JP6566545B2 - Demand forecasting device, demand forecasting method, and demand forecasting program - Google Patents

Description

  The present invention relates to a demand prediction apparatus, a demand prediction method, and a demand prediction program.

  Technologies for predicting demand are known. For example, Patent Document 1 discloses a technique for predicting product demand by multiplying a planned number of product payouts by a coefficient corresponding to the length of a period until a scheduled delivery date and the expected degree of orders.

JP 2013-156893 A

With the technique disclosed in Patent Document 1, demand can be predicted with high accuracy. On the other hand, since the technology disclosed in Patent Document 1 does not consider the advance and postponement of the delivery date of the product, when the delivery date is evaluated monthly, there is a possibility that a month with high accuracy may occur.
An object of the present invention is to provide a demand forecasting device, a demand forecasting method, and a demand forecasting program that accurately forecast demand in a period divided every unit time such as every month.

According to the first aspect of the present invention, the demand prediction device is a demand prediction device that predicts the demand of a product in a prediction target period that is a period divided every unit time, and the product forecast that has been inquired A demand information storage unit that stores demand information in which a delivery date, a scheduled number to be paid out, and an expected degree of inquiry are associated with each other, and a first number corresponding to the expected level according to the expected number of products belonging to a forecast target period. And a value obtained by multiplying the planned number of products belonging to at least one prediction target period immediately before and immediately after the one prediction target period by a second coefficient corresponding to the expectation And a demand prediction unit that predicts the demand for the product by integrating the above.

According to the second aspect of the present invention, in the demand prediction device according to the first aspect , the first coefficient and the second coefficient are specified for each prediction target period, and the demand prediction unit The demand of the product is predicted using the first coefficient relating to one prediction target period and the second coefficient relating to at least one prediction target period immediately before and immediately after the one prediction target period.

According to the third aspect of the present invention, in the demand prediction device according to the first or second aspect , the first coefficient and the second coefficient are specified for each product type or each customer, The demand prediction unit predicts the demand for the product using the first coefficient and the second coefficient related to the product type or customer.

According to the fourth aspect of the present invention, the demand forecasting method is a demand forecasting method for forecasting the demand for a product in a forecast target period, which is a period divided every unit time. A demand forecasting device comprising a demand information storage unit that stores demand information in which a delivery date, a planned number of payouts, and an expected degree of inquiry are associated with each other, the expected degree of the expected number of products belonging to a forecast target period with a scheduled delivery date. A first step of multiplying a first coefficient according to the demand prediction device, wherein the demand forecasting device is configured to calculate the expected number of products belonging to at least one forecast target period immediately before and immediately after the one forecast target period. A second step of multiplying a second coefficient according to the degree, and the demand prediction device integrates the number calculated in the first step and the number calculated in the second step. And a third step of predicting demand.

According to the fifth aspect of the present invention, the demand prediction program stores a computer that stores demand information in which a scheduled delivery date of a product for which an inquiry has been made, a scheduled number to be paid out, and an expected degree of the inquiry are associated with each other, A value obtained by multiplying the planned number of products belonging to one forecast target period by a first coefficient corresponding to the expectation and at least one forecast target immediately before and immediately after the one forecast target period. The product is made to function as a demand prediction unit that predicts the demand of the product by adding the value obtained by multiplying the expected number of products belonging to the period by the second coefficient corresponding to the expectation.

  According to at least one aspect among the above aspects, the demand prediction device is configured to provide a product in one prediction target period based on a planned number of products belonging to at least one prediction target period immediately before and immediately after the one prediction target period. Forecast demand. Thereby, the demand prediction apparatus can predict the demand in the period divided for every unit time with high accuracy.

It is a schematic block diagram which shows the structure of the demand prediction apparatus which concerns on 1st Embodiment. It is a figure which shows the example of demand information. It is a figure which shows the example of track record information. It is a flowchart which shows the procedure of the demand prediction method which concerns on 1st Embodiment. It is a flowchart which shows the procedure of the demand prediction method which concerns on 2nd Embodiment. It is a schematic block diagram which shows the structure of the computer which concerns on at least 1 embodiment.

<< First Embodiment >>
Hereinafter, embodiments will be described in detail with reference to the drawings.
FIG. 1 is a schematic block diagram illustrating a configuration of a demand prediction apparatus according to the first embodiment.
The demand prediction apparatus 100 according to the present embodiment is an apparatus that predicts the total number of payouts per month for products for which inquiries have been made. That is, in this embodiment, “month”, which is a period divided every unit time per month, is set as the prediction target period. The demand prediction apparatus 100 includes a demand information storage unit 101, a demand prediction unit 102, a display control unit 103, a performance information storage unit 104, a coefficient specifying unit 105, and a coefficient storage unit 106.

