JP4866537B2 - Product inventory quantity determination device - Google Patents

Description

  The present invention relates to determination of merchandise inventory at a logistics node.

  2. Description of the Related Art Conventionally, in the field of product distribution management or inventory management, various devices have been devised to supply an optimal amount of an optimal product to a customer at an optimal time at a minimum cost.

  In the conventional merchandise inventory amount determination method, the daily shipment forecast amount is directly calculated with reference to the daily shipment fluctuation amount. Therefore, if there is a sudden peak in the shipment amount in the past data, it is affected by this peak and the like, and as a result, there is a tendency to have excessive inventory, which may be disadvantageous in terms of cost.

  In addition, if the stock quantity is calculated while ignoring the peak or the like, there is a problem that a shortage or the like occurs in the product supply to the customer as a result.

  An object of the present invention is to solve the problems of the prior art.

More specifically, it is an object of the present invention to provide an apparatus for determining an inventory amount of a product that does not cause a shortage or the like to a customer at a low cost.

In order to solve the above problems, the present invention provides:
A product inventory amount determining device for determining an appropriate inventory amount of the product at the distribution base when the product from the production base is stocked at the distribution base and this product is shipped to a retail store,
The first shipment volume memory in which the number of packages is plotted on the vertical axis, the date is plotted on the horizontal axis, and the shipment volume data of products for each month in the past N years is stored , the cumulative shipment volume is plotted on the vertical axis, and the date is plotted on the horizontal axis. the delivery quantity data for each month sequentially in the past N years, first storage means and a second shipments memory the cumulative value for each month of adding is stored as a date each accumulated data in the past N years ,
Cumulative shipments on the vertical axis, taking the date on the horizontal axis, the current year (N + 1) successively the delivery quantity per month of the product in the cumulative value for each month of adding is stored as a date each accumulated data in the current year (N + 1) Second storage means
With
Furthermore, it comprises a rough variation function calculation means, a shipment amount calculation means, a ratio calculation means, and a regression analysis output means,
The rough variation function calculation means is
It said first outline variation curve of the first shipments low order Ferie component shipments data for each date in the memory is stored in the past N years month per shipments of past N years of storage means ( F) and obtaining means for storing in the rough variation function memory,
The ratio calculating means includes
Month every accumulation of said first storage means the date each accumulated data and the second of the year stored in the storage means in the second of the past stored in the shipping volume memory N-year (N + 1) Means for plotting data on the vertical and horizontal axes;
It is determined whether or not the plot data group is indicated by a substantially straight line. If the plot data group is indicated by a substantially straight line, a slope ratio (k) of the substantially straight line is obtained and set in the shipment amount calculating means ;
With
The regression analysis means includes
When the plot data group in plotted by the ratio calculation means is determined not shown in a substantially straight line, the year of the plot data group and date each accumulated data in the past N years among the plot data group (N + 1) more relative to the date each accumulated data, the date each accumulated data of date each subsequent point that the accumulated data and out shift the current year (N + 1) in the past N years later that has issued Re not a the Means for providing a limited period of time,
Means for obtaining a regression line between the monthly cumulative data for each past N years in these limited periods and the monthly cumulative data for each current year (N + 1) in these limited periods ;
Means for determining the respective slope ratios k for these regression lines and determining the correlation function r of these ratios k;
When showing the correlation function r is high correlation, and hand stage to set the inclination ratio k in the ratio memory used by the shipping amount calculation means
With
The shipment amount calculation means
The general variation curve which is stored as required by prior Symbol schematic variation function calculating means to the general variation function memory (F), a schematic variation curve in the ratio (k) the year is multiplied by (N + 1) of said ratio memory Means to seek,
When determining the predicted shipping amount of the specified time period consisting of a predetermined designated month and day specified date after the specified date Toko, the current year (N + 1) wherein each shipments data of the designated period in schematic variation curve in and summarized in that with a <br/> and means for calculating the predicted shipping amount of the specified time period in the year distribution centers (N + 1).

  As described above, according to the present invention, it is possible to determine an inventory amount of a product that does not cause a shortage to a customer at a low cost.

  Hereinafter, an embodiment of the present invention will be described with reference to FIGS.

