Spare parts inventory forecasting method, device, computer equipment and medium

Abstract

The present application relates to a spare parts inventory forecasting method, device, computer equipment, storage medium and computer program product. The method generates data on the variation of spare parts usage over time based on preset maintenance specification data and the historical usage of spare parts; obtains spare parts reservation configuration parameters based on the historical reservation of spare parts, and generates data on the variation of spare parts reservation over time based on the spare parts reservation configuration parameters and the variation of spare parts usage over time; obtains new spare parts procurement data and historical spare parts acceptance data, and generates data on the variation of spare parts order quantity over time based on the new procurement data, historical spare parts acceptance data and the variation of spare parts reserve quantity over time; generates spare parts inventory forecasting model based on the variation of spare parts usage over time, the variation of spare parts reserve quantity over time and the variation of spare parts order quantity over time. By comprehensively considering the impact of spare parts usage, reservation, acceptance and regular order on inventory, accurate spare parts inventory forecasting can be achieved.

Description

Technical Field

[0001] The present application relates to the field of big data technology, and in particular to a spare parts inventory prediction method, apparatus, computer equipment, storage medium, and computer program product. Background Art

[0002] In daily production life, large equipment and systems often need to be repaired and maintained.

[0003] Taking nuclear power plant systems as an example, maintenance work can be mainly divided into two categories: planned maintenance and emergency maintenance. Planned maintenance usually involves reserving the spare parts needed for maintenance activities in advance. Spare parts required for planned maintenance can also be further divided into Class A spare parts and Class B spare parts. Class A spare parts are spare parts that must be replaced during maintenance activities, while Class B spare parts are spare parts that are selectively replaced during maintenance activities based on the status of on-site equipment. Emergency maintenance generally refers to maintenance activities that cannot be estimated in advance. Based on spare parts classification, the demand for different types of spare parts is predicted, which further predicts the spare parts inventory and optimizes spare parts inventory parameters. While ensuring on-site spare parts maintenance needs, spare parts inventory data is optimized to reduce inventory backlogs.

[0004] Currently, conventional spare parts inventory forecasts are generally based on the experience of operation and maintenance personnel or warehouse managers. This method of forecasting based solely on manual experience is easily affected by human experience factors, resulting in inaccurate final forecast results, which is not conducive to the operation and maintenance of large-scale equipment and systems. Summary of the Invention

[0005] Based on this, it is necessary to provide a method, device, computer equipment, storage medium and computer program product that can accurately predict spare parts inventory in order to solve the above technical problems.

[0006] In a first aspect, the present application provides a spare parts inventory forecasting method. The method comprises:

[0007] Generate data on the change in spare parts usage over time based on preset maintenance specification data and historical spare parts usage;

[0008] Based on the historical reserve quantity of spare parts, obtain the spare parts reserve configuration parameters, and generate the spare parts reserve quantity change data over time based on the spare parts reserve configuration parameters and the spare parts use quantity change data over time;

[0009] Obtain new spare parts procurement data and historical spare parts acceptance data, and generate spare parts order quantity change data over time based on the new procurement data, historical spare parts acceptance data, and spare parts reserve quantity change data over time;

[0010] A spare parts inventory forecasting model is generated based on the data on the changes in the spare parts usage quantity, the spare parts reservation quantity and the spare parts regular order quantity over time.

[0011] In a second aspect, the present application also provides a spare parts inventory prediction device. The device includes:

[0012] The requisition module is used to generate data on the change in spare parts requisition quantity over time based on preset maintenance specification data and the historical requisition quantity of spare parts;

[0013] The reservation module is used to obtain spare part reservation configuration parameters based on the historical spare part reservation quantity, and generate spare part reservation quantity change data over time based on the spare part reservation configuration parameters and the spare part use quantity change data over time;

[0014] The order module is used to obtain the new purchase data of spare parts and the historical acceptance data of spare parts. Based on the new purchase data, the historical acceptance data of spare parts and the change data of the spare parts reserve quantity over time, it generates the change data of the spare parts order quantity over time.

[0015] The model building module is used to generate a spare parts inventory forecasting model based on the data of spare parts' usage quantity changing over time, the data of spare parts' reservation quantity changing over time, and the data of spare parts' regular order quantity changing over time.

[0016] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the following steps are performed:

[0017] Generate data on the change in spare parts usage over time based on preset maintenance specification data and historical spare parts usage;

[0018] Based on the historical reserve quantity of spare parts, obtain the spare parts reserve configuration parameters, and generate the spare parts reserve quantity change data over time based on the spare parts reserve configuration parameters and the spare parts use quantity change data over time;

[0019] Obtain new spare parts procurement data and historical spare parts acceptance data, and generate spare parts order quantity change data over time based on the new procurement data, historical spare parts acceptance data, and spare parts reserve quantity change data over time;

[0020] A spare parts inventory forecasting model is generated based on the data on the changes in the spare parts usage quantity, the spare parts reservation quantity and the spare parts regular order quantity over time.

[0021] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the following steps:

[0022] Generate data on the change in spare parts usage over time based on preset maintenance specification data and historical spare parts usage;

[0023] Based on the historical reserve quantity of spare parts, obtain the spare parts reserve configuration parameters, and generate the spare parts reserve quantity change data over time based on the spare parts reserve configuration parameters and the spare parts use quantity change data over time;

[0024] Obtain new spare parts procurement data and historical spare parts acceptance data, and generate spare parts order quantity change data over time based on the new procurement data, historical spare parts acceptance data, and spare parts reserve quantity change data over time;

[0025] A spare parts inventory forecasting model is generated based on the data on the changes in the spare parts usage quantity, the spare parts reservation quantity and the spare parts regular order quantity over time.

[0026] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the following steps:

[0027] Generate data on the change in spare parts usage over time based on preset maintenance specification data and historical spare parts usage;

[0028] Based on the historical reserve quantity of spare parts, obtain the spare parts reserve configuration parameters, and generate the spare parts reserve quantity change data over time based on the spare parts reserve configuration parameters and the spare parts use quantity change data over time;

[0029] Obtain new spare parts procurement data and historical spare parts acceptance data, and generate spare parts order quantity change data over time based on the new procurement data, historical spare parts acceptance data, and spare parts reserve quantity change data over time;

[0030] A spare parts inventory forecasting model is generated based on the data on the changes in the spare parts usage quantity, the spare parts reservation quantity and the spare parts regular order quantity over time.

