A demand prediction method based on equipment maintenance equipment consumption law

Abstract

The application relates to a kind of equipment maintenance equipment consumption law-based demand prediction methods, comprising S1, the normal working time X of equipment and the defect working time Y are collected, the normal working time X of equipment is subjected to normal distribution N1 (mu1, sigma1 2 ), the defect stage working time Y is subjected to normal distribution N2 (mu2, sigma2 2 );S2, the first equipment consumption law model based on time maintenance mode in [0, T] time is constructed, and the second equipment consumption law model under the condition of state maintenance mode at T moment is constructed, then the average value of equipment maintenance consumption in [0, T] time is constructed;S3, the average value of equipment maintenance consumption is combined with the number of equipment served to unit equipment and the equipment safety stock level, and the equipment replenishment forecast model is obtained.The method improves the accuracy and reliability of equipment demand prediction by deeply mining the consumption law of equipment maintenance equipment.

Description

Technical Field

[0001] This invention relates to the field of big data processing technology, specifically to a demand prediction method based on the consumption patterns of equipment maintenance materials. Background Technology

[0002] Equipment maintenance materials are essential resources for equipment maintenance and constitute a major factor in equipment use and maintenance costs. Scientifically analyzing the consumption patterns of equipment maintenance materials and rationally predicting and determining the demand for equipment maintenance materials can not only reduce equipment maintenance costs but also effectively shorten equipment maintenance time and improve equipment readiness.

[0003] Currently, there are many technologies for predicting equipment maintenance materials based on equipment usage time or equipment technical condition. For example, time-based maintenance methods analyze influencing factors such as unit service life and equipment usage time, and establish equipment spare parts prediction models based on unit service life; a joint strategy for maintenance and spare parts ordering based on remaining life prediction is proposed, and the model is optimized using the renewal reward theory; a spare parts prediction model based on condition-based maintenance is established using discrete event simulation methods, which can better solve the problems of accuracy and agility in assessment.

[0004] In general, most equipment maintenance material prediction technologies utilize a series of models, algorithms, and methods to obtain the demand for equipment maintenance materials. However, a method for constructing a maintenance material demand prediction model based on a comprehensive consideration of the common consumption patterns of equipment maintenance materials under two maintenance modes based on time and state within the mission cycle, and thus calculating and predicting maintenance material demand, remains lacking. Therefore, more advanced algorithms and models are needed to conduct in-depth research on the consumption patterns of equipment maintenance materials under the two maintenance modes based on time and state, and to scientifically predict the maintenance material allocation needs of equipment maintenance support organizations, thereby improving the accuracy and reliability of the predictions. Summary of the Invention

[0005] This invention provides a demand forecasting method based on the consumption patterns of equipment maintenance materials. The aim is to scientifically and rationally predict the demand for maintenance materials by deeply exploring the common patterns of maintenance materials under two maintenance modes: time-based and state-based, thus providing a scientific basis for the procurement and allocation of equipment maintenance materials.

[0006] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions:

[0007] A demand forecasting method based on the consumption patterns of equipment maintenance materials includes the following steps:

[0008] S1. Collect the normal working time X and defective working time Y of the equipment, and statistically determine the normal working time X of the equipment according to the normal distribution N1(μ1, σ1). 2 The working time Y during the defect stage follows a normal distribution N2(μ2, σ2). 2 ).

[0009] S2, According to the normal distribution N1(μ1, σ1) 2 Construct a first equipment consumption law model based on time maintenance method within the time interval [0, T], and based on the normal distribution N2(μ2, σ2) 2 A second equipment consumption pattern model based on condition-based maintenance mode is constructed at time T by combining the remaining life threshold of the equipment. Then, the average equipment maintenance consumption within the time period [0, T] is constructed based on the first equipment consumption pattern model and the second equipment consumption pattern model.