  FIG. 1 shows a merchandise distribution route to which a physical distribution management method including an embodiment of a merchandise inventory quantity determination method of the present invention is applied. Here, the product includes, but is not limited to, body care products such as soaps, clothing and home care products, hygiene products such as diapers, various cosmetics, various commercial products, industrial chemical products, and the like. .

  As shown in FIG. 1, the goods are distributed in the distribution route as follows.

  In step S1, raw materials for the product are procured to the production base 11.

  In step S2, a product is produced and stocked by processing the raw material at the production site 11.

  In step S3, the goods are transported or supplied to the wholesaler or the distribution base 13 of the producer by a vehicle or the like.

  In step S4, the transported or supplied goods are stocked at the wholesale or distribution base 13.

  In step S5, a predetermined quantity of products is transported from the wholesale or distribution base 13 to the retail store 15 in response to a request from the retail store. Note that the retail store 15 in FIG. 1 collectively represents a large number of retail stores managed by the distribution base 13 or the like.

  In steps S6 and S7, the merchandise is displayed at the retail store and purchased by the consumer.

  In this distribution management method, the supply amount or shipment amount of goods from the production base 11 to the distribution base 13 is predicted, and the inventory quantity at the distribution base 13 is determined.

More specifically, this physical distribution management method is applied, for example, when a change in the shipment amount of a product at a current predetermined time is expected to be similar to a change in the shipment amount of a product at a certain past time.
That is, this method uses, for example, the fact that the change in the shipment amount of the mid-item product during a mid-year period of a certain year is similar to the change in the shipment amount of the same product in the same period of the previous year.

  FIG. 2 is a block diagram showing a physical distribution management apparatus that executes a physical distribution management method including an embodiment of the commodity inventory quantity determination method of the present invention. Various means other than the memory in this block diagram are realized by, for example, a computer program.

As shown in FIG. 2, the physical distribution management apparatus 20 includes a past shipment amount memory 21 that stores a shipment amount value (also referred to as a shipment amount ) of a predetermined product from a production base 11 to a distribution base 13 at a predetermined past time. . More specifically, the past shipment quantity memory 21 stores, for example, the production base 11 to the distribution base 13 for the same product in the same period (for example, the mid-year period) one year before the year in which the supply quantity or the shipment quantity is to be predicted. The shipment amount fluctuation value is stored.

As shown in FIG. 2, the past shipment amount memory 21 includes a first shipment amount memory 23 that stores daily shipment amounts in a predetermined period (for example, mid-year) in the past (N years), and the daily shipment amount. And a second shipment amount memory 25 for storing the accumulated value (variation thereof).

3, at which it is stored in the first shipments memory 23, indicating the daily shipments in a predetermined period (e.g. Hazime Ataru control timing) of the last (N years). In more detail, the example of FIG. 3 shows the fluctuation of the daily shipment amount between May 1 and August 31, which is the middle period of a certain product. In FIG. 3, the vertical axis represents the number of packages per day.

The upper curve N in FIG. 4 is a graph showing the cumulative shipment amount or cumulative supply amount at the mid-year period in N years, which is stored in the second shipment amount memory 25. The cumulative shipment amount is calculated by sequentially adding the daily shipment amount in FIG.

As shown in FIG. 2, the physical distribution management apparatus has a third shipment amount memory 27. The third shipment quantity memory 27 stores the product from the production base 11 in a predetermined time (for example, the first half of the mid-year) of the year (for example, the (N + 1) year following the N year) in which the shipment quantity is predicted. The accumulated value of the shipment amount to the distribution base 13 is stored. The lower curve (N + 1) in FIG. 4 shows the accumulated shipment quantity value stored in the third shipment quantity memory 27.

That is, each value of the upper curve N in FIG. 4 is stored in the second shipment amount memory 25, and each value of the lower curve (N + 1) is stored in the third shipment amount memory 27.

As shown in FIG. 2, the physical distribution management apparatus 20 has a ratio calculation unit 29. For example, the comparison calculation means 29 includes a cumulative shipment amount fluctuation value in the first half of the mid-year period (for example, from May 1 to July 10 in FIG. 4) stored in the third shipment amount memory 27, 2 The cumulative shipment amount fluctuation values in the first half of the mid-year (from May 1 to July 10) in N years stored in the shipment memory 25 are compared with each other.