[0031] The above-mentioned spare parts inventory prediction method, device, computer equipment, storage medium and computer program product generate data on the change of spare parts usage quantity over time based on preset maintenance specification data and historical spare parts usage quantity; obtain spare parts reservation configuration parameters based on historical spare parts reservation quantity, and generate data on the change of spare parts reservation quantity over time based on the spare parts reservation configuration parameters and the change of spare parts usage quantity over time; obtain new spare parts procurement data and historical spare parts acceptance data, and generate data on the change of spare parts regular order quantity over time based on the new procurement data, historical spare parts acceptance data and the change of spare parts reservation quantity over time; generate a spare parts inventory prediction model based on the change of spare parts usage quantity over time, the change of spare parts reservation quantity over time and the change of spare parts regular order quantity over time. During the whole process, based on the historical spare parts data, we gradually obtain the data of spare parts usage changes over time, reservation changes over time, and regular order changes over time, and generate a spare parts inventory forecast model based on these data. Since the impact of spare parts usage, reservation, acceptance and regular order inventory is comprehensively considered, accurate spare parts inventory forecast can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 2. It is an application environment diagram of a spare parts inventory prediction method in one embodiment;

[0033] Figure 2 1 is a flow chart of a spare parts inventory forecasting method according to an embodiment;

[0034] Figure 3 The following is a schematic diagram of commonly used MRP strategies and their operation logic;

[0035] Figure 4 is a flow chart of a spare parts inventory forecasting method according to another embodiment;

[0036] Figure 5 This is a schematic diagram of the simulation results of total inventory and availability rate;

[0037] Figure 6 The inventory amount and availability rate under different inventory models;

[0038] Figure 7 This is the grid distribution diagram of the model effect;

[0039] Figure 8 is a structural block diagram of a spare parts inventory prediction device in one embodiment;

[0040] Figure 9 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0042] The spare parts inventory forecasting method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, the terminal 102 communicates with the server 104 via a network. The data storage system can store data that the server 104 needs to process. The data storage system can be integrated on the server 104 or placed on the cloud or other network servers. The server 104 may specifically be a server for spare parts inventory management, which stores data related to spare parts inventory changes in historical records. The terminal 102 initiates a spare parts storage prediction request to the server 104. The server 104 generates data on spare parts usage changes over time based on preset maintenance specification data and the historical spare parts usage; obtains spare parts reservation configuration parameters based on the historical spare parts reservation, and generates data on spare parts reservation changes over time based on the spare parts reservation configuration parameters and the spare parts usage changes over time; obtains new spare parts procurement data and spare parts historical acceptance data, and generates spare parts regular order quantity changes over time based on the new procurement data, spare parts historical acceptance data and spare parts reservation change data; generates spare parts inventory prediction model based on spare parts usage changes over time, spare parts reservation changes over time and spare parts regular order quantity changes over time. Terminal 102 may include, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices may include smart speakers, smart TVs, smart air conditioners, and smart car devices. Portable wearable devices may include smart watches, smart bracelets, and head-mounted devices. Server 104 may be implemented as a standalone server or a server cluster consisting of multiple servers.

[0043] like Figure 2 As shown, the present application provides a spare parts inventory forecasting method, the method comprising:

[0044] S200: Generate data on the change in spare parts usage over time based on preset maintenance specification data and the historical usage of spare parts.

[0045] Pre-set maintenance specification data includes industry-specification data such as maintenance outlines and maintenance standard packages. This data allows for understanding the maintenance specifications and standards for large-scale equipment and systems, and thus the types and quantities of spare parts required under normal operation and maintenance conditions. Historical spare parts usage refers to the types and quantities of spare parts used during historical operation and maintenance. By analyzing this data, patterns in spare parts usage can be identified, generating data on spare parts usage over time. In practical applications, spare parts demand includes planned Class A demand, planned Class B demand, and sudden demand. Planned Class A demand refers to spare parts that must be replaced according to the maintenance plan; planned Class B demand refers to spare parts that are selectively replaced according to the maintenance plan; and sudden demand refers to spare parts required for sudden events (failures) during the operation and maintenance process. Aggregating these three types of demand yields data on spare parts usage over time. Specifically, this data on spare parts usage over time can be presented as a future spare parts usage form, displaying spare parts usage through a form.

[0046] S400: Obtain spare part reservation configuration parameters based on historical spare part reservation quantities, and generate spare part reservation quantity change data over time based on the spare part reservation configuration parameters and spare part use quantity change data over time.

[0047] Reservation is the use requirement proposed by the user (operation and maintenance manager) for the future use of spare parts. Specifically, there are three factors that affect reservation: reservation lead time, reservation ratio and reservation accuracy. Therefore, the reservation configuration parameters can specifically include data of these three dimensions. Since reservation is proposed based on use, the data on the change of spare parts reservation quantity over time can be obtained based on the above-mentioned spare parts reservation configuration parameters and the data on the change of spare parts use over time obtained in S200. Furthermore, different types of spare parts may have different reservation lead times, reservation ratios, and reservation accuracy rates. In actual applications, the setting values of the three factors of reservation can be refined according to the spare parts type, or it can be simplified to set the same setting values for all spare parts.

[0048] S600: Acquire new purchase data of spare parts and historical acceptance data of spare parts, and generate data on changes in the order quantity of spare parts over time based on the new purchase data, historical acceptance data of spare parts and changes in the reserved quantity of spare parts over time.

[0049] New purchase data refers to purchases that increase spare parts demand in the current period. This purchase data will affect future spare parts orders. Furthermore, orders are related to acceptance and reserve quantities. By analyzing new purchase data, historical acceptance data, and the evolution of spare parts reserve quantities over time, we can generate data on spare parts orders over time.

[0050] Furthermore, the revision of spare parts is related to the new revision (NPR), the acceptance quantity (Y), and the revision value of the previous period. The new revision (NPR) is related to the spare parts inventory (S), the reserve quantity (R), the inventory parameter (M), and the revision (PR). By simulating the operation logic of MRP, the NPR can be calculated using the formula. Common MRP strategies and their operation logic are as follows: Figure 3 As shown in the table, the logic is converted into a formula, and the newly triggered purchase quantity is calculated by inputting inventory parameters, inventory quantity, regular order quantity, and reserved quantity. The calculation formula for spare parts regular order is PR t =PR t-1 +NPR t-1 -Y t-1 .

[0051] S800: Generate a spare parts inventory forecast model based on the data of spare parts' usage quantity changing over time, the data of spare parts' reservation quantity changing over time, and the data of spare parts' regular order quantity changing over time.

[0052] After obtaining data on the time-varying spare parts usage, spare parts reserve, and spare parts order quantity through the above processing, a spare parts inventory forecasting model can be generated based on this data. This spare parts inventory forecasting model can predict future spare parts inventory parameters. Spare parts inventory parameters can specifically include spare parts inventory quantity, which can be further converted to obtain parameters such as inventory value and availability rate. Furthermore, these spare parts inventory parameters can be further optimized to ensure that the resulting spare parts inventory quantity does not result in long-term inefficient use of spare parts and funds while still meeting the spare parts needs required for daily operations and maintenance.

[0053] Specifically, the inventory of spare parts is related to the inventory, acceptance quantity and issuance quantity of the previous period. The formula for calculating the inventory of spare parts is S t =S t-1 +Y t-1 -L t-1 When the inventory quantity is less than 0, the use of the product in this period cannot be effectively guaranteed. When the inventory quantity is greater than or equal to 0, the use of the product in this period can be effectively guaranteed. Assuming that guaranteed is 1 and unguaranteed is 0, and the availability rate table is A, then if S t ≥0, then A t =1, otherwise A t = 0. To calculate the availability rate of the spare part within the specified time, just divide the number of items in Table A with a value of 1 by the total number in Table A. The inventory amount is the inventory quantity * spare part unit price. When the inventory quantity is less than 0, the inventory amount is 0. The inventory amount table is SA and the spare part unit price is SP. If S t ≥0, then SA t =S t *SP, otherwise SA t= 0. To calculate the average inventory value of the spare part within the specified time, simply use the values in the SA table to find the average value.

