehensive repair evaluation score.

[0027] The negative indicators include: the backlog of work orders for the corresponding maintenance worker and the maintenance time for the work order, which are obtained in advance;

[0028] Each work order in the associated work order set is evaluated according to the evaluation indicators to obtain the comprehensive evaluation parameters of each work order;

[0029] Based on the comprehensive evaluation parameters, the assigned maintenance worker is determined.

[0030] Optionally, the step of evaluating each associated work order in the associated work order set according to the evaluation index to obtain comprehensive evaluation parameters for each associated work order includes:

[0031] Step 1: Construct the original indicator data matrix using the number of rows and columns of associated work orders and evaluation indicators, respectively:

[0032] ;in, Represents the original indicator data matrix. Represents the first element in the original indicator data matrix. The first related work order The numerical values ​​of the evaluation indicators;

[0033] Step 2: Normalize the positive and negative indicators:

[0034] For positive indicators:

[0035] For negative indicators:

[0036] ;in For the first The numerical values ​​of the evaluation indicators For the first The maximum value of each evaluation indicator the first the minimum value of the first

[0037] Step 3, calculating the corresponding proportion of each item in the original index data matrix:

[0038] ; wherein, p ij the proportion of the first the first the total number of work orders in the associated work order set;

[0039] Step 4, establishing a proportion matrix corresponding to the original index data matrix:

[0040] ;

[0041] Step 5, calculating the entropy value of the first the entropy value of the first

[0042] ; wherein, e j the entropy value of the first

[0043] Step 6, calculating the weight coefficient of the first the weight coefficient of each evaluation index, and determining a weight coefficient matrix used for interaction with the original index data matrix according to the weight coefficient of each evaluation index;

[0044] ; wherein, w j the weight coefficient of the first

[0045] Step 7, the original index data matrix and the weight coefficient matrix are multiplied to obtain a weighted matrix, and the calculation formula of the weighted value of the first

[0046] ; wherein, the weighted value of the first the first

[0047] Step 8, calculating the comprehensive evaluation parameter of the first

[0048] ; wherein, the comprehensive evaluation parameter of the first and ​​​​​​respectively represent the distance from the comprehensive evaluation value of the first respectively represent the weighted maximum value and the weighted minimum value of the first

[0049] The method comprises the following steps:

[0050] According to the comprehensive evaluation parameter, the repairman corresponding to each associated work order is arranged in descending order, and the first dispatchable repairman in the sequence is determined as the assigned repairman.

[0051] Optionally, when the fault maintenance difficulty is the third level, the assigned repairman of the to-be-assigned repair order is determined according to the order assigning method, which comprises the following steps:

[0052] An evaluation parameter of each associated work order in the associated work order set is obtained,

[0053] An evaluation value of each associated work order is determined according to the evaluation parameter.

[0054] According to the evaluation value, the repairman corresponding to each associated work order is arranged in descending order, and the first dispatchable repairman in the sequence is determined as the assigned repairman.

[0055] The evaluation parameter comprises: a recall weight of the fault equipment corresponding to the work order and a possible occurrence probability of the suspected fault cause corresponding to the work order in the to-be-assigned repair order.

[0056] The product of the possible occurrence probability and the recall weight is calculated as the evaluation value.

[0057] Optionally,

[0058] The method for determining the possible occurrence probability of the suspected fault cause corresponding to the work order in the to-be-assigned repair order comprises:

[0059] Based on a pre-constructed knowledge graph, the occurrence probability of each suspected fault cause is inferred according to the positioning nodes of the fault equipment, the fault position and the fault phenomenon in the to-be-assigned repair order in the knowledge graph.

[0060] If the work order corresponds to one suspected fault cause, the occurrence probability of the suspected fault cause is determined as the possible occurrence probability.

[0061] If the work order corresponds to multiple suspected fault causes, one with the highest occurrence probability among the multiple suspected fault causes is determined as the possible occurrence probability.

[0062] The method for determining the recall weight of the fault equipment corresponding to the work order comprises: ​​​

[0063] determine the recall weight according to a similarity degree between the fault device corresponding to the work order and the fault device of the to-be-assigned repair work order; the higher the similarity degree, the higher the recall weight.

[0064] Optionally, if the to-be-assigned repair work order does not include a fault position; the method for determining the fault position comprises:

[0065] obtaining a fault picture of the to-be-assigned repair work order;

[0066] comparing the fault picture with each picture in a set of historical fault pictures obtained in advance respectively to obtain a picture similarity between the fault picture and each picture;

[0067] filtering out historical pictures with a picture similarity exceeding a threshold value and a corresponding fault position being three or fewer to construct a comparison picture set;

[0068] obtaining fault positions corresponding to each picture in the comparison picture set, and taking all obtained fault positions as possible fault positions;

[0069] determining a similarity comprehensive score of each possible fault position according to the comparison picture set and the picture similarity;

[0070] determining a possible fault position with the highest similarity comprehensive score as the fault position of the to-be-assigned repair work order.