  FIG. 5 shows the comparison operation performed by the comparison calculation means 29.

  That is, the ratio calculation means 29 plots each value using the shipping date as a parameter, with the cumulative shipment amount fluctuation value for N years and the cumulative shipment amount fluctuation value for N + 1 years as x and y coordinate axes, respectively. For example, in the graph of FIG. 4, the cumulative shipment amount is 100 (in thousands of boxes) in N year on June 12, whereas it is 80 in N + 1 year. Therefore, in the graph of FIG. A plot is attached at 80.

  In the graph of FIG. 5, as shown in FIG. 5, when the plot (with the cumulative shipment quantity fluctuation value in N years and the cumulative shipment quantity fluctuation value in N + 1 years) becomes a straight line, the shipment quantity transition in N + 1 years is N It can be expected to be k1 times the annual shipment volume. Here, k1 is the slope of the straight line in FIG.

  Therefore, the ratio calculation means 29 calculates the ratio k1 based on the plot as shown in FIG.

  This ratio k1 is stored in the ratio memory 29a.

  As shown in FIG. 2, the physical distribution management apparatus 20 includes a rough variation function calculation unit 31.

  The rough fluctuation function calculating means 31 is a rough fluctuation function or rough fluctuation tendency of daily fluctuation values of a predetermined product (for example, from May 1 to August 30) in the past (N years) shown in FIG. Calculate (seasonal trend curve). More specifically, the approximate variation function or the approximate variation tendency is represented by, for example, a Fourier series created by a low-order Fourier component of the daily variation value.

  As a low-order periodic component, it is appropriate to set it to the shortest, about one week.

  In FIG. 3, a waveform (schematic variation function) composed only of low-order Fourier components is indicated by a curve 310.

  The Fourier component and the waveform (rough variation function) composed of these low-order Fourier components are stored in the general variation function memory 31a.

The physical distribution management device 20 calculates a future shipment amount (for example, the latter half of the middle period in N + 1 years ) based on data from the ratio memory 29a and the rough variation function memory 31a ( also referred to as Fourier component memory 31a). First shipping amount calculation means 33 is provided.

  More specifically, when the first shipment amount calculation means 33 calculates the daily product supply amount thereafter on July 10, N + 1, for example, the approximate variation function F (t) from the approximate variation function memory 31a. Is multiplied by the ratio k1 from the ratio memory 29a to calculate the shipping amount after July 10 as k1 × F (t).

  FIG. 6 shows a curve N representing the rough variation function F (t) and a curve N + 1 representing the shipment amount (k1 × F (t)) calculated by the first shipment amount calculation means 33. Here, the case where k1 is smaller than 1 is shown.

  Here, for example, when determining the shipment amount for two weeks from July 10, the curve N + 1 of the period indicated by the arrow T from July 10 to July 24 in FIG. 6 is adopted or used.

This determined shipping amount (k1 × F (t)) is stored in the predicted shipping amount memory 33a.

  According to this determined shipment amount, for example, the approximate shipment amount from the production base 11 to the distribution base 13 in the second half of the middle period of N + 1 can be accurately predicted.

As shown in FIG. 2, the physical distribution management apparatus 20 has an additional shipment amount memory 34. This additional shipment quantity memory 34 is added to the standard shipment quantity calculated by the first shipment quantity calculation means 33 and sent to the distribution base 13 in response to daily additional or temporary requests from the retail store 15 or the like. The amount of additional shipment (for the product) to be shipped is stored.

The distribution management device 20 sends the total shipment amount (from the production base 11 to the distribution base 13) according to the determined shipment quantity from the predicted shipment quantity memory 33a and the additional shipment quantity from the additional shipment quantity memory 34. It has a transport vehicle type / number determination means 35 for determining the type or number of vehicles such as trucks to be transported. More specifically, the transportation vehicle type / number determination means 35 is based on the latest data in the predicted shipment amount memory 33a and the additional shipment amount memory 34, for example, two days after the determination date (for example, July 10) (for example, (July 12) is determined, and based on this determination data, the type and number of vehicles two days later are determined.

  As a result, for example, two days after the decision date, the type and number of vehicles for transporting the shipment product from the production base 11 to the distribution base 13 are determined.

