JPWO2018135526A1 - Standard stock quantity calculation system, standard stock quantity calculation method and program - Google Patents

Abstract

  The reference inventory quantity calculation device (10) includes an acquisition unit (200), an order probability density function generation unit (401), and a reference inventory quantity calculation unit (402). The acquiring unit (200) acquires a target service rate representing a target value of a service rate which is a rate capable of responding to an order without causing a product shortage and order acceptance result data representing an order acceptance result. The order probability density function generation unit (401) generates an order probability density function that is a probability density function representing the likelihood of the number of orders, based on the order record data acquired by the acquisition unit (200). The standard inventory quantity calculation unit (402) is a standard inventory that realizes a service rate equal to or higher than the target service rate acquired by the acquisition unit (200) based on the order probability density function generated by the order probability density function generation unit (401). Calculate the number.

Description

  The present invention relates to a standard inventory quantity calculation system, a standard inventory quantity calculation method, and a program.

  Many conventional stock inventory number calculation systems that perform inventory control of products statistically process the order receipt results and calculate the standard inventory number where the shortage occurrence probability falls below the tolerance value. Approximate. In addition, even in the case where it is considered that the distribution of the number of orders does not follow the normal distribution, the difference from the normal distribution is only corrected.

  Patent Document 1 discloses a technique for creating an expression representing the distribution of the number of orders received. In this technology, when the distribution of the number of orders does not follow the normal distribution, the coefficient of variation of the graph representing the distribution of the number of orders received, the kurtosis is used to evaluate the deviation between the distribution of the number of orders received and the normal distribution, The equation to be expressed is corrected.

Unexamined-Japanese-Patent No. 2006-139796

  By the way, the needs of customers are diversified and high-mix low-volume production is required. As a result, there is an increasing number of cases where the order probability density function indicating the distribution of the number of orders does not follow the normal distribution. The technique disclosed in Patent Document 1 corrects an expression representing a normal distribution using kurtosis or the like. Therefore, when the distribution of the number of orders is left-right asymmetry with respect to the average value of the number of orders, there is a problem that the distribution of the actual number of orders and the distribution represented by the equation after correction are large. Therefore, it is not possible to calculate an appropriate standard inventory number, which causes an occurrence of a shortage or an increase in unnecessary inventory.

  The present invention has been made in view of the above problems, and it is an object of the present invention to provide a reference inventory number calculation system, a reference inventory number calculation method and a program capable of calculating an appropriate reference inventory number.

  In order to achieve the above object, a reference inventory quantity calculation system according to the present invention includes an acquisition means, an order probability density function generation means, and a reference inventory quantity calculation means. The acquiring means acquires a target service rate representing a target value of a service rate which is a rate capable of responding to an order without causing a product shortage, and order acceptance result data representing an order acceptance result. The order acceptance probability density function generating means generates an order acceptance probability density function which is a probability density function representing the likelihood of the number of orders accepted, based on the order acceptance result data acquired by the acquisition means. The reference inventory number calculation means calculates, based on the order probability density function generated by the order probability density function generation means, a reference inventory number for realizing a service rate equal to or higher than the target service rate acquired by the acquisition means.

  According to the present invention, the order probability density function generating means can calculate an appropriate reference stock quantity by generating an order probability density function representing the likelihood of the number of orders received based on the order acceptance data of the product. .

Functional block diagram of reference stock quantity calculation system according to the embodiment of the present invention Diagram showing various masters stored in the input data storage unit of the standard inventory quantity calculation system according to the embodiment of the present invention Diagram showing various masters and data stored in the processing data storage unit of the standard inventory quantity calculation system according to the embodiment of the present invention Hardware configuration diagram of the standard inventory quantity calculation system according to the embodiment of the present invention Flow chart of reference stock quantity calculation processing according to the embodiment of the present invention A diagram showing a display example of an operation mode selection screen according to the embodiment of the present invention Flow chart of generation process of order probability density function according to the embodiment of the present invention Diagram representing an example of order receipt data according to an embodiment of the present invention A figure showing an example of order probability data according to the embodiment of the present invention Diagram showing an evaluation example of an approximate expression based on the method of least squares according to an embodiment of the present invention A diagram representing an evaluation example of an approximate expression based on the maximum likelihood method according to the embodiment of the present invention A figure showing an example of transition of the number of stocks concerning an embodiment of the invention A diagram showing an example of the AIC correction amount according to the embodiment of the present invention Flow chart of calculation process of reference stock quantity according to the embodiment of the present invention Flow chart of creation processing of standard inventory adjustment proposal according to the embodiment of the present invention Flow chart of decision processing of standard inventory adjustment proposal according to the embodiment of the present invention Diagram showing an example of a base-based inventory adjustment plan according to the embodiment of the present invention Diagram showing an example of a standard inventory adjustment plan according to an embodiment of the present invention A figure showing an example of a list table of a standard inventory adjustment proposal concerning an embodiment of the invention Diagram showing an example in which the standard inventory adjustment plan according to the embodiment of the present invention is rearranged A diagram showing another example of the standard inventory adjustment plan according to the embodiment of the present invention The figure showing another example of the standard inventory adjustment proposal concerning the embodiment of the present invention. A figure showing an example of a standard inventory adjustment proposal concerning a modification of the present invention Diagram showing the result of verification of Example 1 Plot of order probability density function of product A in Example 1 Plot of order probability density function of product B in Example 1 Plot of order probability density function of product C in Example 1 Plot of order probability density function of product D in Example 1 Plot of order probability density function of product E in Example 1 Plot of order probability density function of product F in Example 1 Plot of order probability density function of product G in Example 1 Plot of order probability density function of product H in Example 1 Plot of order probability density function of product I in Example 1 Plot of order probability density function of product J in Example 1 Diagram showing the result of verification in Example 2 Diagram showing the result of verification in Example 2

Embodiment 1
The standard inventory quantity calculation system according to the present embodiment is a product based on the target service ratio, which is the target value of the service ratio that is the ratio that can correspond to the order without causing a shortage in the product, and the product order record data. Is a system to calculate the standard stock quantity of Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  The reference inventory quantity calculation system 1 according to the present embodiment includes a reference inventory quantity calculation apparatus 10 which is an apparatus for calculating a reference inventory quantity, and an input apparatus 100 which outputs information input by a user to the reference inventory quantity calculation apparatus 10. , A display device 600 for displaying information, and an external storage device 700 which is a storage device external to the reference inventory quantity calculation device 10.

  As shown in FIG. 1, the reference inventory number calculation device 10 has an acquisition unit 200 for acquiring the condition for calculating the reference inventory number, a storage unit 300 for storing various information, and a processing unit 400 for calculating the reference inventory number. , And an output unit 500 that outputs the calculation result.

  The input device 100 is an external input device of the reference inventory quantity calculation device 10, and includes a keyboard. The input device 100 outputs information representing characters or numerical values input by the user to the reference inventory number calculation device 10. This information includes an instruction to execute the calculation process of the standard stock quantity.

  The acquisition unit 200 is connected to the external storage device 700, acquires data necessary for calculation of the standard stock quantity, and stores the data in the storage unit 300. The acquisition unit 200 is also connected to the input device 100, acquires data from the input device 100, and stores the data in the storage unit 300. Furthermore, the acquisition unit 200 outputs, to the processing unit 400, an instruction to execute the process of calculating the standard inventory quantity acquired from the input device 100. The acquisition unit 200 functions as an acquisition unit that acquires various data input from the external storage device 700 and the input device 100.

