CN113723978B

CN113723978B - Bearing bush demand prediction method and system

Info

Publication number: CN113723978B
Application number: CN202010452767.3A
Authority: CN
Inventors: 张博宇
Original assignee: BMW Brilliance Automotive Ltd
Current assignee: BMW Brilliance Automotive Ltd
Priority date: 2020-05-26
Filing date: 2020-05-26
Publication date: 2023-10-24
Anticipated expiration: 2040-05-26
Also published as: CN113723978A

Abstract

The disclosure provides a bearing bush demand prediction method and a bearing bush demand prediction system. The bearing bush demand prediction method comprises the following steps: acquiring historical data of tolerances of a crankcase and a crankshaft of a scanned engine; determining bearing bush types and historical data of bearing bush demand corresponding to each bearing bush type according to the tolerance of a crankcase and a crankshaft; establishing at least one time sequence prediction model for historical data of bearing bush demand according to each bearing bush type; receiving input from a user regarding a predicted time period and a particular bearing shell type; and predicting bearing shell demand during the received predicted time period using the at least one time series prediction model for the particular bearing shell category.

Description

Bearing bush demand prediction method and system

Technical Field

The present disclosure relates to the field of time series prediction, and more particularly, to a method and system for predicting axle shoe demand using a time series prediction model.

Background

Bearing shells are widely used in automotive engines. The bearing bush is a tile-shaped part positioned between a crank case and a crankshaft of the engine and plays a role in fixing and lubricating. Since the crankcase and crankshaft of each engine have different tolerances, the variety and number of bearing shells required is also different. The bearing bush comprises a normal bearing bush and a thrust bearing bush. Each bearing shell comprises a plurality of types, for example, 8 thrust bearing shells and 16 normal bearing shells can exist.

These bushes are usually imported from abroad, for example germany. The transportation time is about three weeks or so, and engine manufacturers typically order four weeks in advance. There is a need to accurately predict the type and number of bushings to be purchased. Because one to two engine changes can affect the type and number of bearing shells required, if one or more types of bearing shells are purchased too much, this can result in the engine being produced without being so much and left, resulting in wasted resources and economic losses for the manufacturer. However, if the number of purchase is insufficient, but one of the bearing shells is worn out more quickly, the cost of purchasing and transporting the one bearing shell alone is higher. Currently, it is often determined empirically by considering existing stock data, engine data, and the like. However, the manual judgment method has a large error, and the bearing bush demand of several weeks in the future cannot be objectively and accurately predicted.

Accordingly, there is a need for a method and system for objectively and accurately predicting future bearing shell demand.

Disclosure of Invention

The present disclosure provides a novel bearing shell demand prediction method and system.

According to a first aspect of the present disclosure, there is provided a bearing shell demand prediction method, including: acquiring historical data of tolerances of a crankcase and a crankshaft of a scanned engine; determining bearing bush types and historical data of bearing bush demand corresponding to each bearing bush type according to the tolerance of a crankcase and a crankshaft; establishing at least one time sequence prediction model for historical data of bearing bush demand according to each bearing bush type; receiving input from a user regarding a predicted time period and a particular bearing shell type; and predicting bearing shell demand during the received predicted time period using the at least one time series prediction model for the particular bearing shell category.

When the at least one time series prediction model is a plurality of time series prediction models, the method further comprises: predicting the bearing bush demand in a specific time period before the predicted time period by using the plurality of time sequence prediction models respectively; comparing the bearing bush demand predicted by using each time sequence prediction model in the specific time period with the real demand respectively to obtain each error; determining a time sequence prediction model corresponding to the minimum error as an optimal time sequence prediction model; and predicting a bearing shell demand during the received predicted time period for the particular bearing shell category using the optimal time series prediction model.

The specific time period is the previous week of the predicted time period.

The plurality of time series prediction models includes an integrated moving average autoregressive (ARIMA) model, a Bayesian Analysis Time Series (BATS) model, a trend error season (ETS) model, and a Holt-windows model.

The method further comprises the steps of: acquiring bearing bush inventory data; calculating a minimum safety stock value according to the bearing bush demand during the predicted time period and the bearing bush stock data; and when the difference between the current bearing bush stock data and the minimum safety stock value is within a preset threshold value, sending out a warning message.