[0071] Optionally, each possible fault position is sequentially taken as a to-be-scored fault position for scoring;

[0072] determining a similarity comprehensive score of each to-be-scored fault position according to the comparison picture set and the picture similarity comprises:

[0073] Step 1, dividing the comparison picture set into a first set, a second set and a third set according to a relationship between a fault position corresponding to a historical picture and a to-be-scored fault position;

[0074] if one historical picture corresponds to two fault positions, one of which is a to-be-scored fault position and the other has a constituting relationship with the to-be-scored fault position, or one historical picture corresponds to only one fault position and the fault position is a to-be-scored fault position, the historical picture is added to the first set; wherein the constituting relationship means that one fault position belongs to a part of another fault position;

[0075] if one historical picture corresponds to two fault positions, neither of which is a to-be-scored fault position and the relationship between the two fault positions is a constituting relationship, or one historical picture corresponds to only one fault position and the fault position is not a to-be-scored fault position, the historical picture is added to the second set;

[0076] If a historical photo corresponds to two faulty parts, one of which is a faulty part to be scored and the two faulty parts are not related, then the historical photo is added to the third set.

[0077] Step 2: The formula for calculating the comprehensive similarity score of the faulty parts to be scored is as follows:

[0078] In the formula, S a This represents the overall similarity score. a Indicates the faulty part to be scored. For the fault image and the first set of images i 1 Image similarity of photos; n 2 This represents the total number of historical photographs in the first set. b This indicates a fault location that is not related to the fault location to be evaluated. The fault image is related to the first one in the second set. i 2 Image similarity of photos; m 2 This represents the total number of historical photos in the second set; The fault image is related to the third set. u Image similarity of photos; k This represents the total number of historical photos in the third set;

[0079] Repeat the above steps to calculate the overall similarity score for each potentially faulty component.

[0080] Secondly, the present invention provides an intelligent maintenance management system for industrial production equipment, comprising:

[0081] The acquisition module is used to acquire repair work orders to be assigned, and the repair work orders to be assigned include at least the following information: faulty equipment, faulty location, and fault phenomenon.

[0082] The first determination module is used to determine multiple suspected causes of failure based on a pre-built knowledge graph, according to the faulty equipment, the faulty location, and the faulty phenomenon.

[0083] The first construction module is used to select all work orders associated with the suspected fault cause from a pre-established set of historical repair work orders, and construct a set of associated work orders;

[0084] The second construction module obtains the maintenance workers corresponding to each work order in the associated work order set, and constructs a maintenance worker set using all the obtained maintenance workers.

[0085] A second determining module is configured to determine the difficulty of fault maintenance according to the number of repairmen in the repairman set and the total number of repairmen obtained in advance.

[0086] A third determining module is configured to determine an order dispatching method according to the difficulty of fault maintenance, and determine the assigned repairman according to the order dispatching method.

[0087] Compared with the prior art, the present application has the following beneficial effects:

[0088] 1. The present application implements a fault maintenance difficulty driven differentiated order dispatching strategy. The to-be-assigned work orders are divided into three difficulty levels of a first level (general type), a second level (technical type) and a third level (professional type) based on the fault maintenance response rate, and each level corresponds to different order dispatching rules. The general type fault has a low technical difficulty, and the repairmen are directly dispatched in a reverse order according to the current work order quantity to simplify the order dispatching steps. The technical type fault considers technical knowledge and experience, and a comprehensive evaluation rule based on multiple evaluation indexes is adopted to finally determine a better repairman comprehensive adaptability. The professional type fault focuses more on fault matching degree, and the similarity of the fault equipment is considered more highly, and a corresponding order dispatching rule is adopted to ensure that the most suitable repairman is dispatched to improve the maintenance efficiency and accuracy.

[0089] 2. The present application is based on a fault part identification method based on fault picture similarity. In the case that the repair work order fault description is incomplete, the similarity calculation is performed by obtaining the fault picture and the historical fault picture set, and the fault part is determined by further analyzing and processing the similarity calculation result, so that the automatic emergency processing can be performed in the case that the fault description cannot directly determine the fault part, without manual intervention, and the intelligent degree is further improved. BRIEF DESCRIPTION OF DRAWINGS

[0090] Figure 1 is a flowchart of the industrial production equipment maintenance intelligent management method in embodiment 1;

[0091] Figure 2 is a knowledge graph model of fault maintenance in embodiment 1;

[0092] Figure 3 is a flowchart of a specific embodiment of the industrial production equipment maintenance intelligent management method in embodiment 1. DETAILED DESCRIPTION

[0093] It should be noted that:

[0094] The technical solutions of the present application will be described in detail below with reference to the drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solutions of the present application, and are not limitations of the technical solutions of the present application. In the case of no conflict, the technical features in the embodiments of the present application and the embodiments can be combined with each other.

[0095] The term "and / or", only describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the three cases of A alone, A and B together, and B alone. In addition, the character " / " generally represents that the associated objects before and after it are in an "or" relationship. Embodiments

[0096] In combination Figure 1 The present embodiment introduces an intelligent maintenance dispatching method for industrial production equipment, which comprises:

[0097] Step S1, obtaining a to-be-assigned repair work order, the to-be-assigned repair work order at least comprising the following information: fault equipment, fault position, fault phenomenon;

[0098] When repair is needed, the repair personnel need to fill in / modify the equipment number, and according to the equipment number, the basic information of the fault equipment can be automatically obtained, including equipment number, equipment category, installation position, start-of-use month, etc. The repair personnel also fill in the fault information, including emergency degree, fault type, whether to repair immediately, appointment repair time, fault description containing fault position and fault phenomenon, etc., and upload fault pictures, and perform repair submission operation. After repair, the work order automatically generates a repair order number, and it is suggested that one fault is associated with one repair each time. The above is a specific embodiment. In actual work, the information of fault equipment, fault position and fault phenomenon belongs to the information that must be recorded in the repair work order, and other information can be set as mandatory or non-mandatory according to actual situation.