  Further, the reference stock quantity calculation device 10 creates an adjustment proposal for adjusting by increasing or decreasing the reference stock quantity that has been calculated once, selects one of the created adjustment plans, and adjusts the reference stock quantity. The basic inventory quantity calculation device 10 includes a target inventory quantity mode for creating an adjustment proposal based on a designated target inventory quantity, and a target stock value mode for creating an adjustment proposal based on a designated target inventory value. It has a mode. The target stock quantity is an index set by the user in consideration of the balance between the shortage risk and the unnecessary stock risk. Further, since the unnecessary stock accumulated more than necessary is a defective asset, the target inventory amount is an index set by the user such that the amount of the defective asset is equal to or less than a target amount. The user can select an operation mode from the input device 100. The acquisition unit 200 acquires information indicating the operation mode selected by the user via the input device 100.

  In addition, the external storage device 700 connected to the acquisition unit 200 stores master data (hereinafter, simply referred to as a master) that manages various types of information including data necessary for calculating the standard inventory quantity. The master includes a production capacity master including production capacity data indicating the ability to produce a product within a period targeted for calculation of the standard inventory quantity, a production plan master including production plan data, a product master including a product unit price, a fixed period An order receipt actual master including the actual value of the number of orders received for each product, an inventory master including the number of products in stock, and the like are included.

  The storage unit 300 includes an input data storage unit 301 that stores data acquired from the input device 100 or the external storage device 700 via the acquisition unit 200, and a processing data storage unit 302 that stores data calculated by the processing unit 400. Prepare.

  As shown in FIG. 2, the input data storage unit 301 is a product master including a product unit price, a target service rate master including a target service rate, a target inventory amount master including a target inventory amount, and a target inventory quantity master including a target inventory quantity. Cumulative order acceptance probability density function including setting data such as production capacity master including production capacity data, distribution model master including known distribution model data, approximate expression for generating cumulative order probability density function to be described later, and distribution division number A setting master, a production plan master including production plan data, and an order actual result master including an actual value of the number of orders received for each fixed period are stored. The order probability density function is a function that represents the probability of the number of orders received as a probability. The acquisition unit 200 stores various data acquired from the external storage device 700 in the input data storage unit 301. The target service rate, the target inventory amount, and the target inventory number are input by the user via the input device 100. The reference inventory number calculation device 10 may acquire known distribution model data, various setting data used for generating the cumulative order probability density function, etc. from the external storage device 700, or store them in the storage unit 300 in advance. You may leave it.

  The processing data storage unit 302, as shown in FIG. 3, includes an order probability density function master including an order probability density function for each product generated by the processing unit 400, a reference inventory quantity master including a reference inventory quantity for each product, and a standard Stores the standard inventory adjustment proposal master including the inventory adjustment proposal. The processing data storage unit 302 also stores an AIC correction amount α, which is a correction amount for correcting an AIC (Akaike's Information Criterion: Akaike's information amount reference) to be described later. The AIC correction amount α is generated by an AIC correction amount calculation unit 800 described later.

  Note that AIC is an index for evaluating a statistical model, and in the present embodiment, a plurality of order probability density functions generated by order probability density function generation unit 401 are evaluated, and an appropriate order probability density function is evaluated. Used to select In addition, AIC is generally defined by AIC = -2lnL + 2K, where the maximum likelihood is L and the number of free parameters is K. And the one with the smallest AIC is considered to be a suitable statistical model.

  The processing unit 400 generates an order probability density function generating unit 401 that generates an order probability density function that conforms to the distribution of order reception data for each product, a reference inventory number calculation unit 402 that calculates a reference inventory number for each product, and a target If the standard inventory quantity calculated based on the service rate is larger than the target inventory quantity, or if the standard inventory amount, which is the amount equivalent to the standard inventory quantity, is larger than the target inventory value, create a proposal for adjusting the standard inventory quantity A reference inventory adjustment unit 403, and an AIC correction amount calculation unit 800 that generates an AIC correction amount α used to correct the AIC for selecting the order probability density function.

  The order acceptance probability density function generation unit 401 receives the order acceptance result data included in the order acceptance result master of the input data storage unit 301, the distribution model data included in the distribution model master, and the approximate expression included in the cumulative order acceptance probability density function setting master, An order probability density function is generated based on setting data such as the distribution division number. Specifically, the order acceptance probability density function generation unit 401 performs approximation by the least squares method and approximation by the maximum likelihood method, and based on order acceptance result data, a plurality of order acceptance probability density functions representing the likelihood of the number of orders accepted. Generate a candidate. Then, the order probability density function generation unit 401 calculates AIC and evaluates each order probability density function. At this time, the AIC is corrected by the AIC correction amount α generated by the AIC correction amount calculation unit 800. Details of the calculation of the AIC and the correction will be described later. Then, the order acceptance probability density function generation unit 401 selects an order acceptance probability density function to be used for calculation of the standard stock quantity by selecting one having a minimum AIC from among a plurality of candidates. Then, the order probability density function generation unit 401 stores the determined order probability density function in the processing data storage unit 302.

  The standard inventory quantity calculation unit 402 acquires the target service rate from the input data storage unit 301, and acquires the order probability density function from the processing data storage unit 302. Then, the standard inventory quantity calculation unit 402 calculates, for each product, a standard inventory quantity that achieves a service rate equal to or higher than the target service rate, and stores the calculated standard inventory quantity in the processing data storage unit 302.

  The reference inventory adjustment unit 403 acquires the number of planned productions, the target inventory number, the product price, and the target inventory amount from the input data storage unit 301, and acquires reference inventory number data and order acceptance probability density function data from the processing data storage unit 302. . Then, the standard inventory adjustment unit 403 calculates an adjustment proposal of the reference inventory number based on the acquired data, and selects an appropriate adjustment proposal from among the calculated adjustment proposals. The standard inventory adjustment unit 403 stores the selected adjustment proposal in the processing data storage unit 302. Then, the reference inventory adjustment unit 403 applies the adjustment proposal to the reference inventory number stored in the processing data storage unit 302 to adjust the reference inventory number, and displays the adjusted reference inventory number through the output unit 500. Output to the device 600.

  The AIC correction amount calculation unit 800 calculates an AIC correction amount α to be used for correcting the evaluation formula for selecting the order acceptance probability density function to be used for calculating the reference inventory number from a plurality of candidates for order acceptance probability density functions. The data storage unit 302 stores the AIC correction amount α. The AIC correction amount α is used to evaluate an order probability density function described later. As an example, the AIC correction amount calculation unit 800 updates the AIC correction amount α at the beginning of the month and stores the updated AIC correction amount α in the processing data storage unit 302. Details will be described later.

  The output unit 500 is connected to the display device 600, and outputs, to the display device 600, the reference inventory number calculated by the processing unit 400 and the proposed adjustment of the reference inventory number.

  The display device 600 includes a liquid crystal display, a projector, and the like, and is a display device that displays the information output from the output unit 500.

  Next, the hardware configuration of the reference stock quantity calculation device 10 will be described with reference to FIG. The reference inventory quantity calculation device 10 includes a central processing unit (CPU) 11 which executes various processes, a random access memory (RAM) 12 which is a volatile memory, and a read only memory (ROM) 13 which is a non-volatile memory. And a hard disk drive 14 for storing various information, a network card 15 as an interface with a communication line, and a graphic card 16 for converting information into a video signal and outputting it.