[0010] The first equipment consumption pattern model is as follows:

[0011]

[0012] The second equipment consumption pattern model is as follows

[0013]

[0014] The average maintenance cost of the equipment is

[0015]

[0016] S3. The average value of equipment maintenance consumption is combined with the number of equipment serving the unit's equipment and the equipment safety stock level to obtain the equipment replenishment prediction model.

[0017]

[0018] In the formula:

[0019] Q represents the predicted quantity of equipment needed.

[0020] S represents the annual safety stock level of the maintenance and support organization;

[0021] q represents the annual consumption of maintenance equipment.

[0022] W represents the quantity of equipment deployed by the unit;

[0023] L represents the number of equipment components in a unit.

[0024] Furthermore, before constructing the second equipment consumption law model, the probability of equipment consumption at time T under normal and defective conditions should be established separately.

[0025] The probability of equipment consumption during normal operation is:

[0026]

[0027] When the equipment is in a defective working state and its remaining lifespan is greater than the equipment's lifespan threshold, the probability of equipment consumption for repair is:

[0028]

[0029] When the equipment is in a defective working state and its remaining lifespan is less than the equipment's lifespan threshold, the probability of equipment consumption for repair is:

[0030]

[0031] By combining the probability of equipment consumption under normal working conditions and defective working conditions, a second equipment consumption law model is constructed, namely q2=P1×0+P2×0+P3×1.

[0032] The embodiments of the present invention have the following advantages:

[0033] This invention proposes a demand forecasting method based on the consumption patterns of equipment maintenance materials. First, it collects maintenance material consumption data for the equipment configured in a specific unit, including equipment model, the quantity of a single unit within that equipment, normal operating time and defective operating time under time-based maintenance, remaining lifespan threshold under condition-based maintenance, and the safety stock level of the maintenance support organization for that unit. This data lays the foundation for analyzing and mining the consumption patterns and demand forecasting of maintenance materials for that unit. Then, based on the collected data, a mathematical model for material demand forecasting based on the consumption patterns of equipment maintenance materials over a time interval [0, T] is constructed. This model includes a mathematical model for material consumption patterns based on time-based maintenance, a mathematical model for material consumption under condition-based maintenance, and a maintenance material demand forecasting model.

[0034] This method improves the accuracy and reliability of equipment demand forecasting by deeply exploring the consumption patterns of equipment maintenance materials. This technology has broad application prospects and significant practical value, further promoting equipment support efficiency. It can provide a scientific basis for the procurement and allocation of equipment maintenance materials, and provide strong support for equipment support work. Furthermore, it can be continuously trained and iterated with more sample data to continuously optimize the algorithm, providing a scientific basis and theoretical support for the formulation of equipment maintenance material supply plans and the demonstration of equipment maintenance support systems. Attached Figure Description

[0035] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0036] The structures, proportions, sizes, etc. illustrated in this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed herein, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0037] Figure 1 For the equipment degradation process over training time

[0038] Figure 2 Here is a flowchart of a demand forecasting method based on the consumption patterns of equipment maintenance materials provided in Embodiment 1 of the present invention;

[0039] Figure 3 This is the result of equipment demand forecasting based on the consumption patterns of equipment maintenance materials. Detailed Implementation

[0040] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] In this technology, the equipment's transition from "normal" to "faulty" is not instantaneous, but rather a gradual deterioration process with increasing training time. This involves moving from a normal working stage to a defective working stage, and then to a functional failure stage. Since the occurrence of equipment defects and failure points has a certain degree of randomness, the normal working stage and the defective working stage are not fixed values ​​but also random. Through repeated iterations of the equipment's transition from "normal" to "faulty," certain patterns can be abstracted. These patterns can then guide the prediction and decision-making in equipment maintenance. Figure 1 This refers to the process of equipment deterioration over training time.

[0042] Since most equipment currently undergoes both post-failure maintenance and preventative maintenance, this technology fully considers and analyzes the consumption patterns of maintenance equipment based on time and condition-based maintenance methods.

[0043] The time-based maintenance method mainly involves the equipment going through normal working stage and defective working stage as training time progresses, and then being replaced naturally after reaching the functional failure stage, also known as post-failure maintenance.