The physical distribution management device 20 as shown in FIG. 2 further comprises a respective vehicle loading Product Type-stacking amount determination means 37. Each vehicle loaded product type / loading amount determination means 37 is based on the latest additional shipment amount (data of July 11) from the additional shipment amount memory 34 on the next day (July 11) of the determination date. If the vehicle is not fully loaded when loading (the required amount of the distribution base based on the latest additional shipment (data on July 11)) is loaded for the determined vehicle type / number The planned shipment amount after July 12 is incorporated into the shipment amount on July 12 so that the vehicle is fully loaded. As a result, the transportation vehicles can be fully loaded. As a result, the transportation vehicles can be reduced and the transportation cost can be reduced.

  As shown in FIG. 2, the physical distribution management apparatus 20 includes regression analysis means 39.

  In the calculation by the ratio calculation means 29, the regression analysis means 39 determines that the relationship between the cumulative shipment quantity fluctuation in N years and the cumulative shipment quantity fluctuation in N + 1 years is not the linear relationship shown in FIG. When a deviation occurs from the linear relationship from a certain point in time (FIG. 7)), the shipment quantity after the point at which the deviation occurs is determined by regression analysis.

  The regression analysis means 39 operates as follows, and predicts the shipping amount after the refraction point of the contrast line.

The regression analysis means 39 calculates the slope ratio k and the correlation function r (with the degrees of freedom corrected) by the least square method using the cumulative shipment quantities in N years and N + 1 years. As the cumulative shipment amount data used at this time, for example, the slope ratio k and the correlation coefficient r are calculated using data in which the period of the past two weeks, three weeks, and four weeks is limited. When the correlation coefficients of these three types of periods show a sufficiently high correlation (for example, 0.98 or more), the slope ratio k is used instead of the slope coefficient k1 of the ratio memory.

  As shown in FIG. 2, the third shipment memory 27 includes an initial shipment memory 27a and a repeat shipment memory 27b. The initial shipment memory 27a and the repeat shipment memory 27b are used, for example, for predicting a shipment amount for a product that is newly released in a certain year.

  FIG. 8 is a bar graph showing the cumulative shipment amount up to May 30 for a new product that was started on April 20, N + 1, for example. Here, the length below the alternate long and short dash line in each bar graph represents the amount initially delivered from the distribution base 13 to each retail store 15. That is, it represents the total quantity of products that are initially delivered (or shipped) to each retail store. Further, the length of the portion above the one-dot chain line in each bar graph represents the total of the second and subsequent distribution amounts (repeat distribution amounts) to each retail store.

  FIG. 9 is a graph obtained by extracting the cumulative value of the initial delivery amount in FIG. 8, and FIG. 10 is a graph obtained by extracting the cumulative value of the repeat delivery amount in FIG.

The initially Haini memory 27a, the accumulated value data of the initial Haini amount shown in FIG. 9 is stored. On the other hand, the repeat delivery memory 27b stores repeat delivery amount fluctuation data shown in FIG. As described above, each data in FIG. 9 and FIG. 10 is cumulative delivery amount data from the product release date.

  The physical distribution management apparatus includes an initial distribution store number memory 42 that stores a cumulative value of the number of retail stores that have distributed the new product at least once.

  FIG. 11 shows cumulative value data of the number of retail stores that have delivered the new product at least once, which is stored in the initial delivery store number memory 42.

  The physical distribution management apparatus shown in FIG. 2 calculates the future initial delivery amount based on the initial delivery amount fluctuation data from the initial delivery memory 27a and the initial delivery store number data from the initial delivery store number memory 42. An initial delivery amount prediction means 41 for prediction is provided.

More specifically, the initial delivery amount prediction unit 41 includes an initial delivery store / initial delivery amount comparison unit 43. The initial delivery shop / initial delivery quantity comparison means 43 compares the cumulative value data of the initial delivery quantity from the initial shipment memory 27a with the initial delivery shop fluctuation data from the initial delivery shop number memory 42. To do.

FIG. 12 shows a comparison between the initial delivery amount fluctuation data and the initial delivery shop fluctuation data performed by the initial delivery shop / initial delivery quantity comparison means 43.

That is, the initial distribution store / initial distribution amount comparison means 43 checks whether the initial shipment amount and the initial shipment number are in a linear relationship or a proportional relationship, with the number of days from the new product release date as a parameter.