  The CPU 11 reads out the program stored in the hard disk drive 14 to the RAM 12 and executes the program to execute various processes described later.

  The RAM 12 is used as a work area of the CPU 11.

  The ROM 13 stores a control program for basic operation of the standard stock quantity calculation device 10 executed by the CPU 11, a BIOS (Basic Input Output System), and the like.

  The CPU 11, the RAM 12 and the ROM 13 cooperate to function as a processing unit 400.

  The hard disk drive 14 stores information, and stores various masters, an AIC correction amount α, a program executed by the CPU 11, and the like. The hard disk drive 14 functions as the storage unit 300.

  The network card 15 is an interface with the communication line, and is communicably connected to the input device 100 and the external storage device 700. The network card 15 may be directly connected to the input device 100 and the external storage device 700, or may be connected via a network such as the Internet, an intranet, a VPN (Virtual Private Network), or a LAN (Local Area Network). It is good. The network card 15 functions as an acquisition unit 200.

  The graphic card 16 converts information into a video signal and outputs it, converts the information acquired or generated by the processing unit 400 into a video signal, and outputs the video signal to the display device 600. The graphic card 16 functions as the output unit 500.

  Next, the process of calculating the standard stock quantity by the standard stock quantity calculation device 10 having the above configuration will be described with reference to the drawings. Here, the case where the user selects the target stock quantity mode will be described. An example of the operation mode selection screen is shown in FIG. In the selection screen shown in FIG. 6, the target service rate is set to 0.95 by the user. When the user selects the "target stock number" shown in FIG. 6, inputs the target stock number 49000, and presses the "reference stock number calculation button", the reference stock number calculation process shown in FIG. 5 is started. . Note that a part of various data such as order acceptance result data and inventory data stored in the external storage device 700 is acquired by the acquisition unit 200 as necessary and is already stored in the input data storage unit 301. Do.

  When the reference inventory number calculation process starts, the acquisition unit 200 acquires the target service rate and the target inventory number shown in FIG. 6 input by the user (step S1100), and stores them in the input data storage unit 301.

  Next, the order probability density function generation unit 401 takes in the order record data from the input data storage unit 301, and generates an order probability density function representing the likelihood of the number of orders received for each product based on the order record data (steps) S1200). In step S1200, the order probability density function generation unit 401 functions as an order probability density function generation unit that generates an order probability density function. Also, this step S1200 functions as an order probability density function generating step. The details of the process of generating the order probability density function will be described below with reference to the flowchart shown in FIG.

  When the process of generating the order acceptance probability density function starts, the order acceptance probability density function generating unit 401 takes in the order acceptance result data from the input data storage unit 301 (step S1210). An example of the received order result data is shown in FIG. The order acceptance result data shown in FIG. 8 is order acceptance result data regarding one product. The order record data includes the number of orders received in a fixed period (in this case January), the maximum number of orders received Smax during the period targeted by the received order record data, the average number of orders received μ in the period, the standard deviation σ in the period including.

  Next, the order probability density function generation unit 401 generates order probability density function by the least squares method, so the order actual result data is divided into M sections for each fixed order number, and the order probability is determined for each order number category. Calculate The number of divisions M is a parameter that determines how to divide sample data when the order probability density function is generated by the least squares method. The value of the division number M may be stored in advance in the storage unit 300 or may be set by the user from the input device 100. Moreover, it is also possible to set multiple division number M.

  Here, the order acceptance probability for each division when the division number M = 10 is described with reference to FIG. The order acceptance probability for each category, specifically, indicates the category number indicating the number of each category, the category number indicating the upper limit of the number of orders accepted for each category, the order probability for indicating the probability of order acceptance for each category, and the category number 1 The cumulative order probability indicating the cumulative order probability up to each category is included.

  The order probability density function generation unit 401 sets a value Smax / M obtained by dividing the maximum number of orders received Smax by the number of divisions M shown in FIG. Since Smax shown in FIG. 8 is 160 and the division number M = 10, the division unit of the division order number is 160/10 = 16. Therefore, the division order number of division number 1 is from 0 to 16, the division order number of division number 2 is from 17 to 32, and the division order number of division number 3 is from 33 to 48. In the same manner, the number of divided orders received for the division number 10 is 145 to 160.

  The order probability density function generation unit 401 statistically processes the received order result data and counts the number of orders received for each category of the number of orders received. Then, the order probability for each category is calculated by dividing the number of orders received for each category by the total number of orders received. The order probability density function generation unit 401 may obtain order probability data statistically processed in advance from the order record master of the input data storage unit 301. Here, it is assumed that the order acceptance probability of each category is the value shown in FIG. The order probability density function generation unit 401 sequentially adds up the order probability of each section to generate a cumulative order probability.

  Next, the order probability density function generation unit 401 generates candidates for the order probability density function by approximation using the least squares method based on the calculated order probability for each section (step S1220).

  Specifically, the order probability density function generation unit 401 generates a cumulative order probability density function F (x) as an approximate expression that approximates the cumulative order probability. The order probability density function generation unit 401 acquires a plurality of approximate expressions approximating the cumulative order probability from the cumulative order probability density function setting master, applies parameters for each of them, and generates a cumulative order probability density function F (x).

For example, the following (Equation 1) to (Equation 6) are approximation equations that approximate the cumulative order acceptance probability.
· Linear first-order F (x) = β _0i + β 1 _i x (Equation 1)
· Linear second order F (x) = β _0i + β 1 _i x + β ² _i x ² + S _i (x) (2)
- Linear tertiary _{F (x) = β 0i +} β 1i x + β 2i x 2 + β 3i x 3 + S i (x) ··· ( Equation 3)
- Linear next six _{F (x) = β 0i +} β 1i x + β 2i x 2 + β 3i x 3 + β 4i x 4 + β 5i x 5 + β 6i x 6 + S i (x) ··· ( Equation 4)
Log F (x) = β _0i + β _1i * ln (x) + S _i (x) (Equation 5)
Exponent F (x) = β _0i + β _1i * exp (−x) + S _i (x) (6)
Here, i is the number of the divided section. Further, S _i (x) is a spline function and is expressed by the following equation.
S _i (x) = a + bx + cx ² (Equation 7)
In addition, you may use approximation formulas other than said example, such as a linear fourth order and a linear fifth order, as an approximation formula which approximates cumulative order acceptance probability.

The order probability density function generation unit 401 acquires the distribution division number D from the accumulated order probability density function setting master, and generates an accumulated order probability density function F (x) for each section divided into the acquired distribution division number D. Then, a spline function S _i (x) is added to each of the above-described approximate expressions in order to smooth the joints in each section. Specifically, when the distribution division number D is 1, S _i (x) is 0. In the case where the distribution division number D is 2, for example, when F (x) is generated using (Expression 2), since the parameter and the spline function are different for each section, the following is obtained.

Similarly, in the case where F (x) is generated using (Expression 2) when the distribution division number D is 3, the following is obtained.

  Next, the order probability density function generation unit 401 derives the order probability density function f (x) corresponding to each of the accumulated order probability density functions F (x) thus generated by (Equation 10). .

