Inventory control method, inventory control device, electronic equipment and storage medium
By constructing decision trees and distributed fuzzy sets in nuclear power systems, and dynamically predicting spare parts demand variables, the problem of insufficient flexibility of traditional inventory management methods in dynamic environments is solved, thus achieving the safe and stable operation of nuclear power plants.
Patent Information
- Application Number
- CN202511758102.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-27
- Publication Date
- 2026-02-13
AI Technical Summary
Traditional inventory management methods are difficult to adapt to the dynamic and changing demand environment of nuclear power systems, resulting in poor inventory control flexibility and a tendency for stockouts or backlogs, which cannot meet the high safety and stability requirements of nuclear power plants.
By constructing decision trees and distributed fuzzy sets, the spare parts demand variables are predicted based on historical datasets, target inventory order quantities are generated, inventory strategies are dynamically adjusted, and nonparametric methods such as kernel estimation are used to model the demand distribution. Considering the complex relationships between multidimensional features, distributed fuzzy sets are constructed to characterize the uncertainty of demand distribution.
It improves the flexibility of inventory control, reduces operational risks, ensures the safe and stable operation of nuclear power plants, avoids shortages or stockpiles, and meets the high safety and stability requirements of nuclear power plants.
Smart Images
Figure CN121526490A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of inventory management technology, and in particular to an inventory control method, inventory control device, electronic device and storage medium. Background Technology
[0002] In the operation and maintenance of nuclear power systems, the inventory management of equipment spare parts is crucial for ensuring system safety and improving equipment availability. Nuclear power systems are highly complex and have extremely high reliability requirements; the demand for equipment spare parts is influenced by various factors, such as equipment operating status, historical failure frequency, seasonal variations, and maintenance cycles. However, traditional inventory management methods often rely on empirical judgment or static statistical models, such as the fixed safety stock method and the historical average method. These methods are ill-suited to dynamically changing demand environments. Summary of the Invention
[0003] The main objective of this application is to provide an inventory control method, inventory control device, electronic device, and storage medium, which aims to improve the dynamic flexibility of inventory control.
[0004] To achieve the above objectives, a first aspect of this application proposes an inventory control method, the method comprising: Obtain historical datasets of nuclear power plants; wherein, the historical datasets include historical samples of the nuclear power plants during historical operating periods, and the historical samples include historical demand-driven characteristics and historical spare parts demand during the historical operating periods; A decision tree is constructed based on the historical demand-driven characteristics and the historical spare parts demand, and the historical leaf nodes of the historical demand-driven characteristics are determined through the decision tree. The target demand-driven characteristics of the nuclear power plant during the target operating period are obtained, and spare parts demand variables for the target operating period are generated; wherein, the target operating period is located after the historical operating period; The target leaf node of the target demand-driven feature is determined by the decision tree; Based on the historical leaf nodes, the target leaf nodes, the historical spare parts demand, and the spare parts demand variables, construct an estimated reference distribution for the spare parts demand variables; Construct a distributed fuzzy set based on the estimated reference distribution; The target inventory order quantity for the nuclear power plant is calculated based on the spare parts demand variable, the distributed fuzzy set, and the preset expected inventory order quantity. The expected inventory order quantity is updated based on the target inventory order quantity.
[0005] In some embodiments, the historical leaf node has a historical index value, the target leaf node has a target index value, and the step of constructing an estimated reference distribution of the spare parts demand variable based on the historical leaf node, the target leaf node, the historical spare parts demand quantity, and the spare parts demand variable includes: The first quantity is obtained by acquiring the number of historical samples in which the historical index value and the target index value are equal and the historical spare parts demand is less than or equal to the spare parts demand variable. Obtain the number of historical samples whose historical index value and target index value are equal, and obtain the second number; The estimated reference distribution is constructed based on the first quantity and the second quantity.
[0006] In some embodiments, constructing a distributed fuzzy set based on the estimated reference distribution includes: Generate a candidate distribution of the spare parts demand variables and determine the fuzzy radius; Calculate the degree of difference between the candidate distribution and the estimated reference distribution; The candidate distributions are filtered based on the difference degree and the fuzzy radius to obtain the target distribution; Construct the distribution fuzzy set based on the target distribution.
[0007] In some embodiments, calculating the target inventory order quantity for the nuclear power plant based on the spare parts demand variable, the distributed fuzzy set, and the preset expected inventory order quantity includes: Calculate the stockout cost based on the spare parts demand variables and the expected inventory order quantity; Calculate the backlog cost based on the spare parts demand variables and the expected inventory order quantity; Obtain the boundary distribution of the fuzzy distribution set; The target inventory order quantity is calculated based on the stockout cost, the backlog cost, and the boundary distribution.
[0008] In some embodiments, the distributed fuzzy set includes multiple target distributions, and obtaining the boundary distribution of the distributed fuzzy set includes: Compare each of the target distributions with the estimated reference distribution; If the target distribution is less than or equal to the estimated reference distribution, then the target distribution is taken as the first distribution, and the smallest of the first distributions is selected as the lower bound distribution. If the target distribution is greater than the estimated reference distribution, then the target distribution is taken as the second distribution, and the largest second distribution is selected as the upper bound distribution; The boundary distribution is determined based on the lower bound distribution and the upper bound distribution.
[0009] In some embodiments, calculating the target inventory order quantity based on the stockout cost, the overstock cost, and the boundary distribution includes: Calculate the critical ratio based on the stockout cost and the backlog cost; Calculate the critical quantile based on the boundary distribution and the critical ratio; The target inventory order quantity is calculated based on the stockout cost, the backlog cost, and the critical quantile.
[0010] In some embodiments, the boundary distribution includes a lower bound distribution and an upper bound distribution, and the calculation of the critical quantile based on the boundary distribution and the critical ratio includes: If the lower bound distribution is greater than or equal to the critical ratio, then the spare parts demand variable corresponding to the lower bound distribution is obtained to obtain the first variable; The lower bound quantile is obtained by selecting the first variable with the smallest value. If the upper bound distribution is greater than or equal to the critical ratio, then the spare parts demand variable corresponding to the upper bound distribution is obtained to obtain the second variable; The upper bound quantile is obtained by selecting the smallest second variable; The critical quantile is determined based on the lower bound quantile and the upper bound quantile.
