Equipment maintenance resource optimization method based on data driving

Abstract

The invention discloses an equipment maintenance resource optimization method based on data driving, which comprises the following steps: collecting and arranging maintenance logs of equipment, and removing abnormal records to obtain reliability data of the equipment; according to the reliability data of the equipment, common service interruption causes and maintenance resources associated with common service interruption types are identified, the warehouse-out time distribution of the maintenance resources is analyzed, and periodic demand changes are found out; training and optimizing the event counting distribution model to obtain a prediction model of maintenance resource consumption; and inputting the historical consumption data of the maintenance resources into the maintenance resource consumption prediction model, outputting predicted consumption data of the maintenance resources, and determining the purchase cycle and the optimal purchase quantity of the maintenance resources according to the predicted consumption data of the maintenance resources. The invention aims to solve the problems of low maintenance efficiency, poor decision-making quality and difficulty in formulating and executing a preventive maintenance strategy caused by the lack of structuralization of an equipment material list and the lack of effective integration of maintenance resource data and an equipment fault library in the existing technical scheme.

Description

Technical Field

[0001] The present invention relates to the field of equipment maintenance resource optimization, and in particular to a data-driven equipment maintenance resource optimization method. Background Art

[0002] Existing technical solutions establish the relationship between maintenance resource loss and equipment failure by deeply mining a large amount of historical maintenance resource in-and-out data of the enterprise, thereby improving the transparency of maintenance resource management, reducing maintenance resource inventory costs, and accelerating maintenance response speed. These technologies have effectively improved the operational efficiency of equipment. The current equipment bill of materials (BOM) lacks structure, and BOM data exists in an unstructured or semi-structured form, resulting in the inability to quickly and accurately locate the required information during maintenance resource management, equipment maintenance, and troubleshooting, affecting maintenance efficiency and decision-making quality. At the same time, the lack of effective integration of maintenance resource data and equipment fault database makes it difficult to formulate and implement preventive maintenance strategies, which in turn affects the reliability and service life of the equipment. Summary of the invention

[0003] The present invention provides a data-driven equipment maintenance resource optimization method to solve the problems in the existing technical solutions, such as the lack of structuring of the equipment bill of materials (BOM), the lack of effective integration of maintenance resource data and the equipment fault library, resulting in low maintenance efficiency, poor decision-making quality, and difficulty in formulating and implementing preventive maintenance strategies.

[0004] A data-driven equipment maintenance resource optimization method comprises the following steps:

[0005] 1) Collect and organize equipment maintenance logs, including equipment service interruption time, service interruption cause, service interruption type, replacement parts specifications, parts requisition quantity, requisition time, and inventory status, remove abnormal records, and obtain equipment reliability data;

[0006] 2) Based on the reliability data of the equipment, identify the common causes of service interruptions and the maintenance resources associated with the common types of service interruptions, analyze the time distribution of maintenance resource outages, and find out the periodic demand changes as the historical consumption data of maintenance resources;

[0007] 3) Use industry maintenance resource outbound data to train and optimize the event count distribution model to obtain a prediction model for maintenance resource consumption;

[0008] 4) Input the historical consumption data of the maintenance resources obtained in step 2) into the prediction model of the maintenance resource consumption obtained in step 3), output the predicted consumption data of the maintenance resources, and determine the procurement cycle and optimal procurement quantity of the maintenance resources based on the predicted consumption data of the maintenance resources.

[0009] The starting point of the present invention is to comprehensively optimize the equipment maintenance resource management through intelligent tools, to help equipment managers make decisions and take actions at the appropriate time, rather than blindly relying on traditional fixed-mode management. The core idea of ​​the method of the present invention is to evaluate the status of equipment and maintenance resources from multiple dimensions, comprehensively analyze the integration of BOM and fault library, regression analysis of maintenance resource data and supplier performance, and automatically generate structured equipment BOM through intelligent data analysis and matching technology, and optimize maintenance resource procurement strategies and supply chain management. Through this system, it is possible to more accurately predict maintenance resource needs, formulate long-term management plans covering inventory optimization, supplier selection and maintenance strategies, reduce resource waste, and improve supply chain stability and equipment management efficiency.