Next, the order acceptance probability density function generation unit 401 generates candidates for the order acceptance probability density function by the maximum likelihood method (step S1230). Specifically, the order probability density function generation unit 401 obtains distribution model data from the distribution model master stored in the input data storage unit 301, and approximates the standard deviation σ and the average number of orders received μ to those distribution models. Assign to an expression. The distribution model data includes, for example, distribution models represented by approximation formulas (Expression 11) to (Expression 14) below.
- normal distribution f (x) = 1 / √ (2πσ 2) * exp (- (μx) 2 / 2σ 2) ··· ( Equation 11)
Log-normal distribution f (x) = 1 / √ (2πσx) * exp (− (ln (x) −μ) ² / 2σ ² ) (12)
Rayleigh distribution f (x) = x / σ ² * exp (−x ² / 2σ ² ) (Equation 13)
・ Uniform distribution

  The order acceptance probability density function f (x) generated by the least square method and the maximum likelihood method described above is a candidate for the order acceptance probability density function used to calculate the reference stock number described later. The order probability density function generation unit 401 stores the generated order probability density function candidate in the order probability density function master of the processing data storage unit 302.

Next, the order probability density function generation unit 401 evaluates the plurality of order probability density functions f (x) generated in step S 1220 and step S 1230 (step S 1240). Specifically, the order probability density function generation unit 401 calculates AIC using (Expression 15).

The first term on the right side of (Equation 15) is a parameter that represents the maximum log likelihood of the generated order probability density function. If it represents by the maximum likelihood L mentioned above, this maximum log likelihood is -2lnL. The order probability density function generation unit 401 calculates, for example, the maximum log likelihood by calculating a squared error. In the specific example, assuming that the number of divided orders received in FIG. 9 is x _i , the occurrence probability is y _i, and x _i and y _i and the number M of divisions are substituted into (Expression 15), the maximum log likelihood is calculated. Since the number M of divisions is 10 in the case shown in FIG. 9, the order probability density function generation unit 401 performs this calculation for the division numbers 1 to 10 and adds 10 square errors to obtain maximum log likelihood. Calculate the degree.

  Further, in the case of the least squares method, the number of free parameters K is the product of the degree of the approximation formula and the number D of distribution divisions. Therefore, in the case of linear first order, exponent, and logarithm, K = 1 * D. Further, in the case of the linear second order, K = 2 * D, in the case of the linear third order, K = 3 * D, and in the case of the linear sixth order, K = 6 * D. On the other hand, in the case of the maximum likelihood method, the number of free parameters K is the order of the approximation formula.

  FIGS. 10A and 10B show examples of AIC evaluation results when the division number M is M = 10 and the distribution division number D is D = 1, 2, 3, 4, 5. FIG. FIG. 10A is an AIC of an order probability density function generated by the least squares method. FIG. 10B is an AIC of an order probability density function generated by the maximum likelihood method. As shown in FIGS. 10A and 10B, the order probability density function generation unit 401 calculates AIC for each of the number of distribution divisions D and the types of order probability density functions.

  Then, the order probability density function generation unit 401 selects an order probability density function with the smallest AIC (step S1250). In the example shown in FIGS. 10A and 10B, the order probability density function generating unit 401 generates an AIC from the approximate equation of logarithm by the least squares method, since the AIC is 90 when the AIC is 89 or less. Select an order probability density function. Then, the order probability density function generation unit 401 stores the selected order probability density function in the processing data storage unit 302.

The order probability density function generation unit 401 may calculate the corrected AIC 2 by adding the AIC correction amount stored in the processing data storage unit 302 to the AIC in step S 1240. Specifically, AIC2 is calculated by the following (Equation 16). That is, the order acceptance probability density function generation unit 401 adds the AIC correction amount α to the AIC of the order acceptance probability density function. In this case, the order probability density function generation unit 401 selects an order probability density function that minimizes AIC2.
AIC2 = AIC + α (Equation 16)

  Here, processing in which the AIC correction amount calculation unit 800 calculates the AIC correction amount α will be described. At the beginning of each month, the AIC correction amount calculation unit 800 is represented by the planned value of the cumulative number of productions represented by the solid line 1001 with black triangle marks shown in FIG. The cumulative number of orders received Then, the AIC correction amount calculation unit 800 calculates a predicted value of the number of used inventory for which the number of inventory represented by the difference between the accumulated production number of the current month and the accumulated order number is the smallest.

The AIC correction amount calculation unit 800 compares the reference stock quantity calculated using each division quantity M of each approximation formula according to the above-mentioned least squares method for all products with the predicted value of the above-mentioned stock usage quantity and calculates the standard The combination of the type of approximation formula and the number of divisions M that makes the difference between the number of stocks and the predicted value of the number of stocks used become minimum or maximum is determined. Furthermore, the AIC correction amount calculation unit 800 calculates an AIC for each combination of the type of approximate expression and the number of divisions M. Then, the AIC correction amount calculation unit 800 calculates the AIC correction amount α of the combination of the type of the approximate expression and the division number M in which the difference between the reference inventory number and the predicted value of the inventory usage number is minimized, Calculate using. Further, the AIC correction amount calculation unit 800 uses the AIC correction amount α of the combination of the type of the approximate expression and the division number M that maximizes the difference between the reference inventory number and the predicted value of the inventory usage number using (Expression 18) Calculate.
Minimum AIC correction amount α = AIC (min) -AIC (i) -Z (17)
AIC correction amount α of the maximum value α = AIC (max) −AIC (i) + Z (Expression 18)

  Here, AIC (i) is the AIC of the approximate expression to which the AIC correction amount α is to be added, and a value calculated using (Expression 15) is used. AIC (min) is the smallest AIC value among the combinations of the types of all approximate expressions generated by the order probability density function generation unit 401 and the number of divisions M, and AIC (max) is the maximum among them Is the value of AIC. Further, Z is a constant stored in advance in the storage unit 300. For example, Z = 10. However, this Z may be input by the user via the input device 100 and stored in the storage unit 300. In the process of calculating the AIC correction amount α, the AIC correction amount calculation unit 800 functions as an evaluation value correction amount calculation unit that calculates a correction amount for each evaluation value of a plurality of candidates for the order probability density function.

  The AIC correction amount α calculated in this manner is stored in the processing data storage unit 302 for each combination of the type of approximate expression and the number M of divisions, as shown in FIG. If there is the AIC correction amount α calculated in the past, the value calculated this time is overwritten and updated. Then, in step S 1240 of the process of generating the order probability density function described above, the order probability density function generation unit 401 adds the corresponding AIC correction amount α for each combination of the type of approximation used and the number of divisions M, Calculate AIC2 which is AIC after correction.

  Next, referring back to FIG. 5, the standard inventory quantity calculation unit 402 performs the standard inventory quantity calculation process using the order probability density function stored in the process data storage unit 302 in step S1250 (step S1300). . In step S1300, the reference stock quantity calculation unit 402 functions as a reference stock quantity calculation unit that calculates the reference stock quantity. Step S1300 functions as a standard stock quantity calculation step. The calculation process of the standard stock quantity will be described with reference to the flowchart shown in FIG.

  When the calculation process of the reference stock number starts, the reference stock number calculation unit 402 takes in the order probability density function used for the calculation processing of the reference stock number from the processing data storage unit 302 (step S1310).

  Next, the standard inventory quantity calculation unit 402 determines the number of production plans from the production plan master of the input data storage unit 301 and also includes production capacity data including lead time (hereinafter referred to as LT) from production instruction to production completion and production frequency. From the production capacity master (step S1320). Then, using the order probability density function acquired in step S1310, the reference inventory number calculation unit 402 calculates the reference inventory number based on the captured production plan number, LT, and production frequency.