[0099] In the present embodiment, when the repair personnel fill in the fault equipment, fault position and fault phenomenon, there are corresponding preset options for the repair personnel to choose from, so as to facilitate subsequent dispatching and distribution and construction of knowledge graph.

[0100] In a specific embodiment, when the actual fault position or fault phenomenon does not have a corresponding preset option, NLP / big model is used for semantic analysis to extract the fault position and fault phenomenon in the fault description.

[0101] In this embodiment, since there can be multiple repair work orders at the same time, the submitted work orders are sorted according to the fault emergency degree filled by the repair personnel, whether immediate repair, reservation repair time and other factors. When the work order needs immediate repair, the reservation repair time is directly set as the reporting time of the work order. In combination with Table 1, the priority of work order repair assignment is determined according to the reservation repair time and the emergency degree, and the repair work orders are sorted and displayed according to the priority, so that emergency work orders can be repaired in time.

[0102] The work order auditors such as the foremen / production team leaders can pass or reject the work orders in this module to timely eliminate false reports from the source.

[0103]

[0104] After the to-be-assigned work order is approved, step S2 is entered.

[0105] In step S2, based on the pre-constructed knowledge graph, a plurality of suspected fault causes are determined according to the fault equipment, the fault position and the fault phenomenon.

[0106] In one specific embodiment, the construction method of the knowledge graph is as follows:

[0107] The historical repair work order information of the production equipment management system is extracted, combined with Figure 3 In a specific embodiment flowchart, NLP, CNN, RNN and other algorithms are used to extract entity relationships from the work order information, and the production equipment fault maintenance knowledge graph is constructed by knowledge fusion, knowledge reasoning and other methods, combined with Figure 2 .

[0108] The types of the entities include: fault equipment, fault position, fault phenomenon, fault cause, work order number, repair method, repair worker and related spare parts.

[0109] In combination with Table 2, the types of each entity are explained as follows:

[0110]

[0111] The relationship types between each entity are as follows:

[0112]

[0113] In one specific embodiment, based on the knowledge graph, according to the positioning nodes of the fault position and the fault phenomenon of the to-be-assigned work order, the suspected fault causes and the possible occurrence probabilities of each suspected fault cause in the current to-be-assigned work order can be inferred, and the inference can be performed by a Bayesian network, a graph neural network and the like.

[0114] Step S3, selecting work orders associated with the suspected fault cause from a pre-established historical repair work order set, and constructing an associated work order set; the association means that the suspected fault cause is one of the fault causes of the associated work order.

[0115] Selecting all work orders associated with the suspected fault cause to construct a preliminary associated work order set , and the fault cause has a many-to-many relationship with the associated work order. In combination Figure 3 with the specific embodiment flowchart, if the associated work order (similar historical work order) set is empty, manually intervene in dispatching.

[0116] Since the purpose of the present embodiment is to finally determine the assigned repairman, in one specific embodiment, the associated work orders corresponding to repairmen who have other repair tasks and do not meet the time requirements are filtered out to obtain the final associated work order set according to whether the repair order to be assigned is immediately repaired, whether the repair time is reserved, and the specific scheduling situation of the corresponding repairman , if is empty, the repair order to be assigned enters the manual dispatch pool, and the repair team leader manually dispatches or the repairman actively grabs the order.

[0117] This specific embodiment is only one specific way to determine the dispatchable repairman. When facing very urgent repair orders, it is also possible to determine whether the repairman is dispatchable only according to whether the repairman is on duty, and all on-duty repairmen are considered dispatchable. In different situations, the standard for whether a repairman is dispatchable can be flexibly adjusted.

[0118] Step S4, obtaining the repairmen corresponding to each work order in the associated work order set, and constructing a repairman set using all the obtained repairmen ;

[0119] Step S5, determining the fault repair difficulty according to the number of repairmen in the repairman set and the pre-obtained total number of repairmen;

[0120] Considering that it is more reasonable to implement differentiated dispatching by using different dispatching rules for different fault levels, the present embodiment uses the fault repair response rate to divide repair faults into three difficulty levels: the first level (general type), the second level (technical type), and the third level (professional type) (the division threshold can be configured according to actual business), wherein is the total number of repairmen, 1 is the number of repairmen;

[0121] Step S6, determining the assigned repairman of the repair order to be assigned according to the corresponding dispatching rule according to the fault repair difficulty.

[0122] In combination with Table 4, in a specific embodiment, the first level (general type), the second level (technical type) and the third level (professional type) are divided as follows:

[0123]

[0124] The first range, the second range and the third range corresponding to the first level (general type), the second level (technical type) and the third level (professional type) are 0.85-1, 0.15-0.85 and 0-0.15 respectively.

[0125] When the fault maintenance difficulty is the first level (general type), the dispatching method is:

[0126] Obtaining the backlog order duration of each maintenance worker in the maintenance worker set, arranging the maintenance workers in descending order according to the backlog order duration, and determining the first dispatchable maintenance worker in the sequence as the assigned maintenance worker;

[0127] In a specific embodiment, the calculation formula of the backlog order duration is:

[0128]

[0129] In the formula, T is the backlog order duration, is the current order quantity of the maintenance worker corresponding to the order, is the maintenance time of the i3th order in the historical orders of the maintenance worker, and n1 is the number of historical orders of the maintenance worker.