The demand information storage unit 101 stores demand information associating a planned payout date and a planned number of products with inquiries.
Based on the demand information stored in the demand information storage unit 101, the demand prediction unit 102 determines the number of on-schedule payouts that are the number of products paid out as scheduled in the forecast target month and the products scheduled to be paid out in the next month of the forecast target. Are calculated as the number of products to be paid out in the month to be predicted, and the number of products to be paid out in the month to be predicted as the number of products to be paid out in the month to be predicted. The demand prediction unit 102 predicts the demand of the product for the prediction target month by calculating the sum of the on-schedule payout number, the advance payout number, and the postpaid payout number.
The display control unit 103 displays the demand prediction result of the product for each month predicted by the demand prediction unit 102 on a display device such as a display.

The track record information storage unit 104 stores track record information indicating the planned number of products for each past month and the actual number of products paid out. The planned number of products indicates the number of products for which a delivery date is set in the month at the time of product inquiry. The actual number of paid-out products indicates the number of products actually delivered in that month.
The coefficient specifying unit 105 specifies a coefficient used for demand prediction based on the record information stored in the record information storage unit 104. Specifically, the coefficient specifying unit 105 calculates the demand coefficient for calculating the number of demands by multiplying the scheduled number of months to be predicted and the number of advance payments by multiplying the scheduled number of the next month to be predicted. And a postponement coefficient for calculating the number of postponements by multiplying the predicted number of the previous month to be predicted. Note that the number of on-schedule payouts is calculated by subtracting the advance payout number and the late payout number from a value obtained by multiplying the planned number of forecast target months by the demand coefficient. The coefficient specifying unit 105 specifies a coefficient for each month.
The coefficient storage unit 106 stores the coefficient specified by the coefficient specifying unit 105.

Here, the demand information stored in the demand information storage unit 101 will be described.
FIG. 2 is a diagram illustrating an example of demand information.
The performance information according to the present embodiment is information for storing a product number, an inquiry date, a scheduled delivery date, and a planned number. The product number is a reference number for identifying the product. The product type and model are specified by the product number. The inquiry date indicates the date when the product was inquired. The scheduled delivery date indicates the delivery date set at the time of product inquiry. The planned number indicates the number of products to be delivered set at the time of product inquiry.
Thus, the demand information is information relating to product inquiries.

FIG. 3 is a diagram illustrating an example of performance information.
The performance information according to the present embodiment is information for storing a product number, an inquiry date, a scheduled delivery date, a planned number, a payout date, and a payout number. The payout date indicates the date when the product is actually delivered. The number of payouts indicates the number of products actually delivered.
Thus, the performance information is information indicating the relationship between product inquiries and payouts.

Operation | movement of the demand prediction apparatus 100 which concerns on this embodiment is demonstrated. The demand prediction apparatus 100 specifies a coefficient used for demand prediction before starting demand prediction.
FIG. 4 is a flowchart illustrating a procedure of the demand prediction method according to the first embodiment.
The coefficient specifying unit 105 of the demand prediction apparatus 100 reads the performance information from the performance information storage unit 104 (step S1). When the coefficient specifying unit 105 reads the result information, the coefficient specifying unit 105 performs the following steps S3 to S5 on the result information of the products having the same product number (step S2).
First, the coefficient specifying unit 105 determines that the scheduled delivery date is m months for each month (for example, each month from January to December) that is a target of demand prediction based on the performance information of the product number selected in step S2. the total number y _m scheduled number of products belonging to the payout date to calculate the total number Y _m of payout of products in month m (step S3). Next, the coefficient specifying unit 105, coefficient prediction error e expressed by the formula (1) shown below is minimum _{α 1 ~α t, β 1 ~β} t + 1, identifies a combination of gamma ₀ to? _T (Step S4). Here, α _m represents the demand coefficient for m month, β _m represents the forward coefficient for m month, and γ _m represents the backward coefficient for m month.

That is, the coefficient specifying unit 105, a multiplied by the demand factor alpha _m to the total number y _m of predetermined number of m months, those that have been multiplied by the accelerated coefficient beta _m to the total number y _m of predetermined number of m month and the month m subtracts the multiplied by the post-kill coefficient gamma _m the total number y _m of predetermined number, the following month predetermined number of the total number y _{m + 1} to accelerated coefficient beta _{m +} after ₁ to the total number y _m-1 as the multiplication and the previous month predetermined number of The predicted demand number is calculated by adding the product obtained by multiplying the decrement coefficient γ _m−1 . Then, the square sum of the difference between the actual payout Y _m and the predicted demand count obtained for each month is calculated as a prediction error e. The coefficient specifying unit 105 can specify a combination of coefficients that minimizes the prediction error using, for example, a generalized reduced gradient method (GRG). The forward coefficient and the forward coefficient are positive numbers smaller than the demand coefficient. The coefficient identifying unit 105 records the identified demand coefficient, forward coefficient, and backward coefficient for each month in the coefficient storage unit 106 (step S5).