  Specifically, the standard inventory quantity calculation unit 402 calculates the safety inventory quantity and the cycle inventory quantity based on the production plan quantity, the order probability density function to be used, the LT, and the target service rate. The number of safety stocks is the number of stocks with allowance for preventing out-of-stock in consideration of the variation in daily shipment volume. The cycle stock number is a stock number which has a margin for preventing stockout in consideration of the LT.

The standard inventory quantity calculation unit 402 calculates the safety inventory quantity using (Expression 19) (step S1330).

The standard inventory quantity calculation unit 402 calculates the value of a in (Expression 19) so as to satisfy (Expression 20).

Further, the standard inventory quantity calculation unit 402 calculates the cycle inventory quantity using (Expression 21) (step S1340).

  Next, the reference inventory number calculation unit 402 calculates the number obtained by adding the calculated safety inventory number and the cycle inventory number as the reference inventory number (step S1350). The reference inventory number calculation unit 402 stores the calculated reference inventory number for each product in the processing data storage unit 302, and shifts the processing to step 1400 in FIG.

  Next, the processing unit 400 determines whether it is necessary to adjust the reference inventory number calculated in step S1300 (step S1400). Depending on the setting of the target service rate, the base stock quantity calculated above may cause a large amount of unnecessary stock. In order to prevent the generation of the unnecessary inventory, it is necessary to confirm on the screen shown in FIG. 6 that the reference inventory quantity is equal to or less than the target inventory quantity set in advance by the user. To that end, the processing unit 400 fetches target inventory quantity data from the input data storage unit 301, and compares the target inventory quantity with the reference inventory quantity calculated in step S1300. The processing unit 400 stores the value of the difference c (= target inventory number−reference inventory number) in the standard inventory adjustment plan master in the processing data storage unit 302. If the difference c is a positive value, the processing unit 400 determines that there is no need for adjustment (step S1400: No), and transitions the process to step S1600. On the other hand, when the difference c is a negative value, it is determined that the adjustment of the standard stock number is necessary (step S1400: Yes), the processing unit 400 shifts the process to step S1500.

  The process of creating the standard inventory adjustment plan in step S1500 will be described with reference to the flowchart shown in FIG. When the process of creating the standard inventory adjustment plan starts, the standard inventory adjustment unit 403 takes in the value of the difference c from the processing data storage unit 302 (step S1510).

  Next, the reference inventory adjustment unit 403 executes a decision processing of a reference inventory adjustment proposal representing to which product this difference c is to be distributed to eliminate the difference (step S1520). For the preparation of this standard inventory adjustment plan, known optimization methods such as GA (Genetic Algorithm) method and annealing method are used. The GA method prepares multiple "individuals" where the solution candidate is expressed by a gene, preserves high fitness "individuals", and updates low fitness "individuals" to new "individuals" It is a method of searching for a suboptimal solution by repeating gene recombination among each other. When the GA method is applied, a gene group given to each product corresponds to an individual, and a standard inventory adjustment proposal to be created corresponds to a solution. On the other hand, the annealing method is a method of obtaining an optimal solution using random numbers.

  Here, an example of determination processing of the standard inventory adjustment proposal using the GA method will be described with reference to FIG. The standard inventory adjustment unit 403 creates a standard inventory adjustment plan that eliminates the P types of differences c. The standard inventory adjustment unit 403 repeatedly performs the above process while limiting the number of times specified by the user or the number of times that can be processed in the specified processing time, in order to create a final adjustment plan in a finite time. . Then, the standard inventory adjustment unit 403 selects the adjustment proposal having the highest evaluation point among the plurality of adjustment proposals. Thus, the reference inventory adjustment unit 403 does not use an infinite processing time to find an optimal solution, but uses a finite processing time to find a semi-optimal solution.

  Here, the number of times of processing is B times. The reference inventory adjustment section 403 sets the value j of the processing frequency counter for counting the processing frequency B to j = 1. Further, the reference inventory adjustment unit 403 sets the value i of the number counter for counting the number P of the reference inventory adjustment proposal group to be created to i = 1. (Step S1521).

Next, the standard inventory adjustment unit 403 creates a base standard inventory adjustment plan as a basis of the standard inventory adjustment plan (step S1522). Specifically, the standard inventory adjustment section 403 regards each product as one gene for a product group constituted by n pieces of products, and calculates a gene value for each product using (Equation 22). The reference inventory number i is the reference inventory number of the product i calculated by the reference inventory number calculation unit 402 in the calculation processing of the reference inventory number in step S1300 shown in FIG.

Next, the standard inventory adjustment unit 403 calculates the distribution ratio of allocating the difference c to each product using (Equation 23) for all the P standard inventory adjustment plans.

Next, the standard inventory adjustment unit 403 calculates the number of adjustments for each product using (Expression 24).
Adjustment number of product i = proportion ratio of product i * difference c (Equation 24)

  An example of the base standard inventory adjustment plan created by the above process is shown in FIG. Here, it is assumed that the sum of the standard inventory numbers of the target product group calculated in step S1300 is 50,000, the target inventory number is 49,000, and the difference c = −1,000.

  In the example shown in FIG. 16, the sum of the base stock numbers is 50,000, and the base stock number of the product 1 is 1,000. Therefore, the gene value of the product 1 calculated by (Expression 22) is 0.02, and the distribution ratio of the product 1 calculated by (Expression 23) is 0.02. Therefore, the adjustment number of the product 1 calculated by (Expression 24) is −20 (= −1,000 * 0.02). Therefore, the adjusted standard inventory quantity of product 1 is 980. The same applies to products 2 and later. The adjustment proposal prepared above is considered as the base standard inventory adjustment proposal.

  Next, the standard inventory adjustment unit 403 creates P adjustment proposal groups (step S1523). Although the standard inventory adjustment unit 403 calculates the gene value of each product using (Equation 22) in the process of step S 1522, here, it is 0.00 or more and 1.00 or less in 0.01 in the gene value. Random numbers are given to calculate the number of adjustments for each product, and the standard inventory number after adjustment. Then, the reference inventory adjustment unit 403 calculates the distribution ratio i of the product i using (Equation 23) by the gene value using random numbers, and calculates the adjustment number of the product i using (Equation 24). At this time, the upper limit value of the gene value is set to | (difference c) | / (reference stock number of gene i), and when the upper limit value is exceeded, the upper limit value is set to the gene value.

  An example of the standard inventory adjustment plan created by giving a random number of 0.00 or more and 1.00 or less in 0.01 steps to the gene value is shown in FIG. The standard stock quantity for each product shown in FIG. 17 is the same as the example shown in FIG. However, since the gene value is calculated based on random numbers, the gene value and the distribution ratio are different from those shown in FIG. Therefore, the number of adjustments for each product is also different from the numerical value shown in FIG.

  The reason for creating the standard inventory adjustment plan by giving the gene value a random number of 0.00 or more and 1.00 or less in steps of 0.01 in this way is to create a more suitable suboptimal solution within a finite time by the GA method It is for.

  The reference inventory adjustment unit 403 repeats the process of step S1523 P times (step S1524: NO, step S1525, step S1522, step S1523). That is, P new adjustment proposals shown in FIG. 17 are created by giving new 0.01-increment random numbers of 0.00 or more and 1.00 or less as gene values of n products. Then, the standard inventory adjustment unit 403 creates P adjustment plans together with the base standard inventory adjustment plan created in step S1522.