The method further comprises preprocessing the historical data prior to establishing the at least one time series prediction model, the preprocessing comprising excluding weekdays and holidays and their corresponding data. The preprocessing also includes excluding data in the historical data outside of a confidence interval having a confidence level of 95%.

The ARIMA model is represented as ARIMA (p, d, q), where p is the number of autoregressive terms, q is the number of moving average terms, d is the number of differences made to make the time series of the historical data a stationary series, and building the ARIMA model comprises: determining whether the time sequence of the historical data is a stable sequence by adopting a unit root test method; if the sequence is a non-stationary sequence, performing differential processing on the time sequence, and checking again until the sequence becomes a stationary sequence, wherein the number of times of performing the differential is determined as the value of d; determining the values of q and p through the recognition rule of the time sequence; and training and testing the ARIMA model with 70% of the historical data as training data and the remaining 30% of the historical data as test data to obtain a final ARIMA model.

The input of the predicted time period is received via a slider on the user interface representing the time period. The predicted period of time is in days or weeks.

The user interface further includes an input for selecting a particular bearing shell family from one or more bearing shell families, each of the one or more bearing shell families including one or more bearing shell categories, the method further comprising: predicting bearing shell demand during the received predicted time period using the at least one time series prediction model for each bearing shell category in the particular bearing shell series in response to a user selection of the particular bearing shell series; displaying on the user interface historical data of bearing shell demand and a time-dependent plot of predicted bearing shell demand for each bearing shell category of the particular bearing shell series.

According to a second aspect of the present disclosure, there is provided a non-transitory computer-readable medium having stored thereon computer-executable instructions that, when executed by one or more computing devices, cause the one or more computing devices to perform the method according to any one of claims 1-11.

According to a third aspect of the present disclosure, there is provided a computer system comprising: at least one processor; and at least one non-transitory computer-readable medium having stored thereon computer-executable instructions that, when executed by the at least one processor, cause the at least one processor to perform the method according to any of claims 1-11.

Other features of the present invention and its advantages will become more apparent from the following detailed description of exemplary embodiments of the invention, which proceeds with reference to the accompanying drawings.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description, serve to explain the principles of the disclosure.

The disclosure may be more clearly understood from the following detailed description taken in conjunction with the accompanying drawings in which:

fig. 1 shows a flowchart of a bearing shell demand prediction method according to an exemplary embodiment of the present invention.

Fig. 2 shows a flowchart of a bearing shell demand prediction method according to another exemplary embodiment of the present invention.

Fig. 3 shows a flow chart of the establishment of an ARIMA model according to an exemplary embodiment of the invention.

FIG. 4 shows a schematic diagram of a user interface according to an exemplary embodiment of the invention.

FIG. 5 illustrates an exemplary configuration of a computing device in which embodiments according to the invention may be implemented.

Detailed Description

Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Details and functions not necessary for the invention are omitted so as not to obscure the understanding of the present invention.

Note that like reference numerals and letters refer to like items in the figures, and thus once an item is defined in one figure, it is not necessary to discuss it in subsequent figures.

In this disclosure, the terms "first," "second," and the like are used merely to distinguish between elements or steps and are not intended to indicate a temporal order, priority, or importance.

According to the invention, the time series modeling is carried out on the historical data of the bearing bush demand of various types, so that the future bearing bush demand is predicted. Furthermore, the invention can build different time series models for the historical data, and automatically select the optimal time series model for prediction.

Fig. 1 shows a flowchart of a bearing shell demand prediction method according to an exemplary embodiment of the present invention. The method may be performed by a computer system.

As shown in fig. 1, at step S101, the computer system obtains historical data of the tolerances of the crankcase and crankshaft of the scanned engine from a database. The tolerances of the crankcase and crankshaft are typically in the order of microns and are difficult for the naked eye to identify and are thus obtained by scanning the grooves of the engine with an optical scanning device. An optical scanning device scans each engine on the production line and stores scanned tolerance data for the corresponding crankcase and crankshaft in a database.