[0011] To achieve the above objectives, a second aspect of this application provides an inventory control device, the device comprising: The first acquisition module is used to acquire the historical dataset of the nuclear power plant; wherein, the historical dataset includes historical samples of the nuclear power plant during historical operating periods, and the historical samples include historical demand-driven characteristics and historical spare parts demand during the historical operating periods; The first construction module is used to construct a decision tree based on the historical demand-driven features and the historical spare parts demand, and to determine the historical leaf nodes of the historical demand-driven features through the decision tree. The second acquisition module is used to acquire the target demand-driven characteristics of the nuclear power plant during the target operating period and generate spare parts demand variables for the target operating period; wherein the target operating period is located after the historical operating period; The determination module is used to determine the target leaf node of the target demand-driven feature through the decision tree; The second construction module is used to construct an estimated reference distribution of the spare parts demand variable based on the historical leaf node, the target leaf node, the historical spare parts demand quantity, and the spare parts demand variable; The third construction module is used to construct a distributed fuzzy set based on the estimated reference distribution; The calculation module is used to calculate the target inventory order quantity of the nuclear power plant based on the spare parts demand variable, the distributed fuzzy set, and the preset expected inventory order quantity. An update module is used to update the expected inventory order quantity based on the target inventory order quantity.
[0012] To achieve the above objectives, a third aspect of the present application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the method of the first aspect described above.
[0013] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of the first aspect described above.
[0014] This application provides an inventory control method, inventory control device, electronic device, and storage medium. It acquires historical datasets from a nuclear power plant and constructs a decision tree based on these datasets. The historical datasets contain historical demand-driven features and historical spare parts demand quantities for historical operating periods. A decision tree is constructed based on these historical demand-driven features and historical spare parts demand quantities to divide the entire feature space into multiple complementary and overlapping local feature regions. Each leaf node of the decision tree corresponds to a local feature region, allowing the establishment of a mapping relationship between any input feature vector and the leaf node. By determining the historical leaf node of the historical demand-driven features through the decision tree, a unique leaf node with a mapping relationship to the historical demand-driven features can be obtained. The target demand-driven features of the nuclear power plant during the target operating period are acquired to predict spare parts demand and control inventory based on these target demand-driven features. Affected by seasonal changes, maintenance cycles, and other factors, the spare parts demand of a nuclear power plant is dynamically changing. To enable the inventory control strategy to adapt to the dynamically changing demand environment, spare parts demand variables for the target operating period are generated to simulate the dynamic characteristics of demand changes based on these variables. By determining the target leaf nodes of the target demand-driven features using a decision tree, leaf nodes that have a mapping relationship with the target demand-driven features can be obtained. The demand distribution of spare parts demand variables can then be estimated based on these leaf nodes. A fixed demand distribution lacks adaptability to dynamic demand changes, leading to poor inventory strategy flexibility and increasing the risk of inventory backlogs or stockouts, making it difficult to meet the high safety and stability requirements of nuclear power plants. To improve the flexibility of inventory control and ensure the safe and stable operation of nuclear power plants, an estimated reference distribution of spare parts demand variables is constructed based on historical leaf nodes, target leaf nodes, historical spare parts demand quantities, and spare parts demand variables. A distributional fuzzy set is then constructed based on this estimated reference distribution to characterize the uncertainty of the demand distribution. Based on the spare parts demand variables, the distributional fuzzy set, and the preset expected inventory order quantity, the target inventory order quantity for the nuclear power plant is calculated to determine the corresponding inventory order quantity for the spare parts demand variables. The expected inventory order quantity is updated based on the target inventory order quantity to dynamically adjust the inventory order quantity according to demand changes, improving the flexibility of inventory control and thus achieving the safe and stable operation of the nuclear power plant. Attached Figure Description
[0015] Figure 1 This is a flowchart of the inventory control method provided in the embodiments of this application; Figure 2 yes Figure 1 The flowchart of step S150 in the middle; Figure 3 yes Figure 1 The flowchart of step S160 in the process; Figure 4 yes Figure 1 The flowchart of step S170 in the process; Figure 5 yes Figure 4 The flowchart of step S430 in the middle; Figure 6 yes Figure 4 The flowchart of step S440 in the middle; Figure 7 yes Figure 6 The flowchart of step S620 in the process; Figure 8 This is a schematic diagram of the structure of the inventory control device provided in the embodiments of this application; Figure 9 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation
[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.
[0017] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.
[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.
[0019] In the operation and maintenance of nuclear power systems, the inventory management of equipment spare parts is crucial for ensuring system safety and improving equipment availability. Nuclear power systems are highly complex and have extremely high reliability requirements. The operating environment is highly uncertain, and the demand for equipment spare parts is affected by various factors, such as seasonal load changes and sudden adjustments to operating conditions, which can cause drastic fluctuations in spare parts demand. However, traditional inventory management models often rely on empirical judgment or static statistical models, such as the fixed safety stock method and the historical average method. These methods are ill-suited to the dynamically changing and uncertain demand environment.
[0020] Based on this, embodiments of this application provide an inventory control method, an inventory control device, an electronic device, and a computer-readable storage medium. Based on JW fuzzy sets, the worst-case risk caused by the demand distribution estimation error is fully considered when constructing the inventory strategy, thereby generating a more robust inventory strategy. This enables nuclear power plants to maintain a reasonable inventory level when facing sudden demand fluctuations, avoid shortages or backlogs, reduce operational risks, and improve the dynamic flexibility of inventory control.
[0021] The inventory control method, inventory control device, electronic device, and computer-readable storage medium provided in this application are specifically described through the following embodiments. First, the inventory control method in the embodiments of this application is described.
[0022] The inventory control method provided in this application relates to the fields of inventory management technology and nuclear power plant equipment management and supply chain optimization technology. The inventory control method provided in this application can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application implementing the inventory control method, but is not limited to the above forms.