[0130] In addition, the backlog order duration can also be directly obtained by multiplying the current order quantity of the maintenance worker by a preset coefficient.

[0131] When the fault maintenance difficulty is the second level (technical type), the dispatching method corresponding to rule 2 is:

[0132] Step S61, evaluating each order in the associated order set in turn as a to-be-evaluated order;

[0133] Step S62, obtaining the evaluation index of the to-be-evaluated order; the evaluation index includes positive index and negative index;

[0134] The positive index includes: the occurrence probability of the suspected fault cause corresponding to the order in the to-be-assigned repair order, the recall weight of the fault equipment corresponding to the order, the maintenance one-time qualified rate of the maintenance worker corresponding to the order, and the maintenance comprehensive evaluation score of the order;

[0135] The negative index includes: the backlog order duration and the maintenance time of the maintenance worker corresponding to the order;

[0136] The determination method of each evaluation index is respectively:

[0137] The determination method of the possible occurrence probability of the suspected fault cause corresponding to the work order in the to-be-assigned repair work order includes:

[0138] Based on the pre-constructed knowledge graph, the occurrence probability of each suspected fault cause is inferred according to the positioning node of the fault equipment, fault position and fault phenomenon in the to-be-assigned repair work order; in some specific embodiments, the inference can be performed by using a Bayesian network or a graph neural network;

[0139] If the work order corresponds to one suspected fault cause, the occurrence probability of the suspected fault cause is determined as the possible occurrence probability;

[0140] If the work order corresponds to multiple suspected fault causes, one with the highest occurrence probability among the multiple suspected fault causes is selected as the possible occurrence probability;

[0141] The determination method of the recall weight of the fault equipment corresponding to the work order includes:

[0142] The recall weight is determined according to the similarity between the fault equipment corresponding to the work order and the fault equipment of the to-be-assigned repair work order; in one specific embodiment, the relationship between the fault equipment corresponding to the work order and the fault equipment of the to-be-assigned work order can be divided into four types: same equipment, same model equipment, same type equipment and other equipment; the recall weights corresponding to the four types are 1, 0.9, 0.6 and 0.3 respectively, and the recall weight gradually decreases as the similarity of the equipment decreases.

[0143] The calculation method of the first-time repair qualified rate is:

[0144] In the formula, F is the first-time repair qualified rate, and k1 is the total number of repairs of the work order.

[0145] The repair comprehensive evaluation score is determined according to the repair acceptance result. In one specific embodiment, the repair result can be scored according to the functional recovery degree, appearance recovery degree, whether the structural strength is reduced and the part replacement rate, and the specific scoring standard can be different according to the different needs of different equipment for different parameters, different weights are given to each scoring parameter, and finally the repair comprehensive evaluation score is given by manual or machine. Other scoring methods that can be performed according to the repair acceptance result can also be applied in this embodiment, for example, the satisfaction degree of the repair personnel is directly scored, which is not limited here.

[0146] For the reverse index:

[0147] The calculation formula of the backlog work order duration is:​

[0148]

[0149] In the formula, T represents the duration of the backlog of work orders. This represents the current number of work orders for the maintenance worker corresponding to this work order. Let n1 be the repair time for the i3rd work order in the repairman's historical work orders, and n1 be the number of historical work orders for the repairman.

[0150] Regarding repair time, the repair time is recorded from the start of the repair until the repair is completed, and this time will be included in the work order for use in subsequent related work orders.

[0151] Step S63: Based on the entropy weight method, evaluate each work order in the associated work order set according to the evaluation index to obtain the comprehensive evaluation parameters of each work order;

[0152] Step S63 includes: Step 1, constructing the original indicator data matrix with the associated work orders and evaluation indicators as the number of rows and columns, respectively:

[0153] ;in, Represents the original indicator data matrix. Represents the first element in the original indicator data matrix. The first related work order The numerical value of each evaluation indicator;

[0154] Step 2: Normalize the positive and negative indicators:

[0155] For positive indicators:

[0156] For negative indicators:

[0157] ;in For the first The numerical values ​​of the evaluation indicators For the first The maximum value of each evaluation indicator For the first The minimum value of each evaluation indicator;

[0158] Step 3: Calculate the corresponding weight of each item in the original indicator data matrix:

[0159] In the formula, p ij Indicates the first The first related work order The weight of each evaluation indicator, where n represents the total number of work orders in the associated work order set;

[0160] Step 4: Establish a weight matrix corresponding to the original indicator data matrix:

[0161] ;

[0162] Step 5, calculate the first... Entropy value of the evaluation index:

[0163] In the formula, e j Indicates the first The entropy value of the evaluation index;

[0164] Step 6, calculate the first... The weight coefficients of each evaluation indicator are determined, and a weight coefficient matrix is ​​determined based on the weight coefficients of each evaluation indicator to interact with the original indicator data matrix.

[0165] In the formula, w j Indicates the first The weighting coefficients of the evaluation indicators;

[0166] Step 7: The original index data matrix and the weight coefficient matrix are multiplied by a dot product to obtain a weighted matrix. The corresponding values ​​in the weighted matrix are... The formula for calculating the weighted value is:

[0167] In the formula, Represents the weighted matrix in which the first element is... The first related work order The weighted values ​​of the evaluation indicators;

[0168] Step 8: Calculate the first... Comprehensive evaluation parameters for each associated work order:

[0169] In the formula, Indicates the first Comprehensive evaluation parameters for each associated work order; and They represent the first The distance from the comprehensive evaluation value of each related work order to the positive and negative ideal solutions; and They represent the first The weighted maximum and weighted minimum values ​​of the evaluation indicators;

[0170] Step S64: Based on the comprehensive evaluation parameters, arrange the maintenance workers corresponding to each associated work order in reverse order, and determine the first dispatchable maintenance worker in the sequence as the assigned maintenance worker.