[0054] The above-mentioned spare parts inventory forecasting method generates data on the change of spare parts usage over time based on preset maintenance specification data and the historical spare parts usage; obtains spare parts reservation configuration parameters based on the historical spare parts reservation, and generates data on the change of spare parts reservation over time based on the spare parts reservation configuration parameters and the change of spare parts usage over time; obtains new spare parts procurement data and historical spare parts acceptance data, and generates data on the change of spare parts regular order quantity over time based on the new procurement data, historical spare parts acceptance data and the change of spare parts reservation over time; generates spare parts inventory forecasting model based on the change of spare parts usage over time, the change of spare parts reservation over time and the change of spare parts regular order quantity over time. During the whole process, based on the historical spare parts data, we gradually obtain the data of spare parts usage changes over time, reservation changes over time, and regular order changes over time, and generate a spare parts inventory forecast model based on these data. Since the impact of spare parts usage, reservation, acceptance and regular order inventory is comprehensively considered, accurate spare parts inventory forecast can be achieved.

[0055] In one embodiment, the spare parts inventory forecast further includes: obtaining an initial spare parts reserve quantity at an initial moment; decomposing the initial spare parts reserve quantity into future use quantities to obtain data on changes in the initial use quantity of spare parts over time;

[0056] Generating data on the variation of spare parts usage quantity over time based on preset maintenance specification data and historical spare parts usage quantity includes: generating data on the variation of initial spare parts usage quantity over time based on preset maintenance specification data, historical spare parts usage quantity and data on the variation of initial spare parts usage quantity over time.

[0057] When simulating and predicting spare parts quantities, you can access data such as the current actual inventory quantity, inventory parameters, orders, and reservations. Because reservations are directly involved in spare part inventory calculations, to ensure the stability of spare parts simulations, current reservations must be broken down into future procurement data. Furthermore, for orders, current orders must be broken down into future acceptance data.

[0058] Furthermore, decomposition can be performed in two ways: detailed and simplified. Both methods are described below. Detailed decomposition requires obtaining the specific arrival time of the current order, the type of the current reservation, and the work order start date. If detailed spare parts delivery schedules and detailed reservation lists are unavailable, a simplified method can be used to decompose the order and reservation data.

[0059] In one embodiment, the initial spare parts reservation quantity is decomposed into the future use quantity, and the obtained data of the initial spare parts use quantity changing over time includes:

[0060] Extract the reserved lead time and reserved quantity from the initial spare parts reservation; determine the order of decomposition to different reservation time points based on the reserved lead time; determine the collection quantity to different future time points based on the order and reserved quantity of decomposition to obtain the time-varying data of the initial collection quantity of spare parts.

[0061] The reservation lead time refers to the time period that requires advance reservations. For example, if you need to reserve spare parts for the next 12 months, the reservation lead time is 12 months. Similarly, if you need to reserve spare parts for the next 6 months, the reservation lead time is 6 months. The reserved quantity refers to the initial total reserved quantity. The order in which these reserved quantities are allocated to different reservation time points is determined based on the reservation lead time. Generally, the reservation time points are selected in descending order from the median of the reservation time period. For example, if the reservation lead time is 12 months, June is selected as the starting month for the allocation. The reservation time points are then allocated in descending order, i.e., June-July-May-August-April-September-March-October-February-November-January-December, until the entire reserved quantity has been allocated. For example, if the required reserved quantity is 19, and a monthly average allocation is adopted, the allocated quantity needs to be allocated to 10 monthly periods. The dual-use quantity and reduced reserved quantity after the allocation are shown in Table 1 below.

[0062] Table 1 is a table showing the data of initializing the use and reducing the reservation based on the initial spare parts reservation.

[0063]

[0064] Furthermore, a similar method is used for the decomposition of corrections. The following describes the process of simple decomposition of initial corrections.

[0065] If it is not possible to obtain the detailed delivery plan and detailed reservation list of spare parts, the order and reservation data can be decomposed by a simplified method. For the order PO, the quantity to be delivered N is obtained, and the acceptance quantity is allocated within the procurement cycle to reduce the large fluctuation of inventory in the early stage of simulation due to the initialization of the order. The acceptance quantity for each period is That is, the quantity to be delivered divided by the purchase cycle (rounded up), Time (round up) starts to calculate the acceptance quantity, according to Decompose into the acceptance table until the decomposition of the quantity to be delivered N is completed. For example, the procurement cycle of a spare part is 7 months, the number of POs to be delivered is 6, and the acceptance quantity in each period is 1. If it is 4, it will be allocated according to April, May, March, June, February, and July 2021. One will be accepted in each period, and a total of 6 will be accepted. Similarly, if the number of POs to be delivered is 19, the number of accepted ones in each period is 3, and it will be allocated according to April, May, March, June, February, July, and January 2021. Among them, in January 2021, there is only 1 quota to be allocated, so one is allocated in January 2021, and a total of 19 are accepted.

[0066] Furthermore, in addition to the simple decomposition method, a detailed decomposition method can also be adopted. The following will describe the detailed decomposition method for the initial positive order and the initial standby.

[0067] Extract the positive order (usually called PO) of the currently officially signed order. If the arrival date is before the current date, it will arrive after the planned arrival date + 0.5 * procurement cycle. If the planned arrival date + 0.5 * procurement cycle is still less than the current date, it will arrive at the current date + 1. Combine the arrival date in the order and the order procurement quantity to initialize the data in the acceptance form. Assume that the procurement cycle of a certain spare part is 12 months, there are 6 POs, and the current date is January 2021, as shown in Table 2 below:

[0068] Table 2 shows the POs of the signed orders, their arrival dates and quantities

[0069]

[0070] The initialized acceptance data table is shown in Table 3.

[0071] Table 3 is the initialized acceptance form based on the existing POs

[0072]

[0073] Among them, the acceptance quantity in April 2021 in Table 3 is 7, which is the sum of the arrival quantities in October 2020 and April 2021 in Table 3. The planned arrival date is POT, the current date is NDT, the arrival quantity is n, and the procurement cycle is ΔT. The calculation logic for initializing the acceptance form is as follows: If POT i + 0.5 * ΔT < NDT, then Y2 = n i ; If POT i < NDT, then If POT i > NDT, then Y 0.5*ΔT [[ID=3\6]]= n i .

[0074] For the initial reservation data, extract the currently valid reservation data, i.e., reservations from work order reservations with a start date after the current date. Based on this reservation data, initialize the requisition data, reduce the reservations, and reduce the reservation value. If the reservations involve planned maintenance items, remove the spare parts requirements for the corresponding maintenance items during the requisition form review. As shown in Table 4, the spare parts requirements for maintenance items 4002, 4003, and 4004 are already reflected in the reservations, so the spare parts requirements for these maintenance items need to be removed when reviewing the requisition form.

[0075] Table 4 summarizes the existing reserved data.

[0076]

[0077] Based on the existing reservation data, the collection data, reservation reduction, and initial reservation are initialized. Since the demand date for reservation 101 in Table 4 is before the current date, it is an invalid reservation and is not retained. The remaining reservations are converted into collection data and reservation reduction data, as shown in Table 5. Since the current date is January 2021, the simulation starts from the current date. Therefore, the reservations for January 2021 are the valid reservations in Table 5, totaling 9. Reservations for subsequent periods require iterative calculations using reservations, new reservations, and reductions.

[0078] Table 5 shows the initialization, reduction and sorting of reservations based on existing reservation data.