[0044] Condition-based maintenance primarily utilizes testing instruments or external monitoring equipment to replace or repair equipment after it has entered the defective working stage but before it develops into a functional failure, thus preventing the occurrence of functional failures. It is also known as preventive maintenance or condition-based maintenance.

[0045] Both time-based and state-based maintenance methods are consistent with the realities of equipment maintenance. Analyzing the consumption of maintenance equipment based on its historical maintenance records reveals traceable patterns and certain abstract rules. Based on this, this method constructs an equipment maintenance equipment prediction algorithm based on consumption pattern mining, providing a scientific basis and theoretical support for the procurement and allocation of equipment maintenance equipment.

[0046] However, the transition of equipment from a "normal" state to a "faulty" state involves two phases: a normal working phase and a defective working phase. These two phases have different, independent failure rates that follow a certain distribution. The failure rate during normal working is lower than that during the defective working phase. Therefore, generally, the normal working phase lasts longer than the defective working phase. This analysis distinguishes between two branches: post-fault maintenance methods and preventative maintenance methods, focusing on time-based and state-based maintenance approaches. It investigates the common patterns in equipment maintenance material consumption, providing theoretical support for predicting equipment maintenance material demand.

[0047] Based on the analysis of the consumption of time-based equipment, within the time frame [0, T], the main maintenance method adopted is time-based maintenance, i.e., maintenance after failure. According to the renewal theory, after the first piece of equipment goes through the normal working stage and the defective working stage and reaches the functional failure stage, the second piece of equipment takes over the first one and continues to work. The lifespan of each unit is independent of each other. The amount of maintenance equipment consumed within the time frame [0, T] corresponds to the number of times the equipment is renewed.

[0048] Based on the analysis of equipment consumption at time T, the equipment consumption at time T depends on the functional state of the equipment. Here, we analyze two scenarios at time T: when the equipment is in normal working condition and when it is in a defective working condition. When the equipment identified at time T is in normal working condition, no corresponding maintenance measures are taken. When the equipment is in a defective working condition at time T and its service life is less than the service life threshold, it is replaced; if its service life is greater than the service life threshold, it is not replaced.

[0049] Based on the mining and prediction of equipment consumption patterns in time and state, and according to the consumption of equipment maintenance materials in different training cycles, a (R, S, M) strategy is adopted. That is, after each training cycle R, if the inventory level is lower than S based on the annual consumption of maintenance materials, the inventory level is replenished to S. After analyzing the patterns of equipment consumption and inventory replenishment over many years, the demand for equipment maintenance materials M is predicted. This provides a scientific basis and algorithmic support for the procurement and allocation of maintenance materials in the equipment department.

[0050] Example 1

[0051] Based on the above theories, such as Figure 2 As shown, a demand forecasting method based on the consumption patterns of equipment maintenance materials includes the following steps:

[0052] S1. Collect the normal working time X and defective working time Y of the equipment, and statistically determine the normal working time X of the equipment according to the normal distribution N1(μ1, σ1). 2 The working time Y during the defect stage follows a normal distribution N2(μ2, σ2). 2 ).

[0053] S2, According to the normal distribution N1(μ1, σ1) 2 Construct a first equipment consumption law model based on time maintenance method within the time interval [0, T], and based on the normal distribution N2(μ2, σ2) 2 A second equipment consumption pattern model based on condition-based maintenance mode is constructed at time T by combining the remaining life threshold of the equipment. Then, the average equipment maintenance consumption within the time period [0, T] is constructed based on the first equipment consumption pattern model and the second equipment consumption pattern model.

[0054] The first equipment consumption pattern model is as follows:

[0055]

[0056] The second equipment consumption pattern model is as follows

[0057]

[0058] The average maintenance cost of the equipment is

[0059]

[0060] S3. The average value of equipment maintenance consumption is combined with the number of equipment serving the unit's equipment and the equipment safety stock level to obtain the equipment replenishment prediction model.