In step S102, the computer system determines a bearing shell category and a bearing shell demand corresponding to each bearing shell category based on the tolerances of the crankcase and the crankshaft. The crankcase and crankshaft of each engine have respective different tolerances, and accordingly require different types and numbers of bearing shells. The bearing shell typically includes a normal bearing shell and a thrust bearing shell. Each bearing shell is of various types, for example, 8 thrust bearing shells and 16 normal bearing shells can be present. The computer system calculates the consumption of various bearing bushes required every day according to the engine data scanned every day.

In step S103, the computer system builds at least one time series model for the historical data of the bearing shell demand for each bearing shell category.

Time series refers to a set of digit sequences arranged in time series, and time series analysis is to use the set of digit sequences to apply a mathematical statistical method to process, so as to predict the development of future things. With the development of machine learning in recent years, time series prediction has been widely used in various fields. A number of different time series prediction models have been developed, including an integrated moving average autoregressive (ARIMA) model, a Bayesian Analysis Time Series (BATS) model, a trend error season (ETS) model, a Holt-windows model, and the like.

The ARIMA model is one of the methods of time series predictive analysis, generally denoted ARIMA (p, d, q), where AR is "autoregressive" and p is the number of autoregressive terms; MA is "moving average", q is the number of terms of the moving average, and d is the number of differences (orders) made to make the time series to be analyzed a stationary series.

The BATS model is a prediction using Bayesian statistics. Bayesian statistics is different from general statistical methods, and not only utilizes model information and data information, but also fully utilizes prior information.

The ETS model applies seasonal time series in the smoothing method calculation.

Holt-windows is a three-time exponential smoothing algorithm that can well predict time series.

The above 4 models are all the most common algorithms for time series, but each is slightly different. The ARIMA model is more suitable when the comparison of the ARIMA model is focused on the period and the similarity of the data trend is high in each period. Holt-windows focuses on averaging. BATS is relatively accurate for random fluctuating, varying relatively stable data, predicted by a common mean discount model. ETS may not process or be exceptionally slow for long period sequences due to the very many parameters to be estimated.

For each bearing shell category, the computer system builds one or more of the above time series models for historical data of bearing shell demand. The process of building the time series model will be described in detail later with reference to fig. 3.

It will be appreciated by those skilled in the art that the above-listed time series model is not exhaustive, but is merely exemplary shown for the purpose of illustrating the inventive concepts of the present invention. The inventive concept of the present invention is not limited to these models, but may be applied to other models of time series, such as hw and/or nnetar.

Optionally, prior to this step, the historical data may be pre-processed. For example, weekdays and holidays and their corresponding data may be excluded from the historical data. Since sunday is a non-production day, even if there is data, such data samples will affect the prediction accuracy, and are therefore excluded. Furthermore, discrete values may be excluded, such as taking only data within a confidence interval of 95% confidence level in the historical data.

In step S104, the computer system receives input from the user regarding the predicted time period and the particular bearing shell type.

The prediction period may be in days or weeks. For example, the user may choose to predict the future 20 day bearing shell demand, or choose to predict the future four week bearing shell demand, etc. The user may specify the time period to predict by sliding a slider bar on the display interface representing the time period.

The bearing shell type to be predicted can be input alone or by designating the bearing shell series to which it belongs. In the latter case, the bearing shell types may be grouped in series. For example, several bearing shells are classified into a certain series according to their properties. For example, the bearing shell series A00001 includes three types of bearing shells A00001-01, A00001-02 and A00001-03. The input of the bearing shell category may be accomplished by a user specifying the bearing shell series to which the bearing shell category belongs. For example, the user may select a particular bushing series from a drop down menu of bushing series. In this case, the computer system will make predictions for each of the bearing shell categories of the selected bearing shell series. The prediction process will be described in step S105.

In step S105, the computer system predicts the bearing shell demand during the predicted time period for the particular bearing shell category selected by the user using the at least one time series prediction model established in step S103. The computer system can display the historical data and the predicted bearing bush demand on a user interface, so that a user can visually check the bearing bush demand.