[0079] time Quantity of use Add reservation Reduce reservation Reserved quantity January 2021 0 0 0 9 May 2021 2 0 2 June 2021 3 0 3 July 2021 1 0 1 October 2021 2 0 2 January 2022 1 0 1

[0080] like Figure 4 As shown, in one embodiment, S200 includes:

[0081] S220: Obtaining preset maintenance specification data and historical usage of spare parts within adjacent preset time periods;

[0082] S240: Classify the historical usage of spare parts within adjacent preset time periods into the usage of planned Class A spare parts, the usage of planned Class B spare parts, and the usage of emergency spare parts. Class A spare parts are spare parts that must be replaced by operation and maintenance, and Class B spare parts are spare parts that are selectively replaced by operation and maintenance.

[0083] S260: Generate a Class A spare parts demand model based on the preset maintenance specification data and the planned Class A spare parts usage. Generate a Class B spare parts demand model based on the preset maintenance specification data and the planned Class B spare parts usage. Generate the future emergency spare parts usage using a linear fitting method based on the preset maintenance specification data.

[0084] S280: Generate data on the change in spare parts usage over time based on the Class A spare parts demand model, the Class B spare parts demand model, and the future usage of emergency spare parts.

[0085] The adjacent preset time period refers to the time period adjacent to the current time, such as the historical usage of spare parts in the last 3 years, the last 5 years, or the last 10 years. Specifically, calculating the future usage of spare parts is to predict the future demand for spare parts. The simplest solution is to assume that the future demand for spare parts is consistent with the historical demand. Assuming that 10 years of simulation data is needed, the demand can be sorted out as follows. Export the usage data of the spare parts in the past 10 years from the database. Assume that the usage data for the past 10 years are N1, N2, N3, ..., N 10 , thus generating the data of 11 to 20 years of use, the calculation formula is This generates the spare parts demand for the next 10 years. To improve the effectiveness of the forecast, historical spare parts usage data can be categorized. For planned Class A spare parts demand, the future planned Class A spare parts demand can be generated based on the maintenance program and standard package. For planned Class B spare parts demand, the future planned Class B spare parts demand can be generated based on the maintenance program, standard package, and planned Class B spare parts demand model. For sudden spare parts demand, the above linear fitting method can be used to generate the future spare parts demand, or by reading the historical sudden spare parts usage data for the spare part, a random sudden spare parts demand with the same frequency, mean, and standard deviation can be generated. The planned Class A spare parts demand, planned Class B spare parts demand, and sudden spare parts demand can be summed to obtain the total future spare parts demand.

[0086] In one embodiment, obtaining spare part reservation configuration parameters according to the historical reservation amount of spare parts includes: obtaining a reservation lead time, a reservation ratio, and a reservation accuracy rate according to the historical reservation amount of spare parts.

[0087] The following will introduce in detail the method for determining the three reserved factors and give an example to illustrate the complete calculation process.

[0088] The reservation lead time specifies how far in advance a spare part is reserved. For example, for planned maintenance (PM) projects, NPPs issue work orders and reserve spare parts approximately one year in advance (Daya Bay issues work orders 2C in advance). For routine PM projects, NPPs also issue work orders and reserve spare parts approximately one year in advance. For emergency projects, according to NPP management regulations, projects involving nuclear safety must be completed within 24 hours; relatively important projects must be completed within 3 or 12 weeks; and general projects must be completed by the next FEG (Functional Equipment Group). Therefore, different reservation lead times can be set based on the criticality of the spare part. For critical spare parts (such as CCM and H-class spare parts), the emergency reservation period can be set at one month, while for other spare parts, it can be set at three months. If the criticality of the spare part is not differentiated, the emergency reservation period can be set at two months.

[0089] The reservation ratio refers to the proportion of spare parts that are reserved. The vast majority of planned Category A spares are reserved before an overhaul, while only a small number of planned Category B spares are reserved in advance. Unexpected spares are generally reserved only before they are issued. An analysis of overhaul issuance data from six rounds of overhauls at a nuclear power plant shows that the reservation ratio for planned Category A spares during the lead time for overhauls is approximately 93.3%, while the reservation ratio for planned Category B spares during the lead time for overhauls is approximately 24.5%. Given the current situation at this nuclear power plant, the timely submission rate for routine spare parts requests is 10% lower than that for overhauls. Therefore, the reservation ratio for routine planned Category A spares during the lead time can be set at 83.3%. Since the current reservation ratio for planned Category B spares is low, the reservation ratio for routine planned Category B spares can be set to the same as for overhauls, i.e., 24.5%. After analyzing the emergency spare parts requisition work orders, the overall reservation ratio of emergency spare parts can be set to 79.6%.

[0090] Reservation accuracy refers to the proportion of correctly predicted reserved spare parts. Analysis of procurement data from six rounds of overhauls at a nuclear power plant revealed that the reservation accuracy for Category A overhaul spare parts was 87.4%, and the accuracy for Category B overhaul spare parts was 75.6%. For routine planned spare parts, the accuracy of demand for Category A / B planned spare parts can be set at 64%. Comparing and analyzing reservation and procurement data for emergency spare parts, the emergency reservation accuracy can be set at 64.3%. This means that for every spare part reserved, 0.643 parts are actually procured. If reservations are calculated based on procurement, 1.55 parts are reserved for every spare part procured (a value that can be increased to 2). Because the reservation lead time for emergency spare parts is much shorter than that for planned spare parts, the overall reservation accuracy for emergency spare parts is relatively high. Table 6 summarizes the reservation lead time, reservation ratio, and reservation accuracy.

[0091] Table 6 sets the reference values for the reserved influencing factors

[0092]

[0093] Based on the spare parts requisition table and the defined reservation influencing factors, spare parts reservation data can be calculated. The following example illustrates how to convert a requisition table into a reservation table. To more clearly illustrate the logical relationship of the conversion, the reservation influencing factors are simplified. Assume that there are only two types of requisitions: planned and unplanned. The planned reservation lead time is 2 months, the reservation ratio is 100%, and the reservation accuracy is 100%. The unplanned reservation lead time is 1 month, the reservation ratio is 50%, and the reservation accuracy is 50%. As shown in Table 7, first sort the spare parts requisition data and identify the requisition type. Then, a random number sequence of 0-1 is added based on the reservation ratio. If the reservation ratio is 50%, the ratio of 0 to 1 is 50%. Iteratively calculate the "New Reservation" and "Reduction Reservation" columns. At time t2, there is one sudden request, and the sudden reservation ratio random number is 1. Therefore, at time t1, there will be two reservations (one request, 50% reservation accuracy, and two reservations). This reservation will disappear at any time when the spare part is requested at t2, so the reservation is reduced by two at t2. The sudden request of one at t8 and the planned request of three at t9 will both be reserved at t7, so the reservation at t7 is 5 (2 + 3). Reservation data = previous period's reservation + current period's new reservation - current period's reduction reservation. This can be used to calculate the spare part reservation data.

[0094] If converted to a program flow, set the quantity to be taken (L), reserve the lead time (ΔN i ), reserved ratio (K i ), the reserved proportion random number is (f(K), a random number between 0 and 1, the probability of 0 is 1-K, and the probability of 1 is K), the reserved accuracy (P), the newly added reserved (NR), the reduced reserved (DR), and the reserved (R). Iterative operation, IF L t *f(k) t >0,Then Where t is the time dimension and i is the different types of spare parts (sudden, planned). First, iterate to add and reduce the reserved data, and then use R t =R t-1 +NR t -DR t Calculate the reserved quantity.