When the initial shipping quantity / initial shipping quantity comparison means 43 has a linear relationship or a proportional relationship between the initial shipping quantity fluctuation and the initial shipping quantity fluctuation, the slope k2 in the graph of FIG. Represents. Then, at a certain point in the initial shipment quantity fluctuation process, the problem of predicting the future initial shipment quantity fluctuation results in the problem of predicting the fluctuation of the number of initial shipment shops.

  As shown in FIG. 2 again, the initial delivery amount predicting means 41 determines the initial shipping store number variation function based on the initial shipping store number variation data from the initial shipping store number memory 42. A determination unit 45 is included.

  By the way, the initial shipping number fluctuation function is expected to be expressed by a first order lag step response function y (x) = c−bexp [−ax].

Accordingly, the initial shipping store number fluctuation function determining means 45 is based on the data stored in the initial shipping store number memory 42 (FIG. 11), and the temporary delay step response function y (x) = c−bexp [−ax. ] Coefficients a, b, and c are determined. For this determination, for example, a non-linear least square method is used. At that time, for example, a spline interpolation method is used.

As shown in FIG. 2, the initial delivery quantity predicting means 41 further includes a function G (t) from the initial delivery shop number variation function determining means 45 and a comparison from the initial delivery shop / initial delivery quantity comparison means 43. Based on the coefficient k2, a future initial delivery amount calculation means 47 for calculating a future initial delivery amount variation is provided. More specifically, the calculation unit 47 multiplies the function G (t) from the initial shipment number variation function determination unit 45 by the coefficient k2 from the initial distribution store / initial distribution amount comparison unit 43, for example, FIG. 9, the future initial shipment cumulative quantity (k2 × G (t)) after the prediction time t0 in FIG.

  As shown in FIG. 2, the physical distribution management apparatus further includes a repeat delivery amount predicting unit 49 that predicts and calculates the repeat delivery amount after a predetermined time point based on the repeat delivery memory 27b.

The repeat delivery quantity predicting means 49 has a repeat delivery quantity function determining means 51. As shown in FIG. 10, when the repeat shipment quantity is proportional to the number of days from the release date, the repeat delivery quantity function determining means 51 calculates the slope or coefficient k3 in the data represented by the graph of FIG.

  As shown in FIG. 2, the repeat delivery amount predicting means 49 calculates the future repeat delivery amount after a predetermined prediction time based on the repeat delivery amount function from the repeat delivery amount function determining means 51. More specifically, the future repeat delivery amount is calculated by substituting the number of days elapsed from the release date for the number of days parameter of the repeat shipment amount function.

  As shown in FIG. 2, the physical distribution management apparatus adds the future initial delivery amount from the future initial delivery amount calculation means 47 and the future repeat delivery amount from the future repeat delivery amount calculation means 53 to the future. There is a second shipment amount calculation means 55 for calculating the shipment amount.

  Thereby, after a predetermined period (for example, several days) has elapsed from the date of release of the new product, the subsequent shipment amount to the distribution base 13 can be predicted relatively accurately.

  Note that the predicted shipping amount from the second shipping amount calculation means 55 is also stored in the predicted shipping amount memory 33a.

Further, as shown in FIG. 2, the physical distribution management apparatus is provided with a merchandise stock quantity determining means device 57 for executing the merchandise stock method embodiment of the present invention. This merchandise inventory quantity determination means device 57 decides a merchandise inventory quantity for preventing a stockout or the like from occurring in the inventory quantity at the distribution base 13, for example.

As shown in FIG. 2, the commodity stock quantity determination means 57 includes a first shipment quantity memory 23, a rough variation function calculation means 31, and a rough variation function memory 31a.

Further, the commodity inventory quantity determining means 57 is based on the fluctuation data (shipment quantity) from the first shipment quantity memory 23 and the outline fluctuation function from the outline fluctuation function memory 31a, and the deviation width data comprising the difference between them. Is included.

  FIG. 13 is a graph showing the deviation width (difference) of the daily shipment amount with respect to the rough variation function curve 310 as calculated by the deviation width data calculation means 58.

  FIG. 14 is a graph showing deviation width data with reference to the rough fluctuation curve 310 in FIG.