[0010] In step 1), abnormal records are removed, including:

[0011] If the maintenance log lacks any one or more of the replacement part specifications, part collection quantity, collection time, and inventory status, the maintenance log is abnormal and needs to be deleted.

[0012] In step 2), the common causes of service interruption are wear, corrosion, vibration, leakage or poor lubrication.

[0013] In step 2), the common types of service interruption are function failure, malfunction, and parameter abnormality.

[0014] In step 2), the time distribution of maintenance resource outbound is analyzed to find out the periodic demand changes, including:

[0015] The type of each maintenance resource and the time of its release are respectively the first column and the first row of the table. The cells in the table are the number of parts to be collected. Find out the periodic changes in demand.

[0016] In step 4), the forecast consumption data of the maintenance resources includes: the minimum inventory forecast value of the maintenance resources and the maximum inventory forecast value of the maintenance resources.

[0017] Compared with the common methods currently on the market, the method of the present invention has the following advantages and can solve the technical problems of the existing methods:

[0018] ① Integration of structured BOM and fault library: Combined with the fault library, a structured and hierarchical equipment BOM is constructed to achieve automatic correspondence between maintenance resource names and defect fields, and automatically generate a structured BOM.

[0019] ② Maintenance resource data regression analysis: Based on the maintenance resource replacement records in the maintenance records, data regression analysis is performed at the maintenance resource level and equipment level. Taking the equipment as the object, the maintenance resource entry and exit rules are explored, and the maximum and minimum inventory quantities are recommended. Combined with historical data and prediction models, the optimal procurement time and batch of maintenance resources are determined to balance costs and inventory levels.

[0020] ③Supplier statistical analysis: Statistically analyze the maintenance resource supplier information, compare the performance of different suppliers on the same maintenance resource, and optimize supplier selection and management. Avoid over-reliance on a single supplier and reduce supply chain risks by establishing a diversified supplier system.

[0021] In general, the new solution can provide more comprehensive and optimized equipment maintenance resource management and supply chain strategies. Through automation and intelligent means, it can improve equipment maintenance efficiency, reduce operating costs, and enhance the reliability and flexibility of the supply chain, thus playing an important role in equipment management and operation optimization.

[0022] The method based on the integration of structured BOM and fault library, regression analysis of maintenance resource data and statistical analysis of suppliers has the following beneficial effects:

[0023] 1. Reduce equipment management costs: Through the integration of structured BOM and fault database, the text recognition algorithm is used to achieve automatic matching and mapping of maintenance resource information and fault records, forming a systematic and hierarchical equipment BOM. This process effectively reduces the workload of manual maintenance, improves the automation level of maintenance resource management, and thus reduces equipment management costs.

[0024] 2. Reduce maintenance resource procurement and storage costs: Maintenance resource data regression analysis uses maintenance logs and forecasting models to accurately analyze the periodic changes in maintenance resource demand, and through optimizing inventory strategies, determine the best procurement time and batch. This reduces the uncertainty of maintenance resource entry and exit and redundant inventory, and improves the response speed and efficiency of equipment maintenance.

[0025] 3. Optimize supply chain management: Supplier statistical analysis optimizes supplier selection and management by comprehensively comparing the performance of different suppliers in maintenance resources, avoiding dependence on a single supplier. By establishing a diversified supplier system, supply chain risks are reduced, ensuring the stability and reliability of maintenance resource supply, thereby supporting the long-term healthy operation of equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of the data-driven maintenance resource optimization method in the present invention;

[0027] Figure 2It is a schematic diagram of the construction process of the structured BOM of the device in the present invention;

[0028] Figure 3 A structured BOM list is automatically constructed for the present invention. DETAILED DESCRIPTION

[0029] like Figure 1 As shown, the present invention is based on a data-driven equipment maintenance resource optimization method, and the specific method flow is as follows:

[0030] 1. Integration of structured BOM and fault library:

[0031] Based on the fault library, text recognition technology is used to perform text recognition on the defect fields and BOM material descriptions of the equipment, and the recognition results are maintained in the fault library to form a structured equipment BOM.