[0095] Table 7 Case study of converting withdrawal data into reserved data

[0096]

[0097] In one embodiment, after generating the spare parts inventory forecasting model based on the data of the variation of the spare parts usage quantity, the variation of the spare parts reservation quantity, and the variation of the spare parts order quantity over time, the further step includes:

[0098] The parameters of the spare parts inventory of this period are obtained according to the spare parts inventory forecasting model; the parameters of the spare parts inventory of this period are optimized using a preset parameter optimization method, which includes an exhaustive method, an improved genetic algorithm or a demand probability model optimization method.

[0099] Based on the spare parts inventory forecasting model, accurate spare parts inventory parameters for this period can be obtained. Furthermore, the obtained spare parts inventory parameters for this period can be optimized to find a suitable balance between the total inventory amount of spare parts and the total availability rate. Specifically, according to the simulation operation logic, based on the data of the change of spare parts' usage quantity over time, the change of spare parts' reservation quantity over time, and the change of spare parts' regular order quantity over time, the spare parts inventory forecasting model is constructed and then optimized. Figure 5 Taking spare parts as an example, the simulation is carried out. For two sets of inventory parameter schemes, the total inventory amount and total availability rate are calculated, as shown in Figure 6 As shown, the average inventory amount of the TYP1 group inventory model is 10.94 million yuan, and the average in-stock rate is 97.3%; the average inventory amount of the TYPE2 group inventory model is 11.21 million yuan, and the average in-stock rate is 96.4%.

[0100] By setting spare parts inventory parameters and reserving a certain amount of inventory, we can ensure future demand for spare parts without creating inventory backlogs. The effectiveness of spare parts inventory parameter settings is directly related to the amount of spare parts in stock and their availability. Following the logic of inventory forecasting analysis, spare parts inventory parameters can be treated as variables. A fitness function can be set based on the inventory amount and availability rate, transforming the problem into an optimization problem to find the optimal solution for the inventory parameters.

[0101] Evaluation and Optimization Method of Spare Parts Inventory Forecasting Model

[0102] Currently, commonly used demand forecast accuracy testing methods primarily compare the degree of deviation between two models from the true value. However, since the spare parts inventory strategy models currently used in large-scale systems (such as nuclear power plants) rely on setting inventory parameters and utilizing methods such as inventory buffers to respond to spare parts demand, it is difficult to calculate the exact demand quantity at a specific point in the future based on inventory parameters. Therefore, this evaluation method is not suitable for evaluating current inventory strategy models. To evaluate emergency spare parts inventory strategy models, a comprehensive evaluation can be conducted based on the effectiveness of inventory management and spare parts supply after the model parameters are set.

[0103] To evaluate an inventory model, the average inventory value, average availability, average annual purchase frequency, and the baseline reserve value for spare parts under the model must be calculated. The average inventory value and average availability are derived using the above information. Future inventory and availability are predicted for spare parts under certain inventory parameters to calculate the average inventory value and average availability. The availability and inventory value are mutually constrained indicators. Generally, improving the availability requires increasing inventory reserves, which in turn increases the inventory value. After calculating the availability and inventory value, a grid distribution chart of the model's performance is constructed, with the average inventory value as the horizontal axis and the average availability as the vertical axis. The baseline inventory value (represented by a cell on the horizontal axis) is calculated as the average quantity purchased during the procurement cycle multiplied by the unit price, and the baseline availability rate is set at 7% (represented by a cell on the vertical axis; a range of 72% to 100% covers most availability scenarios). Evaluation is performed using the "distance method," which measures the distance between each model and the theoretical optimal value. The closer to the upper left of the figure (i.e. the lower the parameter reserve amount, the higher the stock availability rate), the better the model effect. The grid distribution diagram of the model effect is as follows Figure 7 As shown in the figure, the X distance is calculated first, then the Y distance, and finally the total distance from the theoretical optimal position is calculated based on the X distance and Y distance. The smaller the total distance, the closer the model is to the theoretical optimal value and the better it is.

[0104] For example, the benchmark inventory amount of a spare part is 6000. By calculating the average availability rate and average inventory amount of each inventory strategy model, and then calculating the X-axis distance and Y-axis distance of the model effect grid distribution diagram, the evaluation score of the model is calculated. The following is an example of Model 1. The X-axis distance is the parameter reserve amount of Model 1 / the benchmark parameter reserve amount, that is, 8000 / 6000 = 1.33, and the Y-axis distance is (1-85%) / 7% = 2.14. The total distance is the square of the X-axis distance plus the square of the Y-axis distance, and then the square root is taken, that is Similarly, the total distances of Model 2 to Model 6 can be obtained as shown in Table 8. Among them, the total distance of Model 3 is the smallest, that is, it is closest to the theoretical optimal position, so the effect of Model 3 is relatively better.

[0105] Table 8 Evaluation scores of a spare part in different inventory strategy models

[0106]

[0107] The evaluation method for the spare parts inventory forecasting model has been clarified and can be converted into a fitness function. A smaller fitness function value indicates a better inventory model. In practical applications, the average annual number of purchases, inventory support correction factors, and purchase frequency correction factors can be added to tailor the current status and management requirements of different nuclear power plants, expanding the method's applicability.

[0108] The fitness function has the following structure: its inputs include average availability, average inventory value, baseline inventory value, average annual purchase frequency, inventory guarantee correction factor, and purchase frequency correction factor; its output includes a comprehensive score (total relative distance). The average availability, average inventory value, and average annual purchase frequency represent the future average availability, average inventory value, and average annual purchase frequency of the spare part calculated by the inventory forecasting platform under the specified inventory parameters. The baseline inventory value is the average future inventory value calculated by the inventory forecasting platform based on the inventory parameters set as ZB + EX (only the minimum inventory value is set). The minimum inventory value is the average quantity of the spare part used during the purchase cycle. This value serves as the baseline inventory value for the spare part. The inventory guarantee correction factor ranges from 0 to 1. If it approaches 0, controlled inventory is prioritized (low inventory value), while if it approaches 1, spare part guarantee is prioritized (high guarantee rate). The purchase frequency correction coefficient takes a value between 0 and 1. If it tends to 0, the annual average number of purchase orders is strictly controlled (the inventory amount may be high). If it tends to 1, the constraint on the number of purchase orders is weaker (the inventory amount may be low).

[0109] After establishing the spare parts inventory forecasting model and fitness evaluation model, the optimal spare parts inventory parameters can be calculated using relevant optimization algorithms. The spare parts inventory forecasting model reads the spare parts inventory parameters, performs operations, and outputs the average inventory amount, average availability rate, and average annual purchase frequency. These are then passed to the fitness evaluation model. The fitness function algorithm calculates the inventory parameter evaluation score and transmits it to the optimization algorithm model in the optimization algorithm model library. New inventory parameters are then calculated and passed to the spare parts inventory forecasting model. Through continuous iterative operations, the optimal spare parts inventory parameters are calculated. The following details the operational logic of the exhaustive method and genetic algorithm in the optimization algorithm model library as examples.