  The deviation width data (FIG. 14) is stored in the deviation width data storage means 58a.

Referring again to FIG. 2, the commodity stock determination unit 57 includes a deviation width distribution calculation unit 59. The deviation width distribution calculation means 59 regards the deviation width data in FIG. 14 as a probability distribution and determines the probability distribution parameter. More specifically, the calculation means 59 uses the gamma distribution in FIG.

とみなしてその分布パラメータα、βを決定する。図１４（Ｂ）はこのようにして決定されたガンマ分布の確率密度を表す。図１４（Ｂ）において、縦軸はズレ量或いは乖離量を表し、横軸は確率密度を表す。

And the distribution parameters α and β are determined. FIG. 14B shows the probability density of the gamma distribution thus determined. In FIG. 14B, the vertical axis represents the amount of deviation or deviation, and the horizontal axis represents the probability density.

  The probability distribution and its distribution parameters α and β are stored in the shift width distribution memory 59a.

  As shown in FIG. 2, the commodity stock quantity determining means 57 is a normal deviation width calculating means 61 for obtaining a deviation width (normal deviation width) that can normally occur based on the probability distribution stored in the deviation width distribution memory 59a. Have The normal deviation width calculation means 61 calculates, for example, the standard deviation of the probability distribution, and obtains the value of the normal deviation width Δm based on the standard deviation.

  This normal deviation amount is stored in the normal deviation width memory 61a.

  In FIG. 13 and FIG. 14, a fluctuation exceeding the normal fluctuation range Δm is determined as a sudden fluctuation in this physical distribution management method. These sudden fluctuations are dealt with by collecting sales information and the like in detail and in detail.

The embodiment of the merchandise inventory quantity determination method of the present invention is executed by the merchandise inventory quantity determination means device 57.

  The normal deviation width can be obtained as a standard deviation of the γ distribution function.

According to this embodiment, the normal deviation width viewed from the rough variation function (see FIG. 14A).
(B)) is the product inventory quantity. Accordingly, the inventory quantity at the distribution base 13 can be made relatively small.

FIG. 1 is an explanatory diagram of a product supply route to which a physical distribution management method including a product inventory method of the present invention is applied. FIG. 2 is a block diagram of a physical distribution management apparatus that executes a physical distribution management method including the commodity inventory method of the present invention. FIG. 3 is a graph showing the daily shipment amount of commodities managed by the physical distribution management method at a predetermined time in a certain year (N years). FIG. 4 is a graph showing the cumulative shipment volume of products managed by the logistics management method at a predetermined time in a certain year (N years) and the cumulative shipment volume of the same products in the same period of the following year ((N + 1) years). It is. FIG. 5 is a graph showing the relationship between the cumulative shipment amount for N years and the cumulative shipment amount for N + 1 years, which is created based on the graph of FIG. FIG. 6 is a graph showing fluctuations in the future shipment amount calculated and predicted by this physical distribution management method. FIG. 7 shows the cumulative shipment quantity for N years and the cumulative shipment quantity for N + 1 years when there is an abnormality or change in the fluctuation trend of the shipment quantity due to external fluctuation factors or the like in the predetermined period of N + 1 years shown in FIG. It is a graph which shows the relationship. FIG. 8 is a graph showing a change history of the cumulative shipment amount from the point of sale for a newly released product. FIG. 9 is a graph showing changes in the cumulative shipment amount shown in FIG. 8 for the first supply or distribution (that is, initial distribution or initial shipment) to each retail store. FIG. 10 is a graph showing fluctuations in the cumulative shipment amount shown in FIG. 8 for the second or subsequent supply or delivery (that is, repeat delivery or repeat shipment) to each retail store. FIG. 11 is a graph showing the change in the cumulative number of retail stores (initial cumulative number of stores) in which the first supply or distribution (that is, initial distribution or initial shipment) has been performed. FIG. 12 is a graph showing the relationship between the cumulative initial shipment amount and the cumulative initial number of shipping stores derived from the data of FIGS. 9 and 11. FIG. 13 is a graph showing the approximate fluctuation range of the product shipment amount shown in FIG. FIG. 14A is a graph showing a deviation width of the shipping amount as viewed on the basis of the low-order Fourier series curve as the rough variation function curve in FIG. 3 or FIG. FIG. 14B is a diagram showing a gamma distribution corresponding to the deviation amount of the shipment amount.