In the case where the user specifies a family of bearing shells rather than a single family of bearing shells, the computer system will utilize the at least one time series prediction model to predict bearing shell demand during the received predicted time period for each family of bearing shells in the selected family of bearing shells. The computer system then displays on the user interface both historical data of bearing shell demand and a time-dependent plot of predicted bearing shell demand for each bearing shell category of the selected particular bearing shell family. A diagram of the user interface will be described below with reference to fig. 4.

Those skilled in the art will appreciate that the steps of the foregoing methods are not necessarily performed in the order described above, but they may be performed simultaneously, in a different order, or in an overlapping manner. For example, step S104 may precede step S101.

As previously described, the computer system may build one or more time series models for each bearing shell category. When a plurality of time series models are built in step S103, the present invention may preferably automatically select an optimal time series model for future prediction. Described in detail below in conjunction with fig. 2.

As shown in fig. 2, the method may begin at the location indicated by reference numeral a in fig. 1. In step S201, the computer system predicts the bearing shell demand for a particular time period prior to the future time period to be predicted using a plurality of established time series prediction models, such as ARIMA model, BATS model, ETS model, and/or Holt-windows model, respectively. The specific time period may be the previous week of the future time period to be predicted, i.e. the previous week of the current time. The specific period of time is not limited to one week, and may be any time designated by the user, for example, ten days, or the like, by those skilled in the art.

In step S202, the computer system compares the respective bearing shell demands for the specific time period predicted by the respective time series prediction models with the actual demands, respectively, to obtain respective errors. Various ways of calculating the error may be used by those skilled in the art, such as sample variance, sample standard deviation, total standard deviation, etc.

In step S202, the computer system determines a time-series prediction model corresponding to the minimum error as an optimal time-series prediction model. For example, if the ARIMA model predicts that the error between the bearing shell demand and the actual consumption is minimal for the last week, i.e., the predicted value is closest to the actual value, the ARIMA model is determined to be the optimal time series prediction model for future predictions.

In step S202, the computer system predicts the bearing shell demand during the future time period for the particular bearing shell category selected by the user using the determined optimal time series prediction model, e.g., ARIMA model.

Furthermore, as previously described, the user may specify a family of bearing shells rather than a single bearing shell category. The computer system will predict the bearing shell demand during the future predicted time period using the determined optimal time series prediction model for each bearing shell category in the selected bearing shell series.

The process of building a time series model is described below in connection with fig. 3. The ARIMA model is described below as an example.

The ARIMA model is generally denoted ARIMA (p, d, q), where p is the number of autoregressive terms, q is the number of moving average terms, and d is the number of differences made to make the time series of the historical data a stationary series. The time series predicted using the ARIMA model is generally denoted as X _t ＝(α ₁ X _t-1 +α ₂ X _t-2 +…+α _p X _t-p )+(β ₁ ε _t-1 +β ₂ ε _t-2 +…+β _q ε _t-q ). Thus, ARIMA is built to determine d, p, q and the respective coefficients α and β. Specifically, establishing the ARIMA model generally includes steps S301-S304.

In step S301, the computer system determines whether the time series of the history data is a stationary series using a unit root check method.

The unit root check refers to checking whether a unit root exists in the sequence, and if so, a non-stationary time sequence. Statistically, the Dickey-Fuller test (Dickey-Fuller test) can test whether an autoregressive modelThere is a unit root (unit root). A simple AR (1) model is y _t ＝ρy _t-1 +u _t 。y _t Is the variable to be checked, t is time, ρ is the coefficient, u _t Is an error term. If |ρ| is equal to or greater than 1, it is stated that the unit root is present. The regression model may be written as deltay _t ＝(ρ-1)y _t-1 +u _t ＝δy _t-1 +u _t Delta is the first order difference. Testing whether there is a unit root is equivalent to testing whether δ=0. Because the dir-fowler test is a residual term, and not raw data, the standard t statistic cannot be used. We need to use the dir-fowler statistic.

In step S302, if the test result is a non-stationary sequence, the time sequence is subjected to differential processing and checked again until the sequence becomes a stationary sequence, wherein the number of times of performing the differential is determined as the value of d.

In step S303, the values of q and p are determined by the recognition rule of the time series. This step is commonly referred to as model scaling. The values of p and q may be determined by one skilled in the art by a variety of recognition rules, such as auto-correlation coefficient (ACF) and partial auto-correlation coefficient (PACF) graphs, red-cell information criterion (AIC), bayesian Information Criterion (BIC), thermodynamic diagrams, and the like.