[0171] Here, the dispatchable standard can be that, according to whether the repair order to be assigned is immediately repaired, the repair time is reserved, and the specific scheduling of the corresponding repairman, the repairman left after filtering out the repairman who has other repair tasks and does not meet the time requirement, or only according to whether the repairman is on duty to determine whether the repairman is dispatchable.

[0172] When the fault repair difficulty is the third level, the dispatching method corresponding to rule 1 includes:

[0173] Step S6.1, obtain the evaluation parameters of each work order in the associated work order set, the evaluation parameters including: recall weight of the fault equipment corresponding to the work order; possible occurrence probability of the suspected fault cause corresponding to the work order in the repair order to be assigned; the determination method of the possible occurrence probability is consistent with the determination method of the occurrence probability of the second level recorded in the foregoing.

[0174] If there are multiple suspected fault causes corresponding to the work order, take the maximum value of the occurrence probabilities of the multiple suspected fault causes as the possible occurrence probability;

[0175] In a specific embodiment, the determination method of the recall weight of the fault equipment is: dividing the relationship between the fault equipment corresponding to the work order and the fault equipment of the repair order to be assigned into four types: same equipment, same model equipment, same type equipment, and other equipment; the recall weights corresponding to the four types are 1, 0.9, 0.6, and 0.3 respectively, and the recall weight gradually decreases as the similarity of the equipment decreases;

[0176] Step S6.2, calculate the product of the maximum probability in the set of possible occurrence probabilities and the recall weight as an evaluation value, and assign scores to the priority of each work order according to the evaluation value;

[0177] Step S6.3, according to the evaluation value, arrange the repairmen corresponding to each associated work order in descending order, and determine the first dispatchable repairman in the sequence as the assigned repairman.

[0178] Step S7, repair execution, the assigned repairman performs repair work, and records information such as start time, end time, records the cause of the fault, the repair method used to solve the fault, and the spare parts replaced during the repair process.

[0179] ​Step S8, acceptance module after repair is completed: the repairer performs acceptance operation on the repair work order completed, and when the acceptance fails, the repairer needs to re-perform the repair, and when the acceptance passes, the repair result needs to be comprehensively scored. In a specific embodiment, the repair result can be scored according to the degree of functional recovery after repair, the degree of appearance recovery, whether the structural strength is reduced, and the part replacement rate, and specific scoring standards can be different for different parameters according to the needs of different equipment, and different weights are given to each scoring parameter, and finally a comprehensive score is given by manual or machine.

[0180] Step S9, repair result confirmation, after acceptance, the repair team leader can audit and modify the repair content of the repair work order such as fault description, fault reason and repair method in this module, to ensure that the fault description can clearly determine the fault position and fault phenomenon, and the description is complete, clear and unambiguous, to support the continuous improvement of the knowledge graph update and the corresponding intelligent dispatching function.

[0181] Embodiment 2

[0182] The difference between this embodiment and embodiment 1 is that if the repair work order to be assigned does not include the fault position and cannot extract the directly associated fault position, the fault position is determined by the fault picture. The specific determination method includes (the fault position mentioned below refers to the directly associated fault position, that is, the path from the fault position node to the work order node in the knowledge graph is the shortest):

[0183] Step Q1, obtaining the fault picture of the repair work order to be assigned;

[0184] Step Q2, comparing the fault picture with each picture in the pre-obtained historical fault picture set respectively to obtain the picture similarity between the fault picture and each picture;

[0185] The historical fault picture set Based on the extraction of the fault position entity category in embodiment 1. In some specific embodiments, the picture similarity between the fault picture and each picture is calculated by SIFT, SURF, Siamese, CNN and the like.

[0186] Step Q3, screening out historical pictures with picture similarity exceeding a threshold value and corresponding fault position being 3 or less to construct a comparison picture set ; obtaining the fault position corresponding to each picture in the comparison picture set, and taking all the obtained fault positions as possible fault positions. Setting the corresponding fault position to be 3 or less mainly considers that when the number of fault positions corresponding to the fault picture is too large, the positioning of the fault position is prone to deviation.

[0187] Step Q4, determining a similarity comprehensive score of each possible faulty component according to the comparison picture set and the picture similarity;

[0188] Step Q4 is performed in sequence for each possible faulty component as a fault site to be scored, and Step Q4 specifically includes:

[0189] Step 41, according to the relationship between the fault site corresponding to the historical picture and the fault site to be scored, dividing the comparison picture set into a first set , a second set , and a third set ; ;

[0190] If one historical picture corresponds to two fault sites, one of which is the fault site to be scored and the other is in a constituting relationship with the fault site to be scored, or one historical picture corresponds to only one fault site which is the fault site to be scored, the historical picture is added to the first set; wherein the constituting relationship means that one of the fault sites is part of the other fault site (see Table 3 in the foregoing);