Claims (3)

A product inventory amount determining device for determining an appropriate inventory amount of the product at the distribution base when the product from the production base is stocked at the distribution base and this product is shipped to a retail store,
The first shipment volume memory in which the number of packages is plotted on the vertical axis, the date is plotted on the horizontal axis, and the shipment volume data of products for each month in the past N years is stored , the cumulative shipment volume is plotted on the vertical axis, and the date is plotted on the horizontal axis. , said delivery quantity data items from each month sequentially in the past N years, first storage means having a second shipments memory the cumulative value for each month of adding is stored as a date each accumulated data in the past N years When,
Cumulative shipments on the vertical axis, taking the date on the horizontal axis, the current year (N + 1) successively the delivery quantity per month of the product in the cumulative value for each month of adding is stored as a date each accumulated data in the current year (N + 1) Second storage means
With
Furthermore, it comprises a rough variation function calculation means, a shipment amount calculation means, a ratio calculation means, and a regression analysis output means,
The rough variation function calculation means is
It said first outline variation curve of the first shipments low order Ferie component shipments data for each date in the memory is stored in the past N years month per shipments of past N years of storage means ( F) and obtaining means for storing in the rough variation function memory,
The ratio calculating means includes
Month every accumulation of said first storage means the date each accumulated data and the second of the year stored in the storage means in the second of the past stored in the shipping volume memory N-year (N + 1) Means for plotting data on the vertical and horizontal axes;
It is determined whether or not the plot data group is indicated by a substantially straight line. If the plot data group is indicated by a substantially straight line, a slope ratio (k) of the substantially straight line is obtained and set in the shipment amount calculating means ;
With
The regression analysis means includes
Said ratio when the plot data group plotted is determined not shown substantially straight line calculating means, date each cumulative data and the year of said plot data group in the past N years among the plot data group ( N + 1) and the date each cumulative data, not a Re out by subsequent points are after that it out shift said the date each accumulated data in the past N years year (N + 1) month each of a plurality of relative and cumulative data Means for providing a limited period;
Means for obtaining a regression line between the monthly cumulative data for each past N years in these limited periods and the monthly cumulative data for each current year (N + 1) in these limited periods ;
Means for determining the respective slope ratios k for these regression lines and determining the correlation function r of these ratios k;
When showing the correlation function r is high correlation, and hand stage to set the inclination ratio k in the ratio memory used by the shipping amount calculation means
With
The shipment amount calculation means
The general variation curve which is stored as required by prior Symbol schematic variation function calculating means to the general variation function memory (F), a schematic variation curve in the ratio (k) the year is multiplied by (N + 1) of said ratio memory Means to seek,
When determining the predicted shipping amount of the specified time period consisting of a predetermined designated month and day specified date after the specified date Toko, the current year (N + 1) wherein each shipments data of the designated period in schematic variation curve in merchandise inventory amount determination apparatus characterized by comprising a <br/> and means for calculating the predicted shipping amount of the specified time period in the year distribution centers (N + 1). The second storage means is
First and shipment memory distribution load number of accumulated data for each month during a predetermined period when the first shipment of new products stored for the retail outlet,
An initial distribution store number memory storing initial distribution store number data of the new product,
further,
When it is determined whether the cumulative data of the number of deliveries per month in the predetermined period at the time of the initial shipment and the initial dispatch store number data are in a linear relationship or a proportional relationship, and are in a linear relationship or a proportional relationship Finds the slope (k2), and the first-order lag step response function G of the initial shipping number data
2. The commodity inventory quantity determining device according to claim 1, wherein the commodity quantity determining unit is an initial delivery quantity predicting unit that obtains an initial shipment quantity of each retail store after the predetermined period by multiplying (t) and the slope (k2). . Means for obtaining a difference between the shipment quantity per day in the past N years and the approximate variation function as a deviation width;
The distribution of the deviation width is

Means for determining the distribution parameters α and β,
A means for obtaining the deviation width based on the obtained distribution parameters α and β, and determining a shipment quantity for each month in the past N years exceeding the deviation width as a sudden change. The product inventory amount determination device according to 1 or 2.

Images