For example, parameters p and q may be found using ACF and PACF graphs. ACF is a complete autocorrelation function that can provide autocorrelation values for any sequence with hysteresis values. In brief, it describes the degree of correlation between the current value of the sequence and its past values. The time series may contain trend, seasonal, periodic, residual, etc. components. ACF considers all of these components in finding correlations. PACF is a partial or partial autocorrelation function. Basically, it does not find the correlation of hysteresis like ACF with the current but the correlation of residual with the next hysteresis value. Thus, if there is any hidden information in the residual that can be modeled by the next lag, we may get a good correlation and we feature the next lag when modeling. For example, if the ACF map shows a tail-biting property from after the 1 st order, the PACF map shows a tail-biting property from after the 1 st order, so it is judged manually that the MA (1) model, ARIMA (1, 0, 1), is used.

If ACF and PACF are progressively smaller, this indicates that time series smoothing is required and the d-parameter is introduced.

According to another embodiment, AIC or BIC information may be used. And calculating the AIC or BIC information quantity of all combinations when p and q are less than or equal to 3, and taking the model order in which the AIC or BIC information quantity reaches the minimum.

After determining d, p, q of the ARIMA model in step S303, step S304 of training and testing the model is now reached. In step S304, the computer system trains and tests the ARIMA model with 70% of the historical data as training data and the remaining 30% as test data to obtain a final ARIMA model.

In one embodiment, the ARIMA model building process described above may be implemented using the R language. The ACF, PACF, AIC, BIC calculation, the training and the testing of the model and the like are realized by calling related functions in the R language. Those skilled in the art will appreciate that the inventive concept is not limited to the R language platform, but may be implemented in other languages known in the art and developed in the future, such as python, etc.

A user interface and corresponding user operations according to an exemplary embodiment of the present invention are described below in connection with fig. 4.

As shown in fig. 4, the user interface may include a time input portion 401 for a future time period to be predicted and a bearing shell input portion 402 for a bearing shell series to be predicted. The time input section 401 may be in the form of a slider. The user inputs time by sliding the slider. The time may be in days or weeks. It will be appreciated by those skilled in the art that the time input section 401 is not limited to a slider bar, but may be in the form of other input times.

The bushing input 402 may be in the form of a drop down menu. The drop down menu may list all bearing shell categories for selection by the user, or may list only bearing shell series. In the latter case, when the user selects a particular bearing shell series from a drop down menu of bearing shell series, the computer system will make a prediction for each bearing shell category of the selected bearing shell series.

The user interface also includes a data diagram display portion 403. In the example shown in fig. 4, the data map display portion 403 displays data maps 406, 407, and 408 corresponding to the three bearing shell types included in the selected bearing shell series. Preferably, different colors are used to represent the data maps 406, 407, and 408 for each bearing shell type. Each of the data maps 406, 407, and 408 includes two parts, a history data part 404 and a prediction data part 405, respectively. The historical data portion 404 refers to a map presented by data before the selected current time, and the predicted data portion 405 refers to a predicted data map of a future time period predicted by using a model. Different colors or shades may be used to distinguish the historical data portion 404 from the predictive data portion 405.

In addition, the user may calculate the safety stock based on the bushing demand predicted by the computer system. For example, the computer system may obtain current inventory data from a database and calculate a minimum safe inventory value in combination with the predicted bearing shell demand over a future period of time. When the difference between the real-time inventory and the minimum safety inventory value is within a predetermined threshold, a warning message is issued. Therefore, an inventory warning message can be sent to the user in advance, and inventory arrangement rationalization is realized.

The method predicts the future bearing bush demand by using the time sequence model, and has higher prediction accuracy than the conventional method which uses experience judgment. The user can make a more reasonable transmission plan and inventory plan according to the method, so that bearing bush backlog and economic loss caused by technical change are reduced.