[0191] If one historical picture corresponds to two fault sites, neither of which is the fault site to be scored and the relationship between the two fault sites is a constituting relationship; or one historical picture corresponds to only one fault site which is not the fault site to be scored, the historical picture is added to the second set;

[0192] If one historical picture corresponds to two fault sites, one of which is the fault site to be scored and the relationship between the two fault sites is not a constituting relationship, the historical picture is added to the third set;

[0193] Step 42, the calculation formula of the similarity comprehensive score of the fault site to be scored is:

[0194]

[0195] In the formula, S a represents the similarity comprehensive score, a represents the fault site to be evaluated, represents the picture similarity between the fault picture and the first set of the first i 1 picture; n 2 represents the total number of historical pictures in the first set; b represents the fault site which is not in a constituting relationship with the fault site to be evaluated; represents the picture similarity between the fault picture and the second set of the first i 2 picture;m 2 represents the total number of historical photos in the second set; represents the picture similarity of the fault picture and the first historical photo in the third set; u represents the picture similarity of the fault picture and the first historical photo in the third set; k represents the total number of historical photos in the third set; b has multiple, respectively, to calculate.

[0196] Repeat the above steps to calculate the similarity comprehensive score of each possible fault component.

[0197] Step Q5, determine the possible fault component with the highest similarity comprehensive score as the fault site of the repair order to be assigned.

[0198] The embodiment can determine a more accurate possible fault site according to the fault picture analysis when the fault site is unclear by the above method. Embodiment 3

[0199] The embodiment provides an intelligent maintenance management system for industrial production equipment, which comprises:

[0200] The repair module: the repair personnel fill in / modify the equipment number, automatically bring out the basic information of the equipment including the equipment number, equipment category, installation position, start using month, etc., fill in the fault information including the emergency degree, fault type, whether to repair immediately, reservation repair time, fault description, etc., upload the fault picture, and perform repair submission operation. After repair, the work order automatically generates a repair order number, and it is suggested to associate one fault with each repair.

[0201] The work order review module: according to factors such as fault emergency degree, whether to repair immediately, reservation repair time, etc., the submitted work orders to be reviewed are sorted, the reservation repair time is set as the work order reporting time when repairing immediately, and the repair work orders are sorted and displayed according to the reservation repair time and the emergency degree. The work order reviewer such as the foreman / production team leader can pass or reject the work order in this module to exclude false reports from the source.

[0202] The third part is the dispatch module: the passed work orders are matched with fault sites according to fault description and fault picture, and historical work orders with the same fault reason are screened out combined with fault phenomena. The corresponding maintenance workers are arranged, the work load, the maintenance one-time qualification rate, the maintenance time, the maintenance comprehensive evaluation score and other factors are comprehensively measured, the optimal maintenance worker is selected based on the differentiated dispatch strategy to realize automatic dispatch. When the historical work order cannot be matched, the maintenance team leader manually dispatches or the maintenance worker independently takes the order strategy.

[0203] The fourth part is the maintenance execution module: used to record the maintenance process of the maintenance worker such as the start time and the end time, record the cause of the fault, the maintenance method used to solve the fault, and the information of the spare parts replaced in the maintenance process.

[0204] The fifth part is an acceptance module: the repairer performs an acceptance operation on the completed repair work order in this module. If the acceptance fails, the repairer needs to perform the repair again. If the acceptance passes, the repair result needs to be comprehensively scored.

[0205] The sixth part is a repair result confirmation module: after acceptance, the repair team leader can audit and modify the repair order filling content such as fault description, fault reason, and repair method in this module to ensure that the description is complete, clear, and unambiguous, and to support the updating of the fault knowledge graph and the intelligent dispatching function. Embodiment 4

[0206] Based on the same inventive concept as Embodiment 1, this embodiment provides an industrial production equipment intelligent repair management device, which comprises:

[0207] An acquisition module is configured to acquire a to-be-assigned repair work order, wherein the to-be-assigned repair work order comprises at least the following information: a fault equipment, a fault position, and a fault phenomenon.

[0208] A first determination module is configured to determine a plurality of suspected fault reasons based on a pre-constructed knowledge graph according to the fault equipment, the fault position, and the fault phenomenon.

[0209] A first construction module is configured to select all work orders associated with the suspected fault reasons from a pre-established historical repair work order set, and construct an associated work order set.

[0210] A second construction module is configured to acquire a repair worker corresponding to each work order in the associated work order set, and construct a repair worker set using all the acquired repair workers.

[0211] A second determination module is configured to determine a fault repair difficulty according to a number of repair workers in the repair worker set and a pre-acquired total number of repair workers.

[0212] A third determination module is configured to determine a dispatching method according to the fault repair difficulty, and determine an assigned repair worker according to the dispatching method.

[0213] Those skilled in the art will appreciate that embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0214] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0215] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0216] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 one or more flow or blocks

[0217] The embodiments of the present application described above are merely intended to illustrate the present application, but not to limit the present application. The skilled in the art can make many modifications and improvements without departing from the spirit and scope of the present application, which should be protected as long as they fall within the scope of the present application and the claims.