When the demand coefficient, the forward coefficient, and the forward coefficient for each month of the product with each reference number are recorded in the coefficient storage unit 106, the demand prediction unit 102 reads the demand information from the demand information storage unit 101 (step S6). Next, the demand prediction part 102 performs the process of the following steps S8-S9 about the demand information of the product which has the same product number (step S7).
Forecast unit 102 first, based on the demand information of the selected product number in step S7, for each month to be predicted demand to calculate the total number x _m of the number of outstanding product (step S8). Next, demand prediction unit 102 uses the coefficients stored in the coefficient storage unit 106, by Equation (2) below to calculate the predicted demand count X _m product of each month (step S9).

That is, the demand prediction unit 102 is a coefficient (first coefficient) obtained by subtracting the forward coefficient β _m and the forward coefficient γ _m from the demand coefficient α _m to the planned number x _{m of} products whose scheduled delivery date belongs to m month. To calculate the number of on-schedule payouts in m months. In addition, the demand prediction unit 102 calculates the number of advance payments in m + 1 month by multiplying the expected number x _{m + 1} of products belonging to the month immediately after m month by the advance coefficient β _{m + 1} (second coefficient). . In addition, the demand prediction unit 102 multiplies the planned number x _m−1 of products belonging to the month immediately before the m month by a postponed coefficient γ _m−1 (second coefficient), so that m−1 month. Calculate the number of payouts. Then, the demand prediction unit 102 predicts the demand for the product in m month by integrating the on-schedule payout number in m month, the advance payout number in m + 1 month, and the post-payout number in m-1 month. .
When the demand prediction unit 102 predicts the monthly demand for each product, the display control unit 103 displays the predicted number of demands for each product month calculated by the demand prediction unit 102 on the display (step S10).

  Thus, according to this embodiment, the demand prediction apparatus 100 predicts the demand for a product in a certain month based on the planned number of products belonging to the month immediately before and after the certain month. Thereby, the demand prediction apparatus 100 can accurately predict the demand for products in each month even when the delivery date is advanced or delayed.

<< Second Embodiment >>
The demand forecasting apparatus 100 according to the second embodiment predicts the total number of products to be paid out per month in consideration of the expectation of inquiries in addition to the advance and postponement of the delivery date.
The demand prediction apparatus 100 according to the present embodiment differs from the first embodiment in the information stored in the record information storage unit 104, the coefficient storage unit 106, and the demand information storage unit 101. Moreover, the demand prediction apparatus 100 according to the present embodiment is different from the first embodiment in the calculation of the coefficient specifying unit 105 and the demand prediction unit 102.

The demand information stored in the demand information storage unit 101 stores an expected degree of inquiry in addition to a product number, an inquiry date, a scheduled delivery date, and a planned number. The degree of expectation is a value indicating a high possibility of receiving an order. The degree of expectation according to the present embodiment can be represented by, for example, r-level values from 1 to r in the order from the low possibility of receiving an order. The degree of expectation is determined based on the judgment of the person in charge who negotiates with the customer.
The record information stored in the record information storage unit 104 also stores an expected degree of inquiry in addition to the product number, the inquiry date, the scheduled delivery date, the planned number, the payout date, and the payout number.
The coefficient storage unit 106 stores coefficients for each month and each degree of expectation of the product with each product number.

FIG. 5 is a flowchart illustrating a procedure of the demand prediction method according to the second embodiment.
The coefficient specifying unit 105 of the demand prediction device 100 reads the record information from the record information storage unit 104 (step S101). When the coefficient specifying unit 105 reads the result information, the coefficient specifying unit 105 performs the following steps S103 to S105 on the result information of the products having the same product number (step S102).
First, the coefficient specifying unit 105, based on the actual information of the product number selected in step S102, for each month m targeted for demand, for each expected degree n, the expected number of products whose scheduled delivery date belongs to m month. the total number _{y m, n} of the payout date calculates the total number _{Y m, n} of the payout of products in month m (step S103). Next, the coefficient specifying unit 105 generates coefficients α _{1,1 to} α _{t, r} , β _{1,1 to} β _{t + 1, r} , γ ₀ that minimize the prediction error e represented by the following equation (3). _{, 1 to} γ _{t, r} are identified (step S104). However, α _{m, n} indicates the demand coefficient related to the inquiry of the expectation degree n in m month, β _{m, n} indicates the forward coefficient related to the inquiry of the expectation degree n in m month, and γ _{m, n} indicates the demand coefficient in the m month. The postponement coefficient related to the expectation n is shown.