[0023] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.
[0024] Figure 1 This is an optional flowchart of an inventory control method provided in an embodiment of this application. Figure 1The method may include, but is not limited to, steps S110 to S180.
[0025] Step S110: Obtain the historical dataset of the nuclear power plant; wherein, the historical dataset includes historical samples of the nuclear power plant during historical operating periods, and the historical samples include historical demand-driven characteristics and historical spare parts demand during historical operating periods; Step S120: Construct a decision tree based on historical demand-driven characteristics and historical spare parts demand, and determine the historical leaf nodes of historical demand-driven characteristics through the decision tree; Step S130: Obtain the target demand-driven characteristics of the nuclear power plant during the target operating period, and generate spare parts demand variables for the target operating period; wherein, the target operating period is located after the historical operating period; Step S140: Determine the target leaf node of the target demand-driven feature through the decision tree; Step S150: Construct an estimated reference distribution for the spare parts demand variables based on historical leaf nodes, target leaf nodes, historical spare parts demand, and spare parts demand variables. Step S160: Construct a distributed fuzzy set based on the estimated reference distribution; Step S170: Calculate the target inventory order quantity for the nuclear power plant based on the spare parts demand variables, the distributed fuzzy set, and the preset expected inventory order quantity. Step S180: Update the expected inventory order quantity based on the target inventory order quantity.
[0026] In step S110 of some embodiments, data information generated during the operation of the nuclear power plant is acquired, and a historical dataset is constructed to model a demand prediction model based on the historical dataset. The historical dataset includes historical samples of the nuclear power plant during various historical operating periods. These historical samples include historical demand-driving characteristics and historical spare parts demand during those periods. The historical demand-driving characteristics are multi-dimensional features related to equipment spare parts demand generated during the operation of the nuclear power plant, and the historical spare parts demand is the consumption of equipment spare parts during those periods. The historical demand-driving characteristics include equipment operating status, equipment failure frequency, maintenance cycle, seasonal factors, environmental parameters, equipment aging level, operating load level, and static covariates. Among these, equipment operating status can refer to the working status of equipment such as main pumps, steam generators, and cooling systems; equipment failure frequency refers to the number of times equipment failures have occurred during historical operating periods; maintenance cycle includes the time of the most recent maintenance and the planned time of the next maintenance; monthly and seasonal changes can affect the operating conditions of nuclear power plant equipment, and seasonal factors can include months and seasons; environmental parameters can include temperature, pressure, and humidity; equipment aging can include the equipment's service life and cumulative operating hours; operating load level includes unit power level and load fluctuation level; static covariates include demand planning type (uncertain demand planning), batch type, reorder point, maximum inventory level, fixed batch size, consumable labels, importance rating, average receiving quantity, and average receiving interval.
[0027] Depend on The historical dataset constructed from each historical runtime segment is represented as follows: ,in Indicates the first Historical samples of a historical runtime period Indicates the first Historical spare parts demand for a historical sample Indicates the first Historical demand-driven characteristics of each historical sample.
[0028] In step S120 of some embodiments, the demand for equipment spare parts is affected by various characteristic factors, and the relationship between the demand for nuclear power equipment spare parts and these characteristic factors often exhibits nonlinear and high-dimensional characteristics. Traditional methods struggle to capture this complex mapping relationship. To accurately predict the demand for equipment spare parts, this application embodiment uses the Classification and Regression Tree (CART) algorithm to construct a decision tree based on each historical demand-driving feature and the corresponding historical spare parts demand in the historical dataset. This decision tree captures the complex nonlinear relationship between the demand for nuclear power equipment spare parts and multidimensional operational characteristics. Compared to traditional linear regression and historical averaging methods, decision trees have stronger fitting ability and generalization performance when facing high-dimensional and heterogeneous features, thereby improving the accuracy of demand prediction.
[0029] The process of constructing a decision tree is as follows: Construct an empty binary decision tree, where the root node's sample set includes every historical sample from the historical dataset. For the set of samples contained in the current node. The feature space is recursively divided until a stopping condition is met, resulting in a decision tree. The stopping condition can be that the number of leaf nodes reaches a preset number, or the number of historical samples in the sample set contained in the current node is less than a preset threshold.
[0030] Historical demand-driven features include multiple sub-features. During recursive feature space partitioning, each sub-feature and all possible split points are traversed. The mean squared error of splitting the current node into two child nodes at the split point is calculated, and the sub-feature with the smallest mean squared error and the split point are selected. Based on the selected sub-feature and split point, the current node is split into left and right nodes. The sample set of the left node includes historical samples where the feature value of the selected sub-feature is less than or equal to that of the selected split point, and the sample set of the right node includes historical samples where the feature value of the selected sub-feature is greater than that of the selected split point.
[0031] After all nodes have completed splitting, all nodes that did not continue splitting are marked as leaf nodes, and each leaf node is numbered to obtain its index value. Each leaf node has a node value, which is the average historical spare parts demand of the sample set it contains. Each leaf node corresponds to a local feature region defined by a series of splitting rules, including the selected sub-features and split points. If the number of leaf nodes is... Then the entire feature space represented by the historical dataset is divided into These are non-overlapping local feature regions. That is: , Where X is the feature space. ; The index value of the leaf node; For the first The local feature region indicated by each leaf node; ∪ represents the empty set; ∪ represents the union; ∩ represents the intersection.
[0032] Decision trees can establish a mapping from any input feature to leaf nodes, thus allowing arbitrary input features to be mapped to leaf nodes. Mapped to a unique leaf node index value That is, to determine the local feature region to which the feature belongs. Specifically, historical demand-driven features are input into a decision tree, and the decision tree maps these features to unique leaf nodes to obtain historical leaf nodes.
[0033] In step S130 of some embodiments, the target demand-driven characteristics of the nuclear power plant during the target operating period are obtained. These target demand-driven characteristics refer to multi-dimensional characteristics related to demand forecasting, and their sub-features can be referenced from sample demand-driven characteristics. The target operating period is the time period following the historical operating period. Equipment spare parts demand values corresponding to the target demand-driven characteristics are randomly generated to obtain the spare parts demand variables for the target operating period, thereby simulating the dynamically changing demand environment during inventory management.