Claims

1. An intelligent maintenance dispatching method for industrial production equipment, characterized in that: The method comprises the following steps: obtain a to-be-assigned repair work order, wherein the to-be-assigned repair work order comprises at least the following information: faulty equipment, fault location, and fault phenomenon; based on a pre-constructed knowledge graph, determine a plurality of suspected fault causes of the to-be-assigned repair work order according to the faulty equipment, the fault location, and the fault phenomenon; select all work orders associated with the suspected fault causes from a pre-established historical repair work order set, and construct an associated work order set; obtain a maintenance worker corresponding to each work order in the associated work order set, and construct a maintenance worker set using all the obtained maintenance workers; determine a fault maintenance difficulty according to the number of maintenance workers in the maintenance worker set and a pre-obtained total number of maintenance workers; determine an assigned maintenance worker of the to-be-assigned repair work order according to a dispatch method determined according to the fault maintenance difficulty.

2. The intelligent maintenance dispatching method for industrial production equipment according to claim 1, characterized in that: The method for determining the fault maintenance difficulty according to the number of maintenance workers in the maintenance worker set and the pre-obtained total number of maintenance workers comprises the following steps: determine a fault maintenance response rate according to the number of maintenance workers and the pre-obtained total number of maintenance workers; The calculation formula of the failure maintenance response rate is: ; in the formula, is the failure maintenance response rate; is the total number of maintenance workers, and m1 is the number of maintenance workers in the maintenance worker set; if the fault maintenance response rate belongs to a pre-set first range, determine that the fault maintenance difficulty is of a first level; if the fault maintenance response rate belongs to a pre-set second range, determine that the fault maintenance difficulty is of a second level; if the fault maintenance response rate belongs to a pre-set third range, determine that the fault maintenance difficulty is of a third level; the first range, the second range, and the third range decrease in turn.

3. The intelligent maintenance dispatching method for an industrial production device according to claim 2, characterized in that: When the fault maintenance difficulty is of the first level, the method for determining the assigned maintenance worker of the to-be-assigned repair work order according to the dispatch method comprises the following steps: obtain the backlog work order duration of each maintenance worker in the maintenance worker set, arrange the maintenance workers in a descending order according to the backlog work order duration, and determine the first dispatchable maintenance worker in the sequence as the assigned maintenance worker.

4. The intelligent maintenance dispatching method for industrial production equipment according to claim 3, characterized in that: When the fault maintenance difficulty is of the second level, the method for determining the assigned maintenance worker of the to-be-assigned repair work order according to the dispatch method comprises the following steps: evaluate each associated work order in the associated work order set in turn as a to-be-evaluated work order; obtain an evaluation index of the to-be-evaluated work order, wherein the evaluation index comprises a positive index and a negative index; the positive index comprises: a possible occurrence probability of the suspected fault cause corresponding to the work order in the to-be-assigned repair work order, a recall weight of the fault equipment corresponding to the work order, a pre-obtained maintenance one-time qualification rate of the work order, and a maintenance comprehensive evaluation score; the negative index comprises: a pre-obtained backlog work order duration of the maintenance worker corresponding to the work order, and a maintenance time of the work order; evaluate each work order in the associated work order set according to the evaluation index to obtain a comprehensive evaluation parameter of each work order; determine the assigned maintenance worker according to the comprehensive evaluation parameter.

5. The intelligent maintenance dispatching method for an industrial production device according to claim 4, characterized in that: The method for evaluating each associated work order in the associated work order set according to the evaluation index to obtain a comprehensive evaluation parameter of each associated work order comprises the following steps: Step 1: construct an original index data matrix with the associated work orders and the evaluation indexes as the number of rows and the number of columns, respectively; ; wherein, denotes the original indicator data matrix, denotes the value of the i-th evaluation indicator of the j-th associated work order in the original indicator data matrix, denotes the value of the i-th evaluation indicator of the j-th associated work order in the original indicator data matrix, denotes the value of the i-th evaluation indicator of the j-th associated work order in the original indicator data matrix. Step 2: normalize the positive index and the negative index: for the positive index: ; for negative indicators: ; wherein is the value of the th evaluation criterion, is the maximum value of the th evaluation criterion, is the minimum value of the th evaluation criterion; Step 3, calculating the corresponding proportion of each item in the original index data matrix: In the formula, p ij Indicates the first The first related work order The weight of each evaluation indicator, where n represents the total number of work orders in the associated work order set; Step 4, establishing a proportion matrix corresponding to the original index data matrix: ； Step 5, calculate the entropy value of the item evaluation index: Entropy value of item evaluation index: ; wherein e j represents the first entropy value of the evaluation index Step 6, calculate the first... The weight coefficients of each evaluation indicator are determined, and a weight coefficient matrix is ​​determined based on the weight coefficients of each evaluation indicator to interact with the original indicator data matrix. ; wherein w j denotes the weighting factor of the item evaluation index; Step 7, the original index data matrix and the weight coefficient matrix are multiplied to obtain a weighted matrix, and the corresponding weighted value in the weighted matrix is calculated according to the following formula: The calculation formula of the weighted value corresponding to the weighted matrix is as follows: ; wherein, represents the weighted value of the evaluation index of the i-th associated work order in the weighted matrix; and represents the weighted value of the evaluation index of the i-th associated work order in the weighted matrix; and represents the weighted value of the evaluation index of the i-th associated work order in the weighted matrix; and Step 8, calculate the comprehensive evaluation parameter of the associated work order: ; wherein, represents a comprehensive evaluation parameter of the th associated work order; and respectively represent the distance from the comprehensive evaluation value of the th associated work order to the positive ideal solution and the negative ideal solution; and respectively represent the weighted maximum value and the weighted minimum value of the th evaluation index; The method further comprises: According to the comprehensive evaluation parameter, the repairman corresponding to each associated work order is arranged in descending order, and the first dispatchable repairman in the sequence is determined as the assigned repairman.