[0061]

[0062] In the formula:

[0063] Q represents the predicted quantity of equipment needed.

[0064] S represents the annual safety stock level of the maintenance and support organization;

[0065] q represents the annual consumption of maintenance equipment.

[0066] W represents the quantity of equipment deployed by the unit;

[0067] L represents the number of equipment components in a unit.

[0068] Example 2

[0069] Since there are not only equipment in normal working condition at time T, but also equipment in defective working condition, Example 2 improves the construction of the second equipment consumption law model based on Example 1.

[0070] The equipment consumption at time T depends on the functional state of the equipment. Here, we analyze two scenarios: the equipment is in normal working condition at time T, and it is in a defective working condition. When the equipment is identified as being in normal working condition at time T, no maintenance measures are taken. When the equipment is in a defective working condition at time T, and its service life is less than the threshold, it is replaced. If its service life is greater than the threshold, it is not replaced.

[0071] Before constructing the second equipment consumption law model, the probability of equipment consumption at time T under normal and defective conditions should be established respectively.

[0072] The probability of equipment consumption during normal operation is:

[0073]

[0074] When the equipment is in a defective working state and its remaining lifespan is greater than the equipment's lifespan threshold, the probability of equipment consumption for repair is:

[0075]

[0076] When the equipment is in a defective working state and its remaining lifespan is less than the equipment's lifespan threshold, the probability of equipment consumption for repair is:

[0077]

[0078] By combining the probability of equipment consumption under normal and defective operating conditions, a second equipment consumption pattern model is constructed, namely...

[0079] The following is a specific example of how this method is implemented:

[0080] Taking the demand forecasting of tank equipment maintenance materials in a certain unit as an example, this invention employs the material demand forecasting technology based on consumption pattern mining algorithm to conduct in-depth analysis of the consumption data of vehicle equipment maintenance materials. This reveals common patterns in the consumption of maintenance materials based on two maintenance methods: time and state, and a maintenance material demand forecasting algorithm is constructed. The implementation process of the demand forecasting algorithm based on the consumption patterns of equipment maintenance materials mainly consists of five steps: data collection, pattern mining, instance substitution, algorithm solution, and demand forecasting. The mathematical symbols and meanings involved in each algorithm during implementation are shown in Table 1.

[0081] Table 1. Mathematical symbols and their meanings involved in the algorithm.

[0082]

[0083] 1. Data Collection

[0084] Eight tanks of a certain type were collected from a certain organizational unit. Each tank underwent a combination of post-failure maintenance and condition-based maintenance. Data collection and preprocessing of a specific unit of this tank type revealed that the normal operating time X of this unit follows a normal distribution of N(1000, 1002), and the working time Y during the defect stage follows a normal distribution of N(100, 102). There are five units of this type. During condition-based maintenance, if the remaining service life is less than 100 hours, it is replaced. The inventory level of this unit's maintenance support organization remains unchanged at 100 units.

[0085] 2. Pattern Discovery

[0086] The consumption patterns of maintenance equipment in the unit within the time period [0, 1500]h are analyzed to determine the quantity of maintenance equipment that the maintenance support organization of this type of equipment needs to store at time 1500h. The consumption patterns and standards of maintenance equipment under different annual training are also analyzed to predict the quantity of maintenance equipment that needs to be applied for by the superior department, thereby ensuring the safety stock level of maintenance equipment.

[0087] 3. Example Substitution

[0088] As mentioned above, the failure probability functions for the normal operating time and defect stage operating time of this type of unit are respectively...

[0089]

[0090] The failure probability functions for the tank equipment's cumulative normal operating time and cumulative defect stage operating time within the time interval [0, 1500]h are respectively...

[0091]

[0092] The average number of repair materials consumed during the repair of tank equipment after a failure within a time period of [0, 1500] hours is:

[0093]

[0094] The average consumption of maintenance equipment for tank equipment during 1500 hours of operation was [amount missing].