[0110] Spare parts inventory parameter optimization method based on exhaustive method

[0111] Inventory parameters usually set minimum inventory and maximum inventory (fixed batch). Therefore, by setting different minimum inventory values, maximum inventory values, and fixed batch values, through a series of combinations, a series of potentially optimal inventory parameter sets can be constructed. For example, the number of spare parts required during the procurement cycle is x. According to experience, the value range of the minimum inventory ZB in the spare parts inventory parameters is 0.1x to 3x, the value range of the maximum inventory HB is 0.5x to 3x (only the procurement batch, the maximum inventory value is the minimum inventory + procurement batch), and the value range of the fixed batch FX is 0.5 to 3x. Therefore, the calculation quantity of the program can be calculated as 18x 2+3x, when x is 5, the number of operations is 495; when x is 50, the number of operations is 45150; and when x is 100, the number of operations is 180300. This shows that if an exhaustive search is performed for all potential parameters, the number of operations increases dramatically quadratically with the increase in the quantity used. Therefore, it is necessary to optimize the exhaustive search method to find the optimal solution. Based on the quantity used, if the quantity used is large, the interval between each potential value is larger; if the quantity used is small, the interval between each potential value is smaller. Based on experience, the optimal value for minimum inventory is generally near the average quantity used during the procurement cycle, and the optimal value for the purchase batch is generally near the average annual quantity used. Therefore, near the potential optimal value, the search density can be increased and the intervals can be reduced.When the average annual quantity k is less than 7, the benchmark interval number of the reorder point is 1, and an exhaustive solution is performed between 1 and 5k; when the average annual quantity k is between 7 and 15, the benchmark interval number of the reorder point is 2, which is divided into three sections. The first section is between 1 and k, with an interval of 1, the second section is between k and 2k, with an interval of 2, and the third section is between 2k and 2.7k, with an interval of 3; when the average annual quantity k is between 15 and 30, the benchmark interval number of the reorder point is 3, which is divided into four sections. The first section is between 1 and k / 6, with an interval of 3, and the second section is between k / 6 and 2k / 3 , the interval is 2, the third section is 2k / 3~5k / 3, the interval is 3, the fourth section is 5k / 3~2.4k, the interval is 4; the average annual quantity k is between 30~101, the benchmark interval quantity of reorder point is steps=4+k / 20, divided into four sections, the first section is between 1~0.5*k / (1.2*steps), the interval is 1.2*steps, the second section is 0.5*k / (1.2*steps)~0.5*k / (1.2*steps)+(k / 2) / (0.5*steps)), the interval is 0. 6*steps, the third section is between 0.5*k / (1.2*steps)+(k / 2) / (0.5*steps)) and 0.5*k / (1.2*steps)+(k / 2) / (0.5*steps)+k / (1.2*steps)), with an interval of 1.2*steps, and the fourth section is between 0.5*k / (1.2*steps)+(k / 2) / (0.5*steps)+k / (1.2*steps)) and 0.5*k / (1.2*steps)+(k / 2) / ( 0.5*steps) + k / (1.2*steps) + 1.5*k / (steps+1), with an interval of 2*steps. When the average annual quantity taken is k > 101, the base reorder point interval is steps = 4*k / 40, divided into three segments: the first segment is from 1 to 0.5*k / steps, with an interval of steps; the second segment is from 0.5*k / steps to 2.5*k / steps, with an interval of 0.4*steps; and the third segment is from 2.5*k / steps to 40, with an interval of 2*steps. The potential optimal values for maximum inventory and fixed lot sizes can be decomposed by referring to the reorder point method.

[0112] By combining different MRP types and using an exhaustive method to calculate the comprehensive scores of various inventory parameters, we can find the optimal inventory parameters. For example, for a spare part at a nuclear power plant, the current inventory parameters are ZB+HB, the reorder point is 115, the maximum inventory is 500, the average inventory value is 241, the average stock availability is 88.3%, and the comprehensive score is 4.22. The optimal inventory parameters are ZB+FX, with a reorder point of 123, a fixed lot size of 116, an average inventory value of 107, a stock availability rate of 98.3%, and a comprehensive score of 1.75.

[0113] Spare parts inventory parameter optimization method based on improved genetic algorithm

[0114] Genetic algorithms are global probabilistic search algorithms based on the laws of natural selection and can be used to solve nonlinear optimization problems. The implementation of a genetic algorithm primarily involves population initialization, numerical conversion, fitness calculation, replication, swapping, and mutation. Current nuclear power spare parts inventory strategies primarily include: ZB+EX (only minimum inventory), ZB+HB (maximum and minimum inventory), ZB+FX (minimum inventory with a fixed purchase quantity upon triggering a purchase), and PD+EX (no minimum or maximum inventory). Since PD+EX triggers purchases through reservations and does not proactively reserve inventory, requiring no inventory parameters to be set, this scenario is not currently considered in inventory parameter optimization. When using a genetic algorithm to calculate optimal spare parts parameters, the optimal parameters for each of the three scenarios, ZB+EX, ZB+HB, and ZB+FX, are calculated. The resulting composite scores for the optimal parameters are then compared, and the inventory parameters for the scenario with the highest composite score are selected.

[0115] The following further introduces the calculation logic of the genetic algorithm.

[0116] S1 is to initialize the population. The population size can be set to 40, and the dimension is 1 or 2 (1 dimension in the case of ZB+EX, and 2 dimensions in the cases of ZB+HB and ZB+FX). The individual initialization position is randomly assigned between ±50% of the standard reorder point, standard maximum inventory, and standard fixed batch. The standard reorder point is the number of parts collected during the procurement cycle, the standard maximum inventory is the number of parts collected during the procurement cycle + the average annual number of parts collected, and the standard fixed batch is the average annual number of parts collected. During initialization, the maximum value of the individual position is limited to 10 to 15 times the average annual number of spare parts collected, that is, the optimal value of the individual is calculated between 1 and the maximum value.

[0117] S2. For numerical conversion, two arrays are created, one of which stores the original decimal position data and the other stores the normalized binary position data. The original decimal position data is used to calculate the fitness value of the inventory parameter, and the binary position data is used to participate in the negative value, crossover, and mutation of the genetic algorithm. The method for converting the original decimal position to the binary position is as follows: It normalizes the decimal position data and converts it to between 0 and 1. The formula can be used Where x is the decimal position of the sample after normalization, X is the decimal position of the sample before normalization, and X min is the minimum value among all sample positions, which can be set to 1. max The maximum value of all sample positions can be set to 15 times the average annual number of spare parts. Convert the decimal value of 0 to 1 to the binary value. The method of converting the binary position to the original decimal position is as follows: convert the binary value to the decimal value of 0 to 1. If the value is less than 0, convert it to 0. If the value is greater than 1, convert it to 1 (in the process of genetic algorithm crossover and mutation, it may cause abnormal conditions greater than 1 or less than 0). Convert the decimal position of 0 to 1 to the original decimal position, and you can use the formula X=(X max -X min )*x+X min .

[0118] S3. To calculate the fitness value, the genetic algorithm passes the location of each sample (spare parts inventory parameter) to the spare parts inventory prediction model. The spare parts inventory parameter evaluation score is calculated by the inventory prediction model and the fitness evaluation model, and the fitness value is passed to the genetic algorithm.

[0119] S4 is sample replication. Based on each sample's fitness value (spare parts inventory parameter evaluation score), samples with good fitness values are more likely to be passed on to the next generation. A "roulette wheel algorithm" can be used to sort samples based on fitness values. Samples with good fitness values have a higher probability of being replicated, while samples with poor fitness values have a lower probability of being replicated.