6. The intelligent maintenance dispatching method for an industrial production device according to claim 5, characterized in that: When the fault repair difficulty is the third level, the assigned repairman of the to-be-assigned repair work order is determined according to the dispatch method, which comprises: Obtaining the evaluation parameter of each associated work order in the associated work order set, According to the evaluation parameter, the evaluation value of each associated work order is determined; According to the evaluation value, the repairman corresponding to each associated work order is arranged in descending order, and the first dispatchable repairman in the sequence is determined as the assigned repairman; The evaluation parameter comprises: the recall weight of the fault equipment corresponding to the work order, and the possible occurrence probability of the suspected fault cause corresponding to the work order in the to-be-assigned repair work order; The product of the possible occurrence probability and the recall weight is calculated as the evaluation value.

7. The intelligent maintenance dispatching method for industrial production equipment according to claim 4 or 6, characterized in that: The method for determining the possible occurrence probability of the suspected fault cause corresponding to the work order in the to-be-assigned repair work order comprises: Based on the pre-constructed knowledge graph, the occurrence probability of each suspected fault cause is inferred according to the positioning nodes of the fault equipment, fault position and fault phenomenon in the to-be-assigned repair work order in the knowledge graph; If the work order corresponds to one suspected fault cause, the occurrence probability of the suspected fault cause is determined as the possible occurrence probability; If the work order corresponds to multiple suspected fault causes, the one with the highest occurrence probability among the multiple suspected fault causes is determined as the possible occurrence probability; The method for determining the recall weight of the fault equipment corresponding to the work order comprises: According to the similarity between the fault equipment corresponding to the work order and the fault equipment of the to-be-assigned repair work order, the recall weight is determined; The higher the similarity, the higher the recall weight.

8. The intelligent maintenance dispatching method for industrial production equipment according to claim 1, characterized in that: If the to-be-assigned repair work order does not include a fault position, the method for determining the fault position comprises: Obtaining the fault picture of the to-be-assigned repair work order; Comparing the fault picture with each picture in the pre-obtained historical fault picture set respectively to obtain the picture similarity between the fault picture and each picture; Screening out historical pictures with picture similarity exceeding a threshold value and corresponding fault position being three or less to construct a comparison picture set; Obtaining the fault position corresponding to each picture in the comparison picture set, and taking all obtained fault positions as possible fault positions; According to the comparison picture set and the picture similarity, determining the similarity comprehensive score of each possible fault component; The possible fault component with the highest similarity comprehensive score is determined as the fault position of the to-be-assigned repair work order.

9. The intelligent maintenance dispatching method for an industrial production device according to claim 8, characterized in that: Each possible fault component is sequentially taken as a to-be-scored fault position for scoring; According to the comparison picture set and the picture similarity, determining the similarity comprehensive score of each to-be-scored fault position comprises: Step 1, according to the relationship between the fault parts corresponding to the historical pictures and the to-be-scored fault parts, the set of comparison pictures is divided into a first set, a second set and a third set; If one historical picture corresponds to two fault parts, one of which is the to-be-scored fault part, and the other is in a constituting relationship with the to-be-scored fault part, or one historical picture only corresponds to one fault part which is the to-be-scored fault part, the historical picture is added to the first set; wherein the constituting relationship means that one of the fault parts belongs to a part of the other fault part; If one historical picture corresponds to two fault parts, both of which are not to-be-scored fault parts and the relationship between the two fault parts is a constituting relationship, or one historical picture only corresponds to one fault part which is not a to-be-scored fault part, the historical picture is added to the second set; If one historical picture corresponds to two fault parts, one of which is the to-be-scored fault part and the relationship between the two fault parts is not a constituting relationship, the historical picture is added to the third set; Step 2, the calculation formula of the similarity comprehensive score of the to-be-scored fault part is: In the formula, S a This represents the overall similarity score. a Indicates the faulty part to be scored. For the fault image and the first set of images i 1 Image similarity of photos; n 2 This represents the total number of historical photographs in the first set. b This indicates a fault location that is not related to the fault location to be evaluated. The fault image is related to the first one in the second set. i 2 Image similarity of photos; m 2 This represents the total number of historical photos in the second set; The fault image is related to the third set. u Image similarity of photos; k This represents the total number of historical photos in the third set; Repeat the above steps to calculate the similarity comprehensive score of each possible fault component.

10. An intelligent maintenance management system for industrial production equipment, characterized by: Comprise: An acquisition module is configured to acquire a to-be-assigned repair work order, wherein the to-be-assigned repair work order comprises at least the following information: Fault equipment, fault part, and fault phenomenon; A first determination module is configured to determine a plurality of suspected fault causes based on a pre-constructed knowledge graph and according to the fault equipment, the fault part, and the fault phenomenon; A first construction module is configured to select all work orders associated with the suspected fault causes from a pre-established set of historical repair work orders, and construct an associated work order set; A second construction module is configured to acquire maintenance workers corresponding to each work order in the associated work order set, and construct a maintenance worker set using all the acquired maintenance workers; A second determination module is configured to determine a fault maintenance difficulty according to the number of maintenance workers in the maintenance worker set and a pre-acquired total number of maintenance workers; A third determination module is configured to determine a dispatch method according to the fault maintenance difficulty, and determine an assigned maintenance worker according to the dispatch method.

Images