[0095]

[0096] The average number of tank maintenance materials consumed within the time interval [0, 1500] hours is:

[0097]

[0098]

[0099] Finally, using the data mining tools of the data analysis subsystem for big data management and analysis of joint operations equipment maintenance and support, we performed programming calculations and obtained q≈1.37.

[0100] 4. Algorithm Solution

[0101] According to the data collection results, the maintenance and support organization for this type of tank has an inventory of 100 pieces of this type of equipment. A certain unit is equipped with 8 tanks of this type, and the unit has 5 individual units. The estimated quantity of remaining maintenance equipment stored by the maintenance and support organization within the time period [0, 1500] hours is...

[0102]

[0103] Therefore, based on the equipment maintenance material prediction algorithm using consumption pattern mining, the quantity of maintenance materials that a certain type of tank equipment needs to apply for from higher authorities after 1500 hours is [amount missing].

[0104] M≈SQ≈100-45.2≈55.

[0105] Calculations show that after 1500 hours of operation, a certain type of tank needs to apply to higher authorities for approximately 55 units of maintenance equipment to maintain its safe inventory level.

[0106] 5. Demand Forecasting

[0107] Finally, the data mining tools of the data analysis subsystem for big data management and analysis of joint operations equipment maintenance and support can be used to programmatically predict the number of equipment maintenance and support materials that equipment maintenance and support organizations need to configure for different training periods in different years.

[0108] like Figure 3As shown, the amount of tank equipment consumed by unit maintenance equipment increases with the increase of annual training time. Equipment maintenance and support organizations can scientifically predict the amount of maintenance equipment to be configured based on the consumption patterns of equipment maintenance equipment obtained through analysis and the inventory safety level.

[0109] Although the present invention has been described in detail above with general descriptions and specific embodiments, modifications or improvements can be made to it, which will be obvious to those skilled in the art. Therefore, all such modifications or improvements made without departing from the spirit of the present invention fall within the scope of protection claimed by the present invention.

Claims

1. A demand forecasting method based on the consumption patterns of equipment maintenance materials, characterized in that, Includes the following steps: S1. Collect the normal working time X and defective working time Y of the equipment, and statistically determine the normal working time X of the equipment according to the normal distribution N1(μ1, σ1). 2 The working time Y during the defect stage follows a normal distribution N2(μ2, σ2). 2 ); S2, According to the normal distribution N1(μ1, σ1) 2 Construct a first equipment consumption law model based on time maintenance method within the time interval [0, T], and based on the normal distribution N2(μ2, σ2) 2 A second equipment consumption pattern model based on condition-based maintenance mode is constructed at time T by combining the remaining life threshold of the equipment. Then, the average equipment maintenance consumption within the time period [0, T] is constructed based on the first equipment consumption pattern model and the second equipment consumption pattern model. The first equipment consumption pattern model is as follows: The second equipment consumption pattern model is as follows The average maintenance cost of the equipment is S3. The average value of equipment maintenance consumption is combined with the number of equipment serving the unit's equipment and the equipment safety stock level to obtain the equipment replenishment prediction model. In the formula: Q represents the predicted quantity of equipment needed. S represents the annual safety stock level of the maintenance and support organization; q represents the annual consumption of maintenance equipment; W represents the quantity of equipment deployed by the unit; L represents the number of equipment components in a unit.

2. The demand forecasting method based on the consumption patterns of equipment maintenance materials as described in claim 1, characterized in that: Before constructing the second equipment consumption law model, the probability of equipment consumption at time T under normal and defective conditions should be established respectively. The probability of equipment consumption during normal operation is: When the equipment is in a defective working state and its remaining lifespan is greater than the equipment's lifespan threshold, the probability of equipment consumption for repair is: When the equipment is in a defective working state and its remaining lifespan is less than the equipment's lifespan threshold, the probability of equipment consumption for repair is: By combining the probability of equipment consumption under normal and defective operating conditions, a second equipment consumption pattern model is constructed, namely...

Images