If the standard inventory adjustment section 403 determines that P adjustment plans have been created (step S1524: YES), it next evaluates the created P adjustment plans (step S1526). Specifically, the reference inventory adjustment unit 403 calculates the out-of-stock occurrence probability using (Expression 25). Further, the surplus inventory quantity expected value is calculated using (Equation 26).

  The standard inventory adjustment unit 403 assigns a rank with a high selection priority to the adjustment proposal having a low occurrence probability of shortage. In addition, when there is an adjustment proposal with a similar occurrence probability of shortage, the reference inventory adjustment unit 403 assigns a rank with a high selection priority to an adjustment proposal with a small surplus inventory number expected value. For example, when there are a plurality of adjustment proposals in which the difference in the occurrence probability of out-of-stock occurrences is within 1%, the standard inventory adjustment unit 403 selects an adjustment proposal having the smallest surplus inventory number expected value among them.

  An example of a list of adjustment plans created by the standard inventory adjustment unit 403 is shown in FIG. This list includes the gene value for each product, the probability of occurrence of shortage of the adjustment proposal, the surplus inventory number expected value, and the rank of the adjustment proposal. In addition, the adjustment number for each product may be included instead of the gene value for each product. The standard inventory adjustment unit 403 ranks the adjustment proposals based on the shortage occurrence probability and the surplus inventory number expected value as described above. Then, the reference inventory adjustment unit 403 stores the reference inventory adjustment proposal in which the rank of the selection priority ranks first among the P adjustment proposals in the processing data storage unit 302 (step S1527).

  Referring back to step S1528 in FIG. 15, the reference inventory adjustment unit 403 determines whether the process from step S1522 to step S1527 has been performed B times or more (step S1528). If it is determined that the process has not been performed B times or more (step S1528: No), the reference inventory adjustment section 403 proceeds to step 1540.

  The standard inventory adjustment unit 403 rearranges the adjustment plans to which the ranks have been assigned, in ascending order of rank (step 1540). At this time, "the adjustment proposal" from rank 1 to e-1 is called elite. e is a parameter that determines the number of adjustment plans to be set as elite. The value of e may be stored in advance in the storage unit 300 or may be input from the input device 100 by the user.

  The example which rearranged "the adjustment proposal" shown in FIG. 18 in rank ascending order is shown in FIG. The adjustment proposal to the top e rank of the rank belongs to the elite.

  The standard inventory adjustment unit 403 creates new “adjustment proposals” by giving new gene values to “adjustment proposals” other than the adjustment proposals belonging to the P−e elite. Specifically, the standard inventory adjustment unit 403 randomly selects two adjustment proposals, randomly replaces the genes of the selected adjustment proposals, and creates a new adjustment proposal. Next, the standard inventory adjustment unit 403 performs processing of replacing the inventory adjustment proposal of rank i with a new adjustment proposal.

  The reference inventory adjustment unit 403 performs processing to set the value of the number counter to i = e + 1 in order to replace the “adjustment proposal” other than elite (step 1541).

  In the example shown in FIG. 20, the standard inventory adjustment unit 403 prepares a new adjustment proposal using the adjustment proposal 2 and the adjustment proposal e. Specifically, the standard inventory adjustment unit 403 selects the gene value of an individual based on adjustment proposal 2 for product 1 and product 3, and selects the gene value of an individual based on adjustment proposal e for product 2 and product N. doing. The standard inventory adjustment unit 403 repeats the process of creating a new adjustment proposal P−e times (i = e + 1 to i = P), and creates a new adjustment proposal (step 1542).

  Returning to step S1543 in FIG. 15, the reference inventory adjustment unit 403 determines whether the process of step S1542 has been performed P-e times or more (step S1543). If it is determined that the process has not been performed more than P−e times (step S1543: No), the reference inventory adjustment section 403 performs processing to set the value of the number counter to i = i + 1 (step S1544), step S1542 Repeat the process of

  On the other hand, when it is determined that the above process has been repeated P−e times (step S1543: Yes), the reference inventory adjustment unit 403 performs a process of setting the value of the processing number counter to j = j + 1 (step 1529), and step S1526 Transition to

  The reference inventory adjustment unit 403 repeats the above processing until the value of the processing number counter becomes j ≧ B. In other words, the adjustment proposals from rank 1 to e that belong to the elite are preserved without change, the adjustment proposals from rank e + 1 to P are replaced with new adjustment proposals, and reranking is performed including the new adjustment proposals. An example of the adjustment proposal created by repeating this process is shown in FIG.

  On the other hand, when it is determined that the above process has been repeated B times (step S1528: Yes), the reference inventory adjustment unit 403 determines the adjustment proposal with the highest evaluation among the prepared B * P adjustments. Confirm as. (Step S1545). Specifically, the standard inventory adjustment unit 403 determines the final adjustment proposal by comparing the occurrence probability of shortage of B * P adjustment proposals with the surplus inventory expected value. Above, the standard inventory adjustment part 403 complete | finishes the decision processing of a standard inventory adjustment proposal.

  Referring back to FIG. 14, the standard inventory adjustment unit 403 stores the standard inventory adjustment plan decided at step S 1520 (step S 1530). Above, the standard stock adjustment part 403 finishes the preparation processing of the standard stock adjustment plan. Next, referring back to FIG. 5, the reference inventory adjustment unit 403 stores the adjusted reference inventory quantity based on the reference inventory adjustment proposal created in step S1500 in the processing data storage unit 302 (step S1600). In step S1600, the standard inventory adjustment unit 403 functions as a standard inventory adjustment unit that adjusts the standard inventory number. Then, the output unit 500 outputs the reference inventory number before adjustment and the reference inventory number after adjustment to the display device 600 (step S1700). By the above process, the standard stock quantity calculation device 10 ends the standard stock quantity calculation process.

(Modification)
In the above description, the case where the user has selected the target stock quantity mode has been described. Here, the case where the user selects the target stock value mode will be described. The system configuration is the same as that described in the first embodiment.

  In the target stock amount mode, the target stock amount is acquired instead of the target stock number in step S1100 of the standard stock number calculation process shown in FIG. The process of generating the order probability density function described in step S1200 and the process of calculating the standard stock quantity described in step S1300 are the same as those described in the first embodiment. However, the process of creating the standard inventory adjustment plan is different from the method of preparing the adjustment plan described with reference to FIG. 15 in the first embodiment.

  The process of creating the standard inventory adjustment plan in the target inventory value mode will be specifically described with reference to FIG. Here, it is assumed that the target inventory amount is ¥ 1,100 million. The standard inventory adjustment unit 403 acquires the product price from the product master of the input data storage unit 301. Then, the product price is multiplied by the standard stock quantity calculated in step S1300 of the standard stock quantity calculation process shown in FIG. 5 to calculate the stock price. It is assumed that the total of the calculated inventory amounts is ¥ 1,200 million. In this case, the inventory amount to be adjusted is 100 million yen. The standard stock adjustment unit 403 calculates the adjustment amount of each product by multiplying the stock amount by the distribution ratio.

  As in the example shown in FIG. 18, it is assumed that the sum of the standard stock numbers is 50,000 and the standard stock number of the product 1 is 1,000. In this case, the standard inventory adjustment unit 403 multiplies the standard inventory number by the product price 10,000 yen of the product 1 to obtain the stock product price 10 million yen of the product 1. Since the distribution ratio of the product 1 is 0.20, the adjustment amount of the product 1 is −2.0 million yen (= −10 * 0.2). The standard inventory adjustment unit 403 similarly calculates the products 2 and later. By the above processing, the standard inventory adjustment unit 403 creates a base standard inventory adjustment proposal based on the inventory product price.