FIG. 5 illustrates an exemplary configuration of a computer system in which embodiments according to the invention may be implemented. Computer system 500 is an example of a hardware device in which the above aspects of the invention may be applied. Computer system 500 may be any machine configured to perform processes and/or calculations. The computer system 500 may be, but is not limited to, a workstation, a server, a desktop computer, a laptop computer, a tablet computer, a Personal Data Assistant (PDA), a smart phone, an on-board computer, or a combination thereof.

As shown in FIG. 5, computer system 500 may include one or more elements that may be connected to or in communication with bus 502 via one or more interfaces. Bus 502 can include, but is not limited to, an industry standard architecture (Industry Standard Architecture, ISA) bus, a micro channel architecture (Micro Channel Architecture, MCA) bus, an Enhanced ISA (EISA) bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus. Computer system 500 may include, for example, one or more processors 504, one or more input devices 506, and one or more output devices 508. The one or more processors 504 may be any kind of processor and may include, but are not limited to, one or more general purpose processors or special purpose processors (such as special purpose processing chips). Input device 506 may be any type of input device capable of inputting information to a computing device and may include, but is not limited to, a mouse, keyboard, touch screen, microphone, and/or remote controller. Output device 508 may be any type of device capable of presenting information and may include, but is not limited to, a display, speakers, video/audio output terminals, vibrators, and/or printers.

The computer system 500 may also include or be connected to a non-transitory storage device 514, which non-transitory storage device 514 may be any storage device that is non-transitory and that may enable data storage, and may include, but is not limited to, disk drives, optical storage devices, solid state memory, floppy diskettes, flexible disks, hard disks, magnetic tape, or any other magnetic medium, compact disk or any other optical medium, cache memory, and/or any other memory chip or module, and/or any other medium from which a computer may read data, instructions, and/or code. Computer system 500 may also include Random Access Memory (RAM) 510 and Read Only Memory (ROM) 512. The ROM 512 may store programs, utilities or processes to be executed in a nonvolatile manner. The RAM 510 may provide volatile data storage and stores instructions related to the operation of the computer system 500. Computer system 500 may also include a network/bus interface 516 coupled to a data link 518. The network/bus interface 516 can be any kind of device capable of enabling communication with external devices and/or networksOr a system, and may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication device, and/or a chipset (such as bluetooth ^TM Devices, 1302.11 devices, wiFi devices, wiMax devices, cellular communication facilities, etc.).

The various aspects, embodiments, implementations, or features of the foregoing embodiments may be used singly or in any combination. The various aspects of the foregoing embodiments may be implemented by software, hardware, or a combination of hardware and software.

For example, the foregoing embodiments may be embodied as computer readable code on a computer readable medium. The computer readable medium is any data storage device that can store data which can thereafter be read by a computer system. Examples of a computer readable medium include read-only memory, random-access memory, CD-ROMs, DVDs, magnetic tape, hard drives, solid state drives, and optical data storage devices. The computer readable medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

For example, the foregoing embodiments may take the form of hardware circuitry. The hardware circuitry may include any combination of combinational logic circuits, clock storage devices (such as floppy disks, flip-flops, latches, etc.), finite state machines, memory such as static random access memory or embedded dynamic random access memory, custom designed circuits, programmable logic arrays, etc.

In one embodiment, a hardware circuit according to the present disclosure may be implemented by encoding a circuit description in a Hardware Description Language (HDL) such as Verilog or VHDL. The HDL description may be synthesized for a cell library designed for a given integrated circuit manufacturing technology and may be modified for timing, power, and other reasons to obtain a final design database that may be transferred to a factory for the production of integrated circuits by a semiconductor manufacturing system. Semiconductor manufacturing systems may produce integrated circuits by depositing semiconductor material (e.g., on a wafer that may include a mask), removing material, changing the shape of the deposited material, modifying the material (e.g., by doping the material or modifying the dielectric constant with ultraviolet processing), and so forth. An integrated circuit may include transistors and may also include other circuit elements (e.g., passive elements such as capacitors, resistors, inductors, etc.) and interconnections between transistors and circuit elements. Some embodiments may implement multiple integrated circuits coupled together to implement a hardware circuit, and/or may use discrete elements in some embodiments.