  The coefficient specifying unit 105 records the demand coefficient, the forward coefficient, and the forward coefficient for each specified month and each expected degree in the coefficient storage unit 106 (step S105).

When the demand coefficient, the forward coefficient, and the forward coefficient for each month and each expected degree of the product of each reference number are recorded in the coefficient storage unit 106, the demand prediction unit 102 reads the demand information from the demand information storage unit 101 ( Step S106). Next, the demand prediction part 102 performs the process of the following steps S108-S109 about the demand information of the product which has the same product number (step S107).
First, the demand prediction unit 102 calculates the total number x _{m, n} of the planned number of products for each expectation level for each month that is the target of demand prediction based on the demand information of the product number selected in step S107 ( Step S108). Next, demand prediction unit 102 uses the coefficients stored in the coefficient storage unit 106, by Equation (4) below to calculate the predicted demand count _{X m} product of each month (step S109).

  When the demand prediction unit 102 predicts the monthly demand of each product, the display control unit 103 causes the display to display the predicted demand number of each product for each month calculated by the demand prediction unit 102 (step S110).

  As described above, according to the present embodiment, the demand prediction device 100 predicts the demand for a product in a certain month based on the planned number of products belonging to the month immediately before and after the certain month and the expected degree of inquiry. To do. Thereby, the demand prediction apparatus 100 can accurately predict the demand for products in each month.

As described above, the embodiment has been described in detail with reference to the drawings. However, the specific configuration is not limited to that described above, and various design changes and the like can be made.
For example, the demand prediction apparatus 100 according to the above-described embodiment predicts the demand for each month by integrating the on-schedule payout number, the advance payout number, and the late payout number, but is not limited thereto. The demand prediction apparatus 100 according to another embodiment may predict the demand for each month by, for example, integrating the on-schedule payout number and the advance payout number or the late payout number. The demand prediction device 100 can prevent a shortage of products by predicting demand based on the on-schedule payout number and the advance payout number. Moreover, the demand prediction apparatus 100 can prevent the occurrence of excessive inventory by predicting demand based on the on-schedule payout number and the postponed payout number.

  Moreover, although the demand prediction apparatus 100 which concerns on embodiment mentioned above calculates the coefficient linked | related with each month used as a demand prediction object, it is not restricted to this. For example, the demand prediction apparatus 100 according to another embodiment may predict a demand using a common demand coefficient, a forward coefficient, and a forward coefficient for each month. Specifically, the demand prediction apparatus 100 according to another embodiment may calculate only one demand coefficient, forward coefficient, and backward coefficient based on the performance information. In addition, like the demand prediction apparatus 100 according to the above-described embodiment, the coefficient associated with each month is calculated, and the demand is predicted using the coefficient. Demand can be predicted with high accuracy even during periods when the risk is likely to occur.

  Moreover, although the demand prediction apparatus 100 which concerns on embodiment mentioned above calculates a coefficient and the number of prediction demands for every product number, ie, every product type, it is not restricted to this. For example, the demand prediction apparatus 100 according to another embodiment may calculate a coefficient and the number of predicted demands for each customer.

  Moreover, although the demand prediction apparatus 100 which concerns on embodiment mentioned above calculates the number of forward payments and the number of backward payments based on the demand information of one month before and behind the month of prediction, it is not restricted to this. For example, the demand prediction apparatus 100 according to another embodiment may calculate the number of advance payments and the number of advance payments based on demand information for two months before and after the month to be predicted. In this case, the coefficient specifying unit 105 needs to calculate a forward coefficient related to the previous month and a forward coefficient related to the month before last as the forward coefficient. Further, in this case, the coefficient specifying unit 105 needs to calculate the postponement coefficient for the next month and the postponement coefficient for the following two months as the postponement coefficient.

  Moreover, although embodiment mentioned above makes "month" which is the period divided | segmented every unit time 1 month as a prediction object period, it is not restricted to this. For example, the demand prediction apparatus 100 according to another embodiment includes a “week” that is a period divided every week for a unit time, a “year” or a “year” that is a period divided every year for a unit time, and the like. The demand may be predicted for the prediction target period divided every other unit time.