[0120] S5. For sample crossover, first set the retention ratio of excellent parents. This means that excellent parent samples are screened out and retained directly in the offspring without crossover. This ratio can be set to 0.05. For samples that are not excellent parents, a crossover operation is performed. This means that a sample is randomly crossovered with other samples. Two methods can be used for crossover operation: 1) Select two samples, randomly select half of the nodes in one sample, transfer the values of the corresponding nodes of the other sample to the sample, and perform the crossover operation; 2) Randomly select a node in one sample, transfer the value after the corresponding node in the other sample to the sample, and perform the crossover operation.

[0121] S6 is the mutation of the sample, that is, each node in each sample has a certain probability of mutation, changing from 0 to 1 or from 1 to 0, and the probability of mutation can be selected as 0.005.

[0122] S7 is a new random factor that effectively prevents the algorithm from falling into local extremes. Four samples are randomly selected and reassigned. The position of the first sample is assigned a random number; the positions of the second and third samples are randomly assigned within ±50% of the optimal position; and the position of the fourth sample is directly assigned to the optimal position. This ensures that the optimal fitness value after each iteration is no higher than the optimal fitness value before the iteration.

[0123] S8 is to output the individual with the best fitness value. When the number of iterations reaches the set threshold, the individual with the best fitness value, i.e. the inventory parameters of the spare parts, is directly output.

[0124] Spare parts inventory parameter optimization method based on demand probability model

[0125] For some nuclear power spare parts, due to the low historical usage and low future demand, it's difficult to predict spare parts inventory using a spare parts inventory model. Therefore, a method for determining a demand probability function can be used. By analyzing the historical usage characteristics of spare parts and assuming they follow a certain probability distribution (currently binomial and Poisson distributions are used), the spare parts inventory parameters are calculated by setting a target availability rate. First, the historical usage of the spare parts is analyzed. If the spare parts are high-turnover spare parts (used for two or more years within three years or three or more years within five years), an inventory parameter optimization model based on spare parts inventory forecasting is used. If the spare parts are not high-turnover spare parts, an inventory parameter optimization model based on spare parts demand probability is used.

[0126] Assuming that the demand characteristics of a batch of spare parts conform to a certain probability density, we need to use sample data for parameter estimation and hypothesis testing. By establishing a distribution function library and using methods such as the Kolmogorov-Smirnoff test to determine whether spare parts procurement conforms to the corresponding probability distribution, we can select an appropriate spare parts demand function model library. The distribution function library can be configured with distribution functions such as Poisson, Normal, Binomial, Gamma, Lognorm, and Weibull. The following uses the Poisson distribution function as an example to further explain the spare parts inventory parameter optimization method.

[0127] Poisson distribution refers to the specific probability of an event occurring within a certain period of time. It is a discrete probability distribution commonly used in statistics and probability. Assuming there are n devices on site and the probability of any device failing is a, then within a certain period of time, the average number of failures is λ = an, and the probability of m devices failing is Assuming that the set reserve quantity is s and the spare parts procurement cycle is t days, the out-of-stock probability of the Poisson distribution is: The availability rate is: Therefore, you only need to set the spare part availability rate, extract the spare part procurement cycle, and the quantity issued within the procurement cycle to calculate the required inventory quantity and set the spare part inventory parameters. In practice, in a Poisson distribution model, if spare parts are divided into two levels (CCM / non-CCM), the availability rate of CCM spare parts can be set to 99.5%, and the availability rate of non-CCM spare parts can be set to 90%. If spare parts are divided into three levels (H / M / L), the availability rate of H-level spare parts can be set to 99%, the availability rate of M-level spare parts to 95%, and the availability rate of L-level spare parts to 90%. For example, a spare part is a non-CCM spare part with a procurement cycle of one year. The quantities used in the past five years were 0, 2, 0, 0, and 2, respectively. After the KS test, the use data conforms to the Poisson distribution (the significance level of the KS test is set to 0.05). For non-CCM spare parts, the availability rate is set to 90%. After calculation, when 2 inventory items are reserved, the 90% availability rate requirement is met. Therefore, the inventory parameter of this spare part is ZB+EX, and the reorder point is 2.

[0128] For some non-high-turnover spare parts, the KS test using their procurement data failed. The demand functions in the demand function library all failed the test, so the demand probability model cannot be used to calculate inventory parameters. For these spare parts, the inventory parameters can be set to ZB+EX, with the reorder point being the procurement quantity within the procurement cycle. Based on the importance of the spare part, a certain amount of safety stock can be added.

[0129] It should be understood that, although the steps in the flowcharts of the above embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts of the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily to be performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0130] Based on the same inventive concept, embodiments of the present application also provide a spare parts inventory prediction device for implementing the aforementioned spare parts inventory prediction method. The solution provided by this device is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more of the following embodiments of the spare parts inventory prediction device can be found in the aforementioned limitations of the spare parts inventory prediction method and will not be further elaborated here.

[0131] In addition, if Figure 8 As shown, the present application also provides a spare parts inventory prediction device, the device comprising:

[0132] The requisition module 200 is used to generate the time-varying data of the requisition quantity of spare parts based on the preset maintenance specification data and the historical requisition quantity of spare parts;

[0133] The reservation module 400 is used to obtain spare part reservation configuration parameters based on the historical spare part reservation quantity, and generate spare part reservation quantity change data over time based on the spare part reservation configuration parameters and the spare part use quantity change data over time;

[0134] The ordering module 600 is used to obtain new spare parts procurement data and historical spare parts acceptance data, and generate spare parts order quantity change data over time based on the new procurement data, historical spare parts acceptance data, and spare parts reserve quantity change data over time;

[0135] The model building module 800 is used to generate a spare parts inventory forecasting model based on the data of spare parts' usage quantity changing over time, the data of spare parts' reservation quantity changing over time, and the data of spare parts' regular order quantity changing over time.

[0136] In one embodiment, the above-mentioned spare parts inventory prediction device also includes: an initial decomposition module, which is used to obtain the initial spare parts reserve quantity at the initial moment; decompose the initial spare parts reserve quantity into the future collection quantity to obtain the data of the initial collection quantity of spare parts changing over time; the collection module is also used to generate the data of the initial collection quantity of spare parts changing over time based on the preset maintenance specification data, the historical collection quantity of spare parts and the data of the initial collection quantity of spare parts changing over time.

[0137] In one embodiment, the initial decomposition module is also used to extract the reservation advance period and reservation quantity from the initial spare parts reservation; determine the order of decomposition to different reservation time points based on the reservation advance period; determine the use quantity decomposed to different future time points based on the order of decomposition of different reservation time points and the reservation quantity, and obtain the data on the change of the initial use quantity of spare parts over time.

[0138] In one embodiment, the requisition module is also used to obtain preset maintenance specification data and the historical requisition quantity of spare parts in adjacent preset time periods; classify the historical requisition quantity of spare parts in adjacent preset time periods into the requisition quantity of planned Class A spare parts, the requisition quantity of planned Class B spare parts and the requisition quantity of emergency spare parts, Class A spare parts refer to spare parts that must be replaced by operation and maintenance, and Class B spare parts refer to spare parts that are selectively replaced by operation and maintenance; generate a Class A spare parts demand model based on the preset maintenance specification data and the requisition quantity of planned Class A spare parts, generate a Class B spare parts demand model based on the preset maintenance specification data and the requisition quantity of planned Class B spare parts, and generate the future requisition quantity of emergency spare parts using a linear fitting method based on the preset maintenance specification data; generate data on the change of spare parts requisition quantity over time based on the Class A spare parts demand model, the Class B spare parts demand model and the future requisition quantity of emergency spare parts.