  Next, in the same manner as the processing described with reference to FIG. 16, the standard inventory adjustment unit 403 gives random numbers of 0.00 or more and 1.00 or less in increments of 0.01 to the gene value to adjust the adjustment value of each product. Calculate However, here too, the stock quantity is replaced with stock value and calculated. By the above processing, the reference inventory adjustment unit 403 creates P adjustment proposals created by giving the gene value a random number of 0.00 or more and 1.00 or less in steps of 0.01.

  Then, in the preparation process of the standard inventory adjustment plan, the “surplus inventory number expected value” in the description of the first embodiment using FIG. 17 is replaced with “total amount of inventory amount” to create a final adjustment plan. .

  As described above, the reference inventory quantity calculation system 1 according to the present embodiment performs approximation by the least square method and approximation by the maximum likelihood method, and receives an order indicating the likelihood of the number of orders received based on the order record data. Generate a plurality of candidates for the probability density function, select an order probability density function to be used for calculating the standard inventory number from the plurality of candidates, and calculate the standard inventory number using the selected order probability density function . Thereby, the standard stock quantity calculation system 1 can calculate an appropriate standard stock quantity.

  Further, the order probability density function generation unit 401 evaluates each of the plurality of candidates for the order probability density function representing the likelihood of the number of orders received by calculating AIC, and based on the evaluation, among the plurality of candidates, Select the order probability density function used to calculate the standard inventory quantity. Thereby, the standard stock quantity calculation system 1 can calculate the more appropriate standard stock quantity.

  In addition, the standard inventory quantity calculation system 1 according to the present embodiment calculates "the difference between the target inventory quantity and the reference inventory quantity" and "the distribution ratio of the difference to the total product quantity by the optimization method application". A reference inventory adjustment unit 403 is provided to create a reference inventory adjustment proposal based on the difference between the number and the reference inventory number and the distribution ratio of the difference to the total product number. Thereby, the standard stock quantity calculation system 1 can calculate the proper standard stock quantity for all products.

  Further, the standard inventory adjustment unit 403 creates a standard inventory adjustment proposal based on the inventory amount. As a result, the standard inventory quantity calculation system 1 can achieve a service rate equal to or higher than the target service rate for all products, and can reduce the amount of defective assets due to surplus inventory.

  In the above description, the adjustment proposal is selected based on the evaluation based on either the target stock quantity or the target stock value. However, it is possible to select the adjustment proposal in consideration of the evaluation of both the target stock quantity and the target stock amount. For example, the final adjustment proposal may be selected based on the average rank of the rank rank based on the target stock number described in the first embodiment and the rank rank based on the target stock amount described in the modification.

  In the above description, although the approximation equation to be finally used is selected from both the least squares approximation equation and the maximum likelihood approximation equation, other approximation equations may be used. .

  Further, in the above description performed with reference to FIG. 15, the reference inventory adjustment unit 403 performs a process of steps S 1523 to S 1525 to select a gene value of 0.01 or more and a random number of 0.00 or more and 1.00 or less. An example of creating (P-1) adjustment plans given and created is illustrated. However, the processing from step S1523 to step S1525 may be omitted.

  Each function of the standard stock quantity calculation system 1 of the present invention can also be implemented by a general-purpose PC (Personal Computer). Specifically, in the above embodiment, it has been described that the program of the standard stock quantity calculation process performed by the standard stock quantity calculation system is stored in advance in the ROM of the storage unit 300. However, the program is distributed by being stored in a computer-readable recording medium such as a flexible disk, a compact disc read only memory (CD-ROM), a digital versatile disc (DVD) and a magneto-optical disc (MO). May be read and installed in a computer to configure a computer capable of realizing the functions described above.

Example 1
In order to verify the effect of the standard inventory quantity calculation system 1 according to the present embodiment, the standard inventory quantity is calculated for the servomotors of 10 products by the method of the following two patterns, and the service rate achieved in each pattern is evaluated.
Pattern 1: An order probability density function is generated using the reference inventory number calculation system 1 according to the present embodiment, and the reference inventory number is calculated.
Pattern 2: An order probability density function is generated using approximation by normal distribution to calculate a reference stock quantity.

The items used for the evaluation are as follows.
Product number i (i = 1 to N): The number of the target product.
-Order probability density function calculation month ja (ja = 2014/6 to 2015/1): Target period of order result data as a basis for generating an order probability density function.
Reference stock evaluation month jb (jb = 2015/2 to 2015/9): Target period of order data serving as the basis for evaluating the service rate.
Number of orders received IN _{i, j, k} : Number of orders received on the jth k of the product i.
Base stock quantity IC _i : Base stock quantity of product i.
Service rate IS _i : Service rate of product i.

The method of evaluation is as follows. The standard inventory quantity calculation system 1 generates an order probability density function based on the order quantity IN _{i, ja, k} of the order probability density function calculation month ja. Further, in the pattern 1, the standard inventory quantity calculation system 1 was set to use linear quadratic and linear cubic as an approximation formula. Also, when calculating the number of standard inventory after selection orders probability density function, cumulative orders probability density has set a reference inventory quantity calculation system 1 to 95% become orders number reference inventory quantity IC _i. That is, the standard inventory quantity calculation system 1 calculates the standard inventory quantity of patterns 1 and 2 according to equation (27).
IC _i = (the number of orders received that results in a cumulative order probability density of 95%) formula (27)

The service rate was evaluated by the following method. The standard inventory quantity IC _i was compared with the order quantity IN _{i, jb, k} of the standard inventory evaluation month jb, and when IC _i > IN _{i, jb, k} , it was determined that the delivery date was observed. That is, the service rate of the product i was calculated by equation (28).
IS _i = (the number of days of IC _i > IN _{i, jb, k} ) / total number of days of jb formula (28)

  The result of the verification is shown in FIG. 24A, 24B, 24C, 24D, 24E, 24F, 24G, 24H, 24I, and 24D, 24C, 24D, 24E, 24F, 24G, 24H, and 24I are plots of the order acceptance probability density function for each product generated by the reference inventory quantity calculation system 1 in verification. , And shown in FIG. 24J. The horizontal axis in FIGS. 24A to 24J represents the number of orders received, and the vertical axis represents the order probability. The solid line 1003 a to the solid line 1003 j in FIGS. 24A to 24J represent the order probability density function generated in pattern 2. Further, the solid line 1004 a to the solid line 1004 j represent the order probability density function generated in pattern 1. In the solid line 1004a to the solid line 1004j of FIGS. 24A to 24J, the order received by the reference inventory quantity calculation system 1 by calculating and evaluating the AIC out of any one of the linear quadratic and linear cubic set as the approximation formula The probability density function is illustrated. For example, for product A, the order probability density function generated by the linear cubic approximation formula was selected, and for product B, the order probability density function generated by the linear quadratic approximation formula was selected. As a result, as shown in FIG. 23, in the pattern 1, although the standard inventory number smaller than the pattern 2 was calculated, the service rate higher than the target service rate could still be achieved. In addition, as shown in FIGS. 24A to 24J, in almost all products, the probability of receiving an order when the number of received orders is small is high, and the probability of receiving an order when the number of received orders is large tends to be low. It can be seen that there is a large divergence.