[0034] In step S140 of some embodiments, the target demand-driven features are input into a decision tree, and the decision tree maps the target demand-driven features to unique leaf nodes to obtain target leaf nodes. When performing spare parts demand prediction, the node value of the target leaf node can be used as the spare parts demand quantity or the value of the spare parts demand variable predicted based on the target demand-driven features.
[0035] The task of inventory management is to determine the timing and quantity of replenishment to minimize costs and maximize profits. Existing inventory optimization methods typically rely on historical data to predict demand distribution and formulate inventory strategies based on this prediction. Taking a model-based inventory optimization method as an example, the single-cycle newsboy model periodically reviews the ordering strategy and minimizes the remaining inventory cost (actual demand less than the order quantity) and the stockout penalty cost (actual demand greater than the order quantity) at the review point. The optimal order quantity can be calculated using a fixed formula, expressed as: , Where c is the unit purchase cost; p is the unit stockout penalty cost; and h is the unit remaining inventory cost. It is the inverse function of the demand distribution function.
[0036] However, in practical applications, the demand distribution function is difficult to estimate, especially in the nuclear power sector, where data samples are limited and demand uncertainty is high. This makes existing methods significantly inadequate in terms of demand forecasting accuracy and inventory ordering decisions. When facing sudden operating conditions or when the nuclear power system is in a non-steady-state operation, fixed demand distributions lack adaptability to dynamically changing demand environments, resulting in poor inventory control flexibility and an increased risk of inventory buildup or stockouts. Furthermore, spare parts for critical nuclear power equipment (such as main pump seals and steam generator heat transfer tubes) are highly customized and irreplaceable; their absence directly impacts unit operation, making it difficult to meet the high safety and stability requirements of nuclear power plants.
[0037] Because the demand distribution is uncertain and the distribution function is difficult to estimate, this application starts with historical observation datasets and uses a data-driven approach to find the relationship between the demand distribution and multi-dimensional features. It then uses non-parametric methods such as kernel estimation to predict the demand distribution, obtaining an estimated reference distribution. The estimated reference distribution is the empirical conditional distribution of the spare parts demand variable under given target demand-driven features. If the spare parts demand variable takes discrete values, the estimated reference distribution can be expressed as the cumulative distribution function on the spare parts demand variable.
[0038] Please see Figure 2 In some embodiments, the historical leaf node has a historical index value, which is the index value of the historical leaf node, and the target leaf node has a target index value, which is the index value of the target leaf node. Step S150 may include, but is not limited to, steps S210 to S230: Step S210: Obtain the number of historical samples where the historical index value and the target index value are equal and the historical spare parts demand is less than or equal to the spare parts demand variable, and obtain the first quantity; Step S220: Obtain the number of historical samples whose historical index value is equal to the target index value, and obtain the second number; Step S230: Construct an estimated reference distribution based on the first quantity and the second quantity.
[0039] In step S210 of some embodiments, key equipment in nuclear power plants, such as main pump seals and steam generator heat transfer tubes, are highly customized and irreplaceable. Their operational status is highly dependent on the timely replacement of spare parts; any loss will directly affect unit operation. These devices are typically located in extreme environments such as high temperature, high pressure, and strong radiation. Due to limited maintenance windows, if a malfunction occurs and the required spare parts cannot be obtained in time, it will lead to unplanned shutdowns, seriously affecting unit safety and power generation efficiency. Traditional inventory management models often use historical averaging or fixed safety stock strategies, which are difficult to adapt to dynamically changing demand environments. The embodiments of this application fully consider the unique irreplaceable constraints of nuclear power systems during the modeling process and design corresponding inventory risk control mechanisms. By introducing non-parametric methods such as decision trees and kernel estimation for conditional demand modeling, based on historical observation data... The mapping relationship between features and leaf nodes defined in the constructed decision tree is used to perform non-parametric local kernel density estimation of the conditional distribution of spare parts demand variables, thereby modeling the demand distribution of the newsboy problem containing covariates. For any given target feature vector x, its corresponding leaf node is... Define a kernel weight function of the following form: , in, Features driven by target demand; For the first Historical demand-driven characteristics of a historical sample; For core weights; For historical leaf nodes; The target leaf node; For indicator functions, if ,Right now and If they belong to the same leaf node, the kernel weight is 1; otherwise, it is 0.
[0040] Based on the above kernel weight function, an empirical conditional distribution function for the spare parts demand variable is constructed. The empirical conditional distribution function is expressed as follows: , in, The number of historical samples; For spare parts demand variables; For the first Historical spare parts demand for a historical sample; In a given Down The empirical condition distribution; if Then the indicator function The value is 1 if it is not 0 otherwise; Then the indicator function The value is 1 if it is not 0 otherwise.
[0041] In step S220 of some embodiments, the number of historical samples in the historical dataset whose historical index value and target index value are equal is counted, that is, the number of samples whose historical demand-driven features and target demand-driven features belong to the same leaf node, to obtain a second number.
[0042] In step S230 of some embodiments, the first quantity is used as the numerator and the second quantity is used as the denominator. The ratio between the numerator and the denominator is calculated to obtain the estimated reference distribution.
[0043] This application employs nonparametric methods such as kernel estimation to model the conditional demand distribution. This modeling approach is more flexible and generalizable, overcoming the limitation of traditional linear regression models that can only capture the relationship between features and the demand mean. It can more accurately reflect the complex nonlinear relationship between nuclear power equipment spare parts demand and multidimensional features such as operating status, maintenance cycles, and environmental parameters. Existing technologies generally assume that features only affect the mean of the demand function, i.e., d = f(x) + Where x is the input feature and f is the demand mean function. Here, is the noise term, independent of the input features, and d is the mean demand. This application's embodiments do not make any strong assumptions about the noise structure, thus better adapting to situations where features may affect the overall demand distribution during actual nuclear power system operation.