[0139] In one embodiment, the reservation module is further configured to obtain a reservation lead time, a reservation ratio, and a reservation accuracy rate based on a historical reservation quantity of spare parts.

[0140] In one embodiment, the above-mentioned spare parts inventory prediction device also includes: an optimization module, which is used to obtain the spare parts inventory parameters of this period based on the spare parts inventory prediction model; and optimize the spare parts inventory parameters of this period using a preset parameter optimization method, and the preset parameter optimization method includes an exhaustive method, an improved genetic algorithm or a demand probability model optimization method.

[0141] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 9 As shown. The computer device includes a processor, a memory, and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store historical spare parts inventory-related data. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, it implements a spare parts inventory forecasting method.

[0142] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0143] Generate data on the change in spare parts usage over time based on preset maintenance specification data and historical spare parts usage;

[0144] Based on the historical reserve quantity of spare parts, obtain the spare parts reserve configuration parameters, and generate the spare parts reserve quantity change data over time based on the spare parts reserve configuration parameters and the spare parts use quantity change data over time;

[0145] Obtain new spare parts procurement data and historical spare parts acceptance data, and generate spare parts order quantity change data over time based on the new procurement data, historical spare parts acceptance data, and spare parts reserve quantity change data over time;

[0146] A spare parts inventory forecasting model is generated based on the data on the changes in the spare parts usage quantity, the spare parts reservation quantity and the spare parts regular order quantity over time.

[0147] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0148] Generate data on the change in spare parts usage over time based on preset maintenance specification data and historical spare parts usage;

[0149] Based on the historical reserve quantity of spare parts, obtain the spare parts reserve configuration parameters, and generate the spare parts reserve quantity change data over time based on the spare parts reserve configuration parameters and the spare parts use quantity change data over time;

[0150] Obtain new spare parts procurement data and historical spare parts acceptance data, and generate spare parts order quantity change data over time based on the new procurement data, historical spare parts acceptance data, and spare parts reserve quantity change data over time;

[0151] A spare parts inventory forecasting model is generated based on the data on the changes in the spare parts usage quantity, the spare parts reservation quantity and the spare parts regular order quantity over time.

[0152] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0153] Generate data on the change in spare parts usage over time based on preset maintenance specification data and historical spare parts usage;

[0154] Based on the historical reserve quantity of spare parts, obtain the spare parts reserve configuration parameters, and generate the spare parts reserve quantity change data over time based on the spare parts reserve configuration parameters and the spare parts use quantity change data over time;

[0155] Obtain new spare parts procurement data and historical spare parts acceptance data, and generate spare parts order quantity change data over time based on the new procurement data, historical spare parts acceptance data, and spare parts reserve quantity change data over time;

[0156] A spare parts inventory forecasting model is generated based on the data on the changes in the spare parts usage quantity, the spare parts reservation quantity and the spare parts regular order quantity over time.

[0157] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0158] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0159] The above embodiments merely illustrate several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art may make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A spare parts inventory forecasting method, characterized in that: The method comprises: Generate data on the change in spare parts usage over time based on preset maintenance specification data and historical spare parts usage; According to the historical reservation quantity of spare parts, the spare parts reservation configuration parameters are obtained, and according to the spare parts reservation configuration parameters and the time-varying data of the spare parts' use quantity, the spare parts reservation quantity change data over time is generated; Acquire new purchase data of spare parts and historical acceptance data of spare parts, and generate data on changes in the spare parts' reserved quantity over time based on the new purchase data, the historical acceptance data of spare parts, and the data on changes in the spare parts' reserved quantity over time; Generate a spare parts inventory forecasting model based on the data of the spare parts' usage quantity, the spare parts' reservation quantity and the spare parts' regular order quantity. Before generating the data on the variation of spare parts usage over time based on the preset maintenance specification data and the historical usage of spare parts, the method further includes: obtaining an initial spare parts reservation at an initial moment; extracting a reservation lead period and a reserved quantity from the initial spare parts reservation; determining an order of decomposing the reservations into different reservation time points based on the reservation lead period; and determining the usage quantities decomposed into different future time points based on the order of decomposing the different reservation time points and the reserved quantity, thereby obtaining the data on the variation of the initial spare parts usage over time. Generating the data on the variation of spare parts usage over time based on the preset maintenance specification data and the historical usage of spare parts includes: generating the data on the variation of spare parts usage over time based on the preset maintenance specification data, the historical usage of spare parts and the data on the variation of the initial usage of spare parts over time; The obtaining of spare parts reservation configuration parameters according to the historical spare parts reservation amount includes: obtaining a reservation lead time, a reservation ratio, and a reservation accuracy rate according to the historical spare parts reservation amount.

2. The method according to claim 1, characterized in that After generating the spare parts inventory forecast model based on the data of the spare parts usage quantity changing over time, the spare parts reservation quantity changing over time, and the spare parts regular order quantity changing over time, the method further includes: Obtaining spare parts inventory parameters for this period according to the spare parts inventory prediction model; The parameters of the current spare parts library are optimized using a preset parameter optimization method, wherein the preset parameter optimization method includes an exhaustive method, an improved genetic algorithm, or a demand probability model optimization method.

3. The method according to claim 1, characterized in that The preset maintenance specification data includes data corresponding to the maintenance outline and the maintenance standard package.

4. The method according to claim 1, wherein The data on the change of spare parts usage quantity over time includes the spare parts usage form in the future.

5. A spare parts inventory forecasting device, characterized in that: The device comprises: The requisition module is used to generate data on the change in spare parts requisition quantity over time based on preset maintenance specification data and the historical requisition quantity of spare parts; A reservation module is used to obtain spare part reservation configuration parameters based on the historical spare part reservation quantity, and generate spare part reservation quantity change data over time based on the spare part reservation configuration parameters and the spare part use quantity change data over time; The ordering module is used to obtain new purchase data of spare parts and historical acceptance data of spare parts, and generate time-varying data of the order quantity of spare parts based on the new purchase data, the historical acceptance data of spare parts and the time-varying data of the reserved quantity of spare parts; A model building module is used to generate a spare parts inventory forecasting model based on the data of the spare parts' usage quantity, the spare parts' reservation quantity and the spare parts' regular order quantity. The initial decomposition module is used to obtain the initial spare parts reservation quantity at the initial moment; extract the reservation lead time and reservation quantity from the initial spare parts reservation quantity; determine the order of decomposition to different reservation time points based on the reservation lead time; and determine the allocation quantity to different future time points based on the allocation order and reservation quantity of different reservation time points, thereby obtaining the time-varying data of the initial allocation quantity of spare parts. The requisition module is also used to generate the time-varying data of the requisition quantity of spare parts based on the preset maintenance specification data, the historical requisition quantity of spare parts and the time-varying data of the initial requisition quantity of spare parts; The reservation module is also used to obtain the reservation lead time, reservation ratio and reservation accuracy based on the historical reservation quantity of spare parts.

6. The device according to claim 5, characterized in that The device also includes an optimization module for obtaining the current period spare parts warehouse parameters according to the spare parts inventory prediction model; and optimizing the current period spare parts warehouse parameters using a preset parameter optimization method, wherein the preset parameter optimization method includes an exhaustive method, an improved genetic algorithm, or a demand probability model optimization method.

7. The device according to claim 5, characterized in that The preset maintenance specification data includes data corresponding to the maintenance outline and the maintenance standard package.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 4 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

Images