[0032] The construction process of structured BOM is as follows: Figure 2 As shown in the figure, the relationship between the equipment's trouble ticket, notification ticket, and work order is used to trace the equipment's maintenance resource outbound records. The outbound records of maintenance resources are matched with the defect location, service interruption cause, and consequences of defect improvement, and the maintenance resource coding, technical parameters, quantity and other information are automatically mapped to the fault library to form a hierarchical and structured equipment BOM. For equipment without trouble tickets, the defect location in the fault library is semantically matched with the maintenance resource name. The matching results are reviewed and confirmed in the defect improvement link to form the final structured equipment BOM data.

[0033] 2. Maintenance resource data regression analysis:

[0034] The present invention performs data regression analysis at the maintenance resource level and the equipment level based on the maintenance resource replacement records in the maintenance records. Taking the equipment as the object, the maintenance resource in-and-out rules are mined. Combining historical data and prediction models, the optimal procurement time and batch of maintenance resources are determined, and the maximum and minimum inventory quantities are recommended to balance the cost and inventory level. The main implementation steps are as follows:

[0035] (1) Data preparation and preprocessing:

[0036] Collect and organize equipment maintenance logs, including but not limited to the time of equipment service interruption, cause of service interruption, specifications of replacement parts, quantity of parts issued, time of issuance, existing inventory, etc., remove invalid or abnormal records, ensure data integrity and consistency, and lay the foundation for subsequent analysis.

[0037] (2) Data modeling and analysis:

[0038] Based on the equipment maintenance records, identify common fault types and their associated maintenance resources, analyze the time distribution of maintenance resources out of the warehouse, and find out periodic demand changes.

[0039] (3) Inventory optimization strategy formulation:

[0040] Based on the historical consumption data and prediction model of maintenance resources, combined with factors such as the MTBF, installation quantity, replenishment time and availability of maintenance resources, the Peimu maintenance resource optimization algorithm is used to calculate the minimum and maximum inventory thresholds of maintenance resources, determine the procurement cycle and optimal procurement quantity of maintenance resources, and reduce inventory costs.

[0041] 3. Supplier statistical analysis:

[0042] Through comprehensive statistical analysis of maintenance resource supplier information, the performance of different suppliers on the same maintenance resource can be effectively compared, including detailed statistics on the frequency of use and inventory quantity of maintenance resources, and analysis of the maximum, minimum and average prices of maintenance resources to obtain comprehensive information on price fluctuations. In addition, statistics on the consumption of maintenance resources at the plant level, device level and equipment level can be collected to understand the demand and usage at each level.

[0043] Through statistical analysis of these data, we can optimize the selection and management of suppliers, thereby avoiding over-reliance on a single supplier, establishing a diversified supplier system, effectively reducing supply chain risks, and improving the stability and reliability of maintenance resource supply.

[0044] The method based on the integration of structured BOM and fault library, regression analysis of maintenance resource data and statistical analysis of suppliers can play an important role in the field of equipment maintenance resource management. Through automation and intelligent technology, it can accurately analyze the laws of equipment failure and maintenance resource demand, optimize inventory management and supply chain selection, reduce equipment management costs and improve maintenance efficiency, and enhance the stability and reliability of equipment operation. In future industrial operations, this method will play an increasingly important role, bringing more value and innovation to equipment management and supply chain optimization.

[0045] Specifically, the method of the present invention comprises:

[0046] ①Structured BOM: Combined with the fault database, a structured and hierarchical equipment BOM is built in the defect system. The corresponding relationship between the maintenance resource name and the defect field is sorted out and a structured BOM is automatically generated.