  Moreover, although the demand prediction apparatus 100 which concerns on embodiment mentioned above is provided with the track record information storage part 104 and the coefficient specific | specification part 105, it predicts a demand based on the demand coefficient specified with the own apparatus, the advance coefficient, and the postponement coefficient, It is not limited to this. For example, the demand prediction apparatus 100 according to another embodiment does not specify a demand coefficient, a forward coefficient, and a forward coefficient, and predicts a demand based on the demand coefficient, the forward coefficient, and the forward coefficient specified by the other apparatus. May be.

  Moreover, although the demand prediction apparatus 100 which concerns on embodiment mentioned above specifies a demand coefficient, a forward coefficient, and a backward coefficient every time a demand is predicted, it is not restricted to this. For example, the demand prediction apparatus 100 according to another embodiment may predict the demand based on the demand coefficient, the forward coefficient, and the backward coefficient specified in the past demand prediction.

FIG. 6 is a schematic block diagram illustrating a configuration of a computer according to at least one embodiment.
The computer 900 includes a CPU 901, a main storage device 902, an auxiliary storage device 903, and an interface 904.
The above-described demand prediction apparatus 100 is mounted on a computer 900. The operation of each processing unit described above is stored in the auxiliary storage device 903 in the form of a program. The CPU 901 reads out the demand prediction program from the auxiliary storage device 903, expands it in the main storage device 902, and executes the above processing according to the demand prediction program. A storage area corresponding to each storage unit is secured in the auxiliary storage device 903.

  In at least one embodiment, the auxiliary storage device 903 is an example of a tangible medium that is not temporary. Other examples of the tangible medium that is not temporary include a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, and a semiconductor memory connected via the interface 904. When the demand prediction program is distributed to the computer 900 via a communication line, the computer 900 that has received the distribution may develop the demand prediction program in the main storage device and execute the above processing.

  In addition, the demand prediction program may be for realizing a part of the functions described above. Further, the demand prediction program may be a so-called difference file (difference program) that realizes the above-described function in combination with another program already stored in the auxiliary storage device 903.

DESCRIPTION OF SYMBOLS 100 Demand prediction apparatus 101 Demand information storage part 102 Demand prediction part 103 Display control part 104 Performance information storage part 105 Coefficient specific | specification part 106 Coefficient storage part

Claims (5)

A demand forecasting device for forecasting demand for products in a forecast target period, which is a period divided by unit time,
A demand information storage unit for storing demand information in which a scheduled delivery date of a product that has been inquired, a planned number to be paid out, and an expected degree of inquiries are associated;
A value obtained by multiplying the planned number of products belonging to one forecast target period by a first coefficient corresponding to the expectation and at least one forecast target immediately before and immediately after the one forecast target period. A demand forecasting device comprising: a demand forecasting unit for forecasting demand for the product by adding a value obtained by multiplying a planned number of products belonging to a period by a second coefficient corresponding to the degree of expectation. The first coefficient and the second coefficient are specified for each prediction target period,
The demand prediction unit uses the first coefficient related to the one prediction target period and the second coefficient related to at least one prediction target period immediately before and immediately after the one prediction target period. Forecast demand
The demand prediction apparatus according to claim 1 . The first coefficient and the second coefficient are specified for each product type or each customer, and the demand forecasting unit calculates the first coefficient and the second coefficient related to the product type or customer. demand prediction apparatus according to claim 1 or claim 2 for predicting the demand of the product used. A demand forecasting method for forecasting product demand in a forecast target period, which is a period divided by unit time,
A demand forecasting device having a demand information storage unit that stores demand information in which a planned delivery date, a planned number of products to be paid out, and an expected degree of inquiries are associated with each other, is a product that belongs to a forecast target period with a single scheduled delivery date. A first step of multiplying the expected number of the first factor according to the expectation;
A second step in which the demand forecasting device multiplies the planned number of products belonging to at least one forecast target period immediately before and immediately after the one forecast target period by a second coefficient corresponding to the expectation degree. ,
A demand prediction method comprising: a third step in which the demand prediction device predicts demand for the product by integrating the number calculated in the first step and the number calculated in the second step. Computer
A storage unit that stores demand information in which a scheduled delivery date of a product that has been inquired, a planned number to be paid out, and an expected degree of inquiries are associated,
A value obtained by multiplying the planned number of products belonging to one forecast target period by a first coefficient corresponding to the expectation and at least one forecast target immediately before and immediately after the one forecast target period. The demand prediction program for functioning as a demand prediction part which estimates the demand of the said product by integrating | accumulating the value which multiplied the 2nd coefficient according to the said expectation degree with the expected number of the products which belong to a period.

Images