[0044] Steps S210 to S230 above only consider information from locally similar samples to construct a reference distribution with strong interpretability and adaptability. This distribution can more accurately characterize the complex relationship between spare parts demand and multi-dimensional features such as equipment operating status, maintenance cycle, environmental parameters, and covariates, thereby achieving higher-precision demand forecasting and significantly improving the accuracy of spare parts demand forecasting, the timeliness of spare parts supply, and equipment availability.
[0045] A distributed fuzzy set is constructed based on the estimated reference distribution. This fuzzy set characterizes the uncertainty of the demand distribution, ensuring the solvability of the inventory strategy while effectively controlling the inventory decision risk caused by forecast errors. Furthermore, nonparametric methods such as kernel estimation are combined to model the conditional demand distribution, thereby obtaining a robust inventory strategy with a closed-form under limited data conditions. The construction process of the fuzzy set is described in detail below.
[0046] Please see Figure 3 In some embodiments, step S160 may include, but is not limited to, steps S310 to S340: Step S310: Generate candidate distributions of spare parts demand variables and determine the fuzzy radius; Step S320: Calculate the degree of difference between the candidate distribution and the estimated reference distribution; Step S330: Filter the candidate distributions based on the degree of difference and the fuzzy radius to obtain the target distribution; Step S340: Construct a distributed fuzzy set based on the target distribution.
[0047] In step S310 of some embodiments, a probability distribution of the spare parts demand variable is randomly generated to obtain a candidate distribution. A fuzzy radius is set, which represents the tolerance for deviation from the reference distribution and is used to control the size of the fuzzy set. The fuzzy radius is a non-negative real number, and its value is greater than or equal to 0.
[0048] In step S320 of some embodiments, a distance metric function is determined. This distance metric function measures the degree of difference between the candidate distribution and the estimated reference distribution. Different distance metric functions are suitable for different uncertainty structures and can be flexibly selected according to the actual application scenario, such as the Kullback-Leibler divergence function. Based on the distance metric function, the degree of deviation between other possible demand distributions (candidate distributions) and the estimated reference distribution is measured, forming a quantitative index of distribution uncertainty, the JW difference measure, to obtain the degree of difference.
[0049] The KL divergence function is expressed as: , in, It is a distance metric function; For input parameters.
[0050] For any candidate distribution, the JW difference measure is calculated between the candidate distribution and the estimated reference distribution to obtain the difference. The formula for calculating the JW difference measure is defined as follows: , , in, For spare parts demand variables; To estimate the reference distribution; Candidate distribution; JW difference measure; sup represents the function The upper bound of the spare parts demand variable; and Both are functions Input parameters; This is a distance metric function.
[0051] The JW difference measure reflects any threshold Below, the candidate distribution and the estimated reference distribution divide the random variable into two categories. and The degree of difference between the two parts.
[0052] In step S330 of some embodiments, candidate distributions with a difference degree less than or equal to the fuzzy radius are selected to obtain the target distribution.
[0053] In step S340 of some embodiments, a JW fuzzy set is constructed based on the target distribution to obtain a distribution fuzzy set. The distribution fuzzy set is a set of probability distributions (candidate distributions) that satisfy the condition that the JW difference measure from the estimated reference distribution does not exceed the fuzzy radius.
[0054] The JW fuzzy set defines the set of all distributions that, under the JW difference measure, do not exceed a certain robustness level relative to the estimated reference distribution, as follows: , in, Represents the JW fuzzy set; The fuzzy radius; It is the complete set of spare parts demand variables, containing all possible values; This is the set of probability distributions for spare parts demand variables.
[0055] In steps S310 to S340 above, traditional inventory models often rely on a single reference distribution for decision-making, which is insufficient to address the issues of data sparsity and high demand uncertainty in nuclear power systems. By introducing JW fuzzy sets into the field of nuclear power spare parts inventory management for uncertainty modeling, and using JW fuzzy sets as a tool to characterize demand distribution uncertainty, a set of demand distributions that are similar to the reference distribution under the JW difference measure is constructed. This fuzzy set not only possesses good theoretical properties such as closure and solvability, but also effectively characterizes the distribution uncertainty risk caused by prediction errors, thereby significantly improving the robustness of the inventory strategy and avoiding model errors caused by the assumption of a single distribution.
[0056] When demand distribution is uncertain, inventory ordering is performed based on distributed fuzzy sets to achieve distributed inventory optimization, thereby determining the optimal inventory ordering strategy and obtaining the target inventory ordering quantity for the nuclear power plant. The calculation process for the target inventory ordering quantity is described in detail below.
[0057] Please see Figure 4 In some embodiments, step S170 may include, but is not limited to, steps S410 to S440: Step S410: Calculate the stockout cost based on spare parts demand variables and expected inventory order quantity; Step S420: Calculate the backlog cost based on spare parts demand variables and expected inventory order quantity; Step S430: Obtain the boundary distribution of the distributed fuzzy set; Step S440: Calculate the target inventory order quantity based on stockout costs, overstock costs, and boundary distribution.
[0058] In steps S410 to S420 of some embodiments, stockout cost and overstock cost are predetermined. Stockout cost is the cost incurred when actual demand exceeds the order quantity, and overstock cost is the cost incurred when actual demand is less than the order quantity. To achieve more scientific and secure inventory decisions, stockout cost can be much higher than overstock cost, so as to combine JW fuzzy sets to construct an inventory optimization model suitable for asymmetric loss structures.
[0059] Ordering of spare parts for nuclear power plant equipment is typically done on a single-cycle basis. Only one order is placed within this cycle. Orders are placed with the supplier before the start of the sales cycle, and the order quantity is the expected inventory quantity. Here, the cycle refers to a complete operating cycle of the nuclear power system, the time span from the end of one refueling overhaul to the start of the next, for example, one year.