[0047] ② Maintenance resource data regression: Extract maintenance resource replacement records from maintenance records, perform regression analysis on equipment-level and maintenance resource-level replacement records, and compare the analysis results of defect data and maintenance resource data regression. Use historical data to determine the best procurement time and quantity for maintenance resources. The work can be divided into three steps: 1. Recommendation of maximum and minimum inventory of maintenance resources; 2. Mining of in-and-out inventory rules; 3. Recommendation of maintenance resource procurement cycle.

[0048] ③Supplier statistical analysis: Statistics on supplier information of maintenance resources, and horizontally compare and analyze the indicators of different suppliers of the same maintenance resources. Analyze and count the supplier information of maintenance resource data, such as statistics on the number of maintenance resource replacements, MTBF, price, etc. of different suppliers of the same maintenance resources in the same time period.

[0049] 1. Maintenance resource outbound analysis:

[0050] The outbound records of the general maintenance resource "bearings" are selected as the source data, the consumption of bearings is analyzed, and suggestions for optimizing maintenance resources are given, as shown in Table 1.

[0051] Table 1

[0052]

[0053] 2. Build a structured BOM:

[0054] Use AI text recognition algorithms to match defective parts with maintenance resource names and automatically build a structured BOM list, such as Figure 3 shown.

[0055] 3. Maintain resource inventory optimization calculation:

[0056] Combined with the maintenance resource's MTBF, price, installation quantity, procurement time, availability and other parameters, the minimum and maximum inventory recommendations for maintenance resources are calculated to guide maintenance resource procurement.

[0057] 4. Maintenance resource analysis and optimization suggestions:

[0058] The present invention uses an algorithm model to analyze and calculate the installed quantity, maximum / minimum inventory, MTBF, single purchase quantity, and recommended purchase cycle of maintenance resources. The maintenance resource optimization output data is summarized, including maintenance resource outbound analysis data and inventory optimization suggestions, as shown in Table 2.

[0059] Table 2

[0060]

Claims

1. A data-driven equipment maintenance resource optimization method, characterized in that: The following steps are involved: 1) Collect and organize equipment maintenance logs, including equipment service interruption time, service interruption cause, service interruption type, replacement parts specifications, parts collection quantity, collection time, inventory status, remove abnormal records, and obtain equipment reliability data; 2) Based on the reliability data of the equipment, identify the common causes of service interruptions and the maintenance resources associated with the common types of service interruptions, analyze the time distribution of maintenance resource outages, and find out the periodic demand changes as the historical consumption data of maintenance resources; 3) Use industry maintenance resource outbound data to train and optimize the event count distribution model to obtain a prediction model for maintenance resource consumption; 4) Input the historical consumption data of the maintenance resources obtained in step 2) into the prediction model of the maintenance resource consumption obtained in step 3), output the predicted consumption data of the maintenance resources, and determine the procurement cycle and optimal procurement quantity of the maintenance resources based on the predicted consumption data of the maintenance resources.

2. The data-driven equipment maintenance resource optimization method according to claim 1 is characterized in that: In step 1), abnormal records are removed, including: If the maintenance log lacks any one or more of the replacement part specifications, part collection quantity, collection time, and inventory status, the maintenance log is abnormal and needs to be deleted.

3. The data-driven equipment maintenance resource optimization method according to claim 1 is characterized in that: In step 2), the common causes of service interruption are wear, corrosion, vibration, leakage or poor lubrication.

4. The data-driven equipment maintenance resource optimization method according to claim 1 is characterized in that: In step 2), the common types of service interruption are function failure, malfunction, and parameter abnormality.

5. The data-driven equipment maintenance resource optimization method according to claim 1 is characterized in that: In step 2), the time distribution of maintenance resource outbound is analyzed to find out the periodic demand changes, including: The type of each maintenance resource and the time of its release are respectively the first column and the first row of the table. The cells in the table are the number of parts to be collected. Find out the periodic changes in demand.

6. The data-driven equipment maintenance resource optimization method according to claim 1, characterized in that: In step 4), the forecast consumption data of the maintenance resources includes: the minimum inventory forecast value of the maintenance resources and the maximum inventory forecast value of the maintenance resources.

Images