While certain specific embodiments of the invention have been illustrated in detail by way of example, it will be appreciated by those skilled in the art that the foregoing examples are intended to be illustrative only and not to limit the scope of the invention. It should be appreciated that some of the steps in the foregoing methods are not necessarily performed in the order illustrated, but they may be performed simultaneously, in a different order, or in an overlapping manner. Furthermore, one skilled in the art may add some steps or omit some steps as desired. Some of the components in the foregoing systems are not necessarily arranged as shown, and one skilled in the art may add some components or omit some components as desired. It will be appreciated by those skilled in the art that changes could be made to the embodiments described above without departing from the scope and spirit of the invention. The scope of the invention is defined by the appended claims.

Claims

1. A bushing demand prediction method, comprising:

acquiring historical data of tolerances of a crankcase and a crankshaft of a scanned engine;

determining bearing bush types and historical data of bearing bush demand corresponding to each bearing bush type according to the tolerance of a crankcase and a crankshaft;

establishing a plurality of time sequence prediction models for historical data of bearing bush demand according to each bearing bush type;

receiving input from a user regarding a predicted time period and a particular bearing shell type;

predicting the bearing bush demand in a specific time period before the predicted time period by using the plurality of time sequence prediction models respectively;

comparing the bearing bush demand predicted by using each time sequence prediction model in the specific time period with the real demand respectively to obtain each error;

determining a time sequence prediction model corresponding to the minimum error as an optimal time sequence prediction model; and

the optimal time series prediction model is utilized to predict the bearing shell demand during the received predicted time period for the particular bearing shell category.

2. The method of claim 1, wherein the particular time period is a previous week of the predicted time period.

3. The method of claim 1, wherein the plurality of time series prediction models comprises an integrated moving average autoregressive (ARIMA) model, a Bayesian Analysis Time Series (BATS) model, a trend error season (ETS) model, and a Holt-windows model.

4. The method of claim 1, further comprising:

acquiring bearing bush inventory data;

calculating a minimum safety stock value according to the bearing bush demand during the predicted time period and the bearing bush stock data; and

and when the difference between the current bearing bush stock data and the minimum safety stock value is within a preset threshold value, a warning message is sent out.

5. The method of claim 1, wherein prior to establishing a plurality of time series prediction models, the method further comprises preprocessing the historical data, the preprocessing comprising excluding weekdays and holidays and their corresponding data.

6. The method of claim 5, wherein the preprocessing further comprises excluding data in the historical data outside of a confidence interval that the confidence level is 95%.

7. The method of claim 3, wherein the ARIMA model is represented as ARIMA (p, d, q), where p is the number of autoregressive terms, q is the number of moving average terms, d is the number of differences made to make the time series of the historical data a stationary sequence, and building the ARIMA model comprises:

determining whether the time sequence of the historical data is a stable sequence by adopting a unit root test method;

if the sequence is a non-stationary sequence, performing differential processing on the time sequence, and checking again until the sequence becomes a stationary sequence, wherein the number of times of performing the differential is determined as the value of d;

determining the values of q and p through the recognition rule of the time sequence; and

the ARIMA model is trained and tested with 70% of the historical data as training data and the remaining 30% of the historical data as test data to obtain a final ARIMA model.

8. The method of claim 1, wherein the input of the predicted time period is received via a slider on a user interface representing the time period.

9. The method of claim 8, wherein the predicted period of time is in days or weeks.

10. The method of claim 8, wherein the user interface further comprises an input for selecting a particular bearing shell series from one or more bearing shell series, each of the one or more bearing shell series comprising one or more bearing shell categories, the method further comprising:

predicting bearing shell demand during the received predicted time period using the plurality of time series prediction models for each bearing shell category in the particular bearing shell family in response to a user selection of the particular bearing shell family;

displaying on the user interface historical data of bearing shell demand and a time-dependent plot of predicted bearing shell demand for each bearing shell category of the particular bearing shell series.

11. A non-transitory computer-readable medium having stored thereon computer-executable instructions that, when executed by one or more computing devices, cause the one or more computing devices to perform the method of any of claims 1-10.

12. A computer system, the computer system comprising:

at least one processor; and

at least one non-transitory computer-readable medium having stored thereon computer-executable instructions that, when executed by the at least one processor, cause the at least one processor to perform the method according to any of claims 1-10.