[0060] Stockout costs and overstock costs can also be calculated separately based on spare parts demand variables and expected inventory order quantities. If the nuclear power plant orders from supplier nodes... The unit's products, The expected inventory order quantity is calculated. The spare parts demand variable is subtracted from the expected inventory order quantity to obtain the intermediate difference. If the intermediate difference is greater than 0, it means the spare parts demand variable is greater than the expected inventory order quantity. In this case, the unit stockout penalty cost is multiplied by the intermediate difference to obtain the stockout cost, and the backlog cost is determined to be 0. The unit stockout penalty cost is the penalty cost incurred due to a lack of one unit of product. If the intermediate difference is less than or equal to 0, it means the spare parts demand variable is less than or equal to the expected inventory order quantity. In this case, the unit inventory holding cost is multiplied by the absolute value of the intermediate difference to obtain the backlog cost, and the stockout cost is determined to be 0. The unit inventory holding cost is the holding cost incurred in storing one unit of product.
[0061] Nuclear power plant operations must balance controlling inventory costs with ensuring equipment availability. Traditional inventory management methods often focus on a single objective (such as minimizing inventory holding costs), neglecting considerations for equipment reliability and operational safety. This application's embodiments incorporate multi-objective optimization principles into the inventory optimization model, seeking a balance between stockout costs and overstock costs. It allows users to flexibly adjust robustness parameters (fuzzy radius) and inventory strategies based on safety levels and budget requirements at different operational stages, resulting in greater adaptability and decision-making flexibility when facing diverse operational objectives.
[0062] In step S430 of some embodiments, the distributed fuzzy set is a JW fuzzy set constructed from the target distributions, including multiple target distributions. The boundary distribution of the distributed fuzzy set is selected from the multiple target distributions.
[0063] In step S440 of some embodiments, in order to obtain the actual order quantity corresponding to the spare parts demand variable, a target inventory order quantity is calculated based on the stockout cost, backlog cost and boundary distribution, and an ordering operation is performed based on the target inventory order quantity.
[0064] Through steps S410 to S440 above, inventory ordering can be carried out in an environment of uncertain demand, thereby improving the flexibility of inventory control.
[0065] Construct the boundary distribution of the distributed fuzzy set, perform first-order stochastic dominance (FSD) analysis on the distributed fuzzy set, derive the lower and upper bound distributions, and obtain the boundary distribution.
[0066] Please see Figure 5 In some embodiments, step S430 may include, but is not limited to, steps S510 to S540: Step S510: Compare each target distribution with the estimated reference distribution; Step S520: If the target distribution is less than or equal to the estimated reference distribution, then the target distribution is taken as the first distribution, and the smallest first distribution is selected as the lower bound distribution. Step S530: If the target distribution is greater than the estimated reference distribution, then the target distribution is taken as the second distribution, and the largest second distribution is selected as the upper bound distribution. Step S540: Determine the boundary distribution based on the lower and upper bound distributions.
[0067] In step S510 of some embodiments, for any spare parts requirement variable R, the upper and lower bounds of the distribution of the fuzzy set are defined as follows: , , , in, Indicates the lower bound distribution; Indicates the upper bound distribution; JW Difference Measure Less than or equal to the fuzzy radius The target distribution; To estimate the reference distribution.
[0068] Based on the above definition formulas for upper and lower bound distributions, compare each target distribution and the estimated reference distribution in the fuzzy distribution set.
[0069] In step S520 of some embodiments, if the target distribution is greater than or equal to 0 and less than or equal to the estimated reference distribution, then the target distribution is taken as the first distribution, that is, the first distribution is located at... The target distribution within the interval is determined, and the first distribution with the smallest value is selected as the lower bound distribution.
[0070] In step S530 of some embodiments, if the target distribution is greater than the estimated reference distribution and less than or equal to 1, then the target distribution is used as the second distribution, that is, the second distribution is located at... The target distribution within the interval is determined, and the second largest distribution is selected as the upper bound distribution.
[0071] In step S540 of some embodiments, the lower bound distribution and the upper bound distribution are used as the boundary distributions of the distributed fuzzy set. The lower bound distribution and the upper bound distribution represent the most conservative and the most optimistic demand distributions within the current uncertainty range, respectively.
[0072] Through steps S510 to S540 above, the upper and lower boundary distributions of the distributed fuzzy set can be obtained, so as to determine the optimal inventory strategy based on the upper and lower boundary distributions.
[0073] Please see Figure 6 In some embodiments, step S440 may include, but is not limited to, steps S610 to S630: Step S610: Calculate the critical ratio based on the stockout cost and the backlog cost; Step S620: Calculate the critical quantiles based on the boundary distribution and critical ratio; Step S630: Calculate the target inventory order quantity based on stockout costs, overstock costs, and critical quantiles.
[0074] In step S610 of some embodiments, the stockout cost and the overstock cost are added together to obtain the intermediate cost, and the ratio between the stockout cost and the intermediate cost is calculated to obtain the critical ratio. The formula for calculating the critical ratio is expressed as: , in, The critical ratio; Costs related to stockouts; This is to cover accumulated costs.
[0075] In step S620 of some embodiments, the boundary distribution includes a lower bound distribution and an upper bound distribution, and the critical quantile includes a lower bound quantile and an upper bound quantile. The lower bound quantile is calculated based on the lower bound distribution and critical ratio of the JW fuzzy set, and the upper bound quantile is calculated based on the upper bound distribution and critical ratio of the JW fuzzy set.
[0076] In step S630 of some embodiments, the optimal robust inventory ordering quantity is calculated based on stockout cost, overstock cost, lower bound quantile, and upper bound quantile to obtain the target inventory ordering quantity. The formula for calculating the target inventory ordering quantity is defined as follows: , in, Order the target inventory quantity; Costs related to stockouts; For accumulated costs; The lower bound quantile; This is the upper bound quantile.
[0077] This closed-form expression for the optimal inventory strategy minimizes the expected loss in the worst case and eliminates the need to solve complex numerical optimization problems. It enables a rapid solution for the optimal inventory strategy, greatly improving computational efficiency.
[0078] Existing inventory optimization methods require solving complex convex optimization problems or rely on numerical simulations, resulting in high computational costs and slow response times. This makes it difficult to meet the real-time, high-frequency inventory adjustment needs of nuclear power systems, especially when dealing with a large variety of spare parts, where algorithm execution time may exceed acceptable limits, restricting practicality. Steps S610 to S630 apply JW fuzzy sets to the sub-Bruker inventory optimization problem. Given shortage and backlog costs, a sub-Bruker optimization model is established based on the constructed JW fuzzy sets, and an optimal inventory strategy expression with a closed-form is derived. This can be applied to demand forecasting and dynamic inventory optimization management of key equipment spare parts during nuclear power plant operation and maintenance. It exhibits stronger robustness while retaining efficient computational capabilities, enabling real-time inventory management response to nuclear power plant needs. The inventory strategy generation time is reduced from minutes to seconds.
[0079] With the support of multi-dimensional features, the embodiments of this application integrate data-driven and distributed optimization for inventory management, which can achieve more accurate demand forecasting and more robust inventory decisions under limited data conditions. This improves the scientific, economic and safety aspects of spare parts management in nuclear power plants, helps to improve the availability of nuclear power plant equipment, reduce redundant inventory levels, and ensure the safety and economy of material supply in complex operating environments.
[0080] Please see Figure 7 In some embodiments, step S620 may include, but is not limited to, steps S710 to S750: Step S710: If the lower bound distribution is greater than or equal to the critical ratio, then obtain the spare parts demand variable corresponding to the lower bound distribution to obtain the first variable; Step S720: Select the smallest first variable to obtain the lower bound quantile; Step S730: If the upper bound distribution is greater than or equal to the critical ratio, then obtain the spare parts demand variable corresponding to the upper bound distribution to obtain the second variable; Step S740: Select the smallest second variable to obtain the upper bound quantile; Step S750: Determine the critical quantile based on the lower and upper bound quantiles.
[0081] In step S710 of some embodiments, based on the boundary distribution of the fuzzy set... , Calculate the upper and lower critical quantiles using the critical ratio. The formulas for calculating the upper and lower critical quantiles are as follows: , , in, The lower bound quantile; The upper bound quantile; The distribution is the lower bound; The distribution is bounded above. The critical ratio; For spare parts demand variables.
[0082] Based on the above formulas for calculating the upper and lower critical quantiles, if the lower bound distribution is greater than or equal to the critical ratio, i.e. Then the lower bound distribution will be... Corresponding spare parts demand variables As the first variable.
[0083] In step S720 of some embodiments, the smallest first variable is selected as the lower bound quantile.
[0084] In step S730 of some embodiments, if the upper bound distribution is greater than or equal to the critical ratio, i.e. Then the upper bound distribution will be... Corresponding spare parts demand variables As the second variable.
[0085] In step S740 of some embodiments, the smallest second variable is selected as the upper bound quantile.
[0086] In step S750 of some embodiments, the lower bound quantile and the upper bound quantile are used as critical quantiles.
[0087] Through steps S710 to S750, the expected loss in the worst-case scenario can be minimized, demonstrating good robustness and computational efficiency, thereby improving the accuracy and efficiency of inventory control.
[0088] In step S180 of some embodiments, the expected inventory order quantity is updated to the target inventory order quantity, which enables optimal inventory ordering in an uncertain demand environment. The inventory control method based on JW fuzzy sets not only improves the accuracy of inventory order quantity prediction, but also shows stronger robustness and adaptability in the face of data sparsity and environmental uncertainty. It is suitable for the inventory control of key spare parts in complex systems with extremely high safety and reliability requirements, such as nuclear power plants.
[0089] Traditional inventory management systems are mostly statically configured, based on offline modeling of historical data, lacking the ability to learn online from new data and the adaptive adjustment function of inventory strategies, making continuous optimization difficult. Considering the frequent changes in the operating environment of nuclear power systems, the inventory control method of this application also supports an online update mechanism. It adjusts the estimated reference distribution and fuzzy set range in real time based on newly collected historical datasets or target demand-driven features, achieving continuous optimization and dynamic adaptive adjustment of inventory strategies. This not only applies to static inventory planning but also flexibly responds to changes in the nuclear power operating environment, such as equipment aging, maintenance cycle adjustments, and load changes, adapting to the high dynamic requirements of nuclear power systems. If dynamic model updates are required, steps S110 to S180 above can be repeated in each sales cycle to adjust the expected inventory order quantity based on the latest data.
[0090] Please see Figure 8 This application also provides an inventory control device that can implement the above-described inventory control method. The inventory control device includes: The first acquisition module 810 is used to acquire the historical dataset of the nuclear power plant; wherein, the historical dataset includes historical samples of the nuclear power plant during historical operating periods, and the historical samples include historical demand-driven characteristics and historical spare parts demand during historical operating periods; The first construction module 820 is used to construct a decision tree based on historical demand-driven features and historical spare parts demand, and to determine the historical leaf nodes of historical demand-driven features through the decision tree. The second acquisition module 830 is used to acquire the target demand-driven characteristics of the nuclear power plant during the target operating period and generate spare parts demand variables for the target operating period; wherein, the target operating period is located after the historical operating period; Module 840 is used to determine the target leaf node of the target demand-driven feature through a decision tree; The second construction module 850 is used to construct an estimated reference distribution of spare parts demand variables based on historical leaf nodes, target leaf nodes, historical spare parts demand quantities, and spare parts demand variables. The third building module 860 is used to construct a distributed fuzzy set based on the estimated reference distribution; The calculation module 870 is used to calculate the target inventory order quantity for the nuclear power plant based on the spare parts demand variables, the distributed fuzzy set, and the preset expected inventory order quantity. Update module 880 is used to update the expected inventory order quantity based on the target inventory order quantity.
[0091] The specific implementation method of this inventory control device is basically the same as the specific implementation method of the above-mentioned inventory control method, and will not be described again here.
[0092] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described inventory control method. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.
[0093] Please see Figure 9 , Figure 9 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 910 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 920 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 920 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 920 and is called and executed by the processor 910 using the inventory control method of the embodiments of this application. The input / output interface 930 is used to implement information input and output; The communication interface 940 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 950 transmits information between various components of the device (e.g., processor 910, memory 920, input / output interface 930, and communication interface 940); The processor 910, memory 920, input / output interface 930 and communication interface 940 are connected to each other within the device via bus 950.
[0094] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described inventory control method.
[0095] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
[0096] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.
[0097] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.
[0098] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.
[0099] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.
[0100] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0101] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.
[0102] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. The coupling or direct coupling or communication connection between the shown or discussed units may be through some interfaces, or indirect coupling or communication connection between the apparatus or units, and may be electrical, mechanical, or other forms.
[0103] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0104] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.
[0105] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0106] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.
Claims
1. An inventory control method, characterized in that, The method includes: Obtain historical datasets of nuclear power plants; wherein, the historical datasets include historical samples of the nuclear power plants during historical operating periods, and the historical samples include historical demand-driven characteristics and historical spare parts demand during the historical operating periods; A decision tree is constructed based on the historical demand-driven characteristics and the historical spare parts demand, and the historical leaf nodes of the historical demand-driven characteristics are determined through the decision tree. The target demand-driven characteristics of the nuclear power plant during the target operating period are obtained, and spare parts demand variables for the target operating period are generated; wherein, the target operating period is located after the historical operating period; The target leaf node of the target demand-driven feature is determined by the decision tree; Based on the historical leaf nodes, the target leaf nodes, the historical spare parts demand, and the spare parts demand variables, construct an estimated reference distribution for the spare parts demand variables; Construct a distributed fuzzy set based on the estimated reference distribution; The target inventory order quantity for the nuclear power plant is calculated based on the spare parts demand variable, the distributed fuzzy set, and the preset expected inventory order quantity. The expected inventory order quantity is updated based on the target inventory order quantity.
2. The method according to claim 1, characterized in that, The historical leaf nodes have historical index values, and the target leaf nodes have target index values. The step of constructing an estimated reference distribution for the spare parts demand variable based on the historical leaf nodes, the target leaf nodes, the historical spare parts demand quantity, and the spare parts demand variable includes: The first quantity is obtained by acquiring the number of historical samples in which the historical index value and the target index value are equal and the historical spare parts demand is less than or equal to the spare parts demand variable. Obtain the number of historical samples whose historical index value and target index value are equal, and obtain the second number; The estimated reference distribution is constructed based on the first quantity and the second quantity.
3. The method according to claim 1, characterized in that, The step of constructing a distributed fuzzy set based on the estimated reference distribution includes: Generate a candidate distribution of the spare parts demand variables and determine the fuzzy radius; Calculate the degree of difference between the candidate distribution and the estimated reference distribution; The candidate distributions are filtered based on the difference degree and the fuzzy radius to obtain the target distribution; Construct the distribution fuzzy set based on the target distribution.
4. The method according to any one of claims 1 to 3, characterized in that, The step of calculating the target inventory order quantity for the nuclear power plant based on the spare parts demand variable, the distributed fuzzy set, and the preset expected inventory order quantity includes: Calculate the stockout cost based on the spare parts demand variables and the expected inventory order quantity; Calculate the backlog cost based on the spare parts demand variables and the expected inventory order quantity; Obtain the boundary distribution of the fuzzy distribution set; The target inventory order quantity is calculated based on the stockout cost, the backlog cost, and the boundary distribution.
5. The method according to claim 4, characterized in that, The distributed fuzzy set includes multiple target distributions, and obtaining the boundary distribution of the distributed fuzzy set includes: Compare each of the target distributions with the estimated reference distribution; If the target distribution is less than or equal to the estimated reference distribution, then the target distribution is taken as the first distribution, and the smallest of the first distributions is selected as the lower bound distribution. If the target distribution is greater than the estimated reference distribution, then the target distribution is taken as the second distribution, and the largest second distribution is selected as the upper bound distribution; The boundary distribution is determined based on the lower bound distribution and the upper bound distribution.
6. The method according to claim 4, characterized in that, The step of calculating the target inventory order quantity based on the stockout cost, the overstock cost, and the boundary distribution includes: Calculate the critical ratio based on the stockout cost and the backlog cost; Calculate the critical quantile based on the boundary distribution and the critical ratio; The target inventory order quantity is calculated based on the stockout cost, the backlog cost, and the critical quantile.
7. The method according to claim 5, characterized in that, The boundary distribution includes a lower bound distribution and an upper bound distribution. The calculation of the critical quantile based on the boundary distribution and the critical ratio includes: If the lower bound distribution is greater than or equal to the critical ratio, then the spare parts demand variable corresponding to the lower bound distribution is obtained to obtain the first variable; The lower bound quantile is obtained by selecting the first variable with the smallest value. If the upper bound distribution is greater than or equal to the critical ratio, then the spare parts demand variable corresponding to the upper bound distribution is obtained to obtain the second variable; The upper bound quantile is obtained by selecting the smallest second variable; The critical quantile is determined based on the lower bound quantile and the upper bound quantile.
8. An inventory control device, characterized in that, The device includes: The first acquisition module is used to acquire the historical dataset of the nuclear power plant; wherein, the historical dataset includes historical samples of the nuclear power plant during historical operating periods, and the historical samples include historical demand-driven characteristics and historical spare parts demand during the historical operating periods; The first construction module is used to construct a decision tree based on the historical demand-driven features and the historical spare parts demand, and to determine the historical leaf nodes of the historical demand-driven features through the decision tree. The second acquisition module is used to acquire the target demand-driven characteristics of the nuclear power plant during the target operating period and generate spare parts demand variables for the target operating period; wherein the target operating period is located after the historical operating period; The determination module is used to determine the target leaf node of the target demand-driven feature through the decision tree; The second construction module is used to construct an estimated reference distribution of the spare parts demand variable based on the historical leaf node, the target leaf node, the historical spare parts demand quantity, and the spare parts demand variable; The third construction module is used to construct a distributed fuzzy set based on the estimated reference distribution; The calculation module is used to calculate the target inventory order quantity of the nuclear power plant based on the spare parts demand variable, the distributed fuzzy set, and the preset expected inventory order quantity. An update module is used to update the expected inventory order quantity based on the target inventory order quantity.
9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 7.
10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1 to 7.