(Example 2)
In order to verify the effect of the standard inventory quantity calculation system 1 according to the present embodiment, the target inventory quantity was set for the servo motor 10 product, the standard inventory quantity was calculated, and the achieved service rate was evaluated.

The items used for the evaluation are as follows.
Target stock quantity: Target stock quantity which is the upper limit of the standard stock quantity.
Product number i (i = 1 to N): The number of the target product.
Gene I _i (i = 1 to N): Inventory adjustment ratio of target products. The maximum is 1, and ΣI _i = 1.
Base stock quantity IC _i : Base stock quantity of product i.
Difference D: Difference between the target stock quantity and the sum of the standard stock quantity IC _i .
Adjusted base stock number NC _i : Base stock number after adjusting for product i.

The method of evaluation is as follows. Similar to the pattern 1 of the first embodiment, the standard inventory quantity calculation system 1 is set to use the linear quadratic and the linear cubic as an approximation equation. Also, when calculating the number of standard inventory after selection orders probability density function, cumulative orders probability density has set a reference inventory quantity calculation system 1 to 95% become orders number reference inventory quantity IC _i. Furthermore, the standard stock quantity calculation system 1 adjusted the standard stock quantity of each product by giving gene _Ii to each product. That is, the standard stock quantity calculation system 1 calculates the standard stock quantity after adjustment using Formula (29).
NC _i = IC _i −I _i * D equation (29)
Then, with respect to the standard stock quantity calculated by the standard stock quantity calculation system 1, the achieved service rate was evaluated in the same manner as in Example 1.

  Specifically, 100 patterns of gene patterns used by the standard inventory quantity calculation system 1 were set. The target inventory quantity was set at 1,500. Of the 100 gene patterns, the top 10 patterns with high service rates are shown in FIG. Further, FIG. 26 shows the service rate by product and the adjusted standard inventory quantity of the gene pattern having the highest service rate. It can be seen from FIG. 26 that, even when the target number of stocks is set, the standard number of stocks exceeding the service rate of 95% can be calculated. From the above results, the effect of the standard inventory quantity calculation system 1 according to the present embodiment is proved.

  The present invention is capable of various embodiments and modifications without departing from the broad spirit and scope of the present invention. In addition, the embodiment described above is for explaining the present invention, and does not limit the scope of the present invention. That is, the scope of the present invention is indicated not by the embodiments but by the claims. And, various modifications applied within the scope of the claims and the meaning of the invention are considered to be within the scope of the present invention.

  This application is based on Japanese Patent Application No. 2017-009553 filed on Jan. 23, 2017. The specification, claims and whole drawings of Japanese Patent Application No. 2017-009553 are incorporated herein by reference.

1 reference stock number calculation system, 10 reference stock number calculation device, 100 input device, 200 acquisition unit, 300 storage unit, 301 input data storage unit, 302 processing data storage unit, 400 processing unit, 401 order probability density function generation unit, 402 reference stock number calculation unit, 403 reference stock adjustment unit, 500 output unit, 600 display device, 700 external storage device, 800 AIC correction amount calculation unit.

In order to achieve the above object, a reference inventory quantity calculation system according to the present invention includes an acquisition means, an order probability density function generation means, a reference inventory quantity calculation means, and an evaluation value correction amount calculation means . The acquiring means is a target service rate representing a target value of a service rate which is a rate capable of responding to an order without causing a product shortage, order reception result data representing an order reception result, and production plan data representing a production plan. And inventory data representing the number of inventory . The order acceptance probability density function generating means generates an order acceptance probability density function which is a probability density function representing the likelihood of the number of orders accepted, based on the order acceptance result data acquired by the acquisition means. The reference inventory number calculation means calculates, based on the order probability density function generated by the order probability density function generation means, a reference inventory number for realizing a service rate equal to or higher than the target service rate acquired by the acquisition means. The order probability density function generating means generates a plurality of candidates for the order probability density function, calculates an evaluation value for each of the plurality of generated candidates, and calculates a reference inventory number from the plurality of candidates based on the calculated evaluation value. Select the order probability density function to use for The evaluation value correction amount calculation means calculates, based on the production plan data, the stock data, and the received order data, a predicted value of the number of used inventory representing the number of used stock, and a standard of each candidate of the received order probability density function. The number of stocks is calculated, and a correction amount for the evaluation value of each candidate of the order probability density function is calculated based on the difference between the calculated predicted value of the number of used stocks and the calculated number of standard stocks. The order acceptance probability density function generating means corrects the evaluation value of the order acceptance probability density function with the correction amount calculated by the evaluation value correction amount calculation means, and uses the order acceptance probability to calculate the reference inventory number based on the corrected evaluation value. Choose a density function.

Claims (9)

Acquisition means for acquiring a target service rate representing a target value of a service rate which is a rate capable of responding to an order without causing a product shortage and order acceptance result data representing an order acceptance result;
Order probability density function generation means for generating an order probability density function that is a probability density function representing the likelihood of the number of orders received based on the order acceptance result data acquired by the acquisition means;
Reference inventory number calculation means for calculating a reference inventory number for realizing a service rate equal to or higher than the target service rate acquired by the acquisition means based on the order acceptance probability density function generated by the order acceptance probability density function generation means; Prepare,
Standard stock quantity calculation system. The order probability density function generating means generates a plurality of candidates for the order probability density function, calculates an evaluation value for each of the plurality of generated candidates, and calculates the criterion from the plurality of candidates based on the evaluation value. Select order probability density function to use for calculation of stock quantity,
The standard inventory quantity calculation system according to claim 1. The order acceptance probability density function generating means generates some candidates among the plurality of candidates for the order acceptance probability density function using an approximation expression by a least squares method.
The standard stock quantity calculation system according to claim 2. The acquisition means further acquires production plan data representing a plan of production and inventory data representing an inventory number,
The standard inventory quantity calculation system further calculates, based on the production plan data, the inventory data, and the received order data, a predicted value of the number of used inventory representing the used inventory quantity, and the order probability density The reference inventory number of each candidate of the function is calculated, and the correction amount for the evaluation value of each candidate of the order probability density function is calculated based on the difference between the calculated predicted value of the inventory usage number and the reference inventory number. An evaluation value correction amount calculation means is provided,
The order probability density function generating means corrects the evaluation value of the order probability density function with the correction amount calculated by the evaluation value correction amount calculating means, and calculates the reference inventory number based on the corrected evaluation value. Select order probability density function to use for
The standard stock quantity calculation system according to claim 2 or 3. The acquisition means further acquires a target stock quantity representing a target value of the upper limit of the stock quantity,
The reference inventory number calculation system further determines that the reference inventory number is the target inventory number when the reference inventory number calculated by the reference inventory number calculation means exceeds the target inventory number acquired by the acquisition means. Further comprising a standard inventory adjustment means for adjusting to a standard inventory quantity within a range not exceeding
The standard stock quantity calculation system according to any one of claims 1 to 4. The acquisition means further acquires a target stock amount representing a target value of the upper limit of the stock amount,
The reference inventory number calculation system further determines that the reference inventory amount corresponding to the reference inventory number calculated by the reference inventory number calculation means exceeds the target inventory amount obtained by the acquisition means. It further comprises a standard inventory adjustment means for adjusting the standard inventory quantity to the standard inventory quantity corresponding to the standard inventory value in the range not exceeding the target inventory value.
The standard stock quantity calculation system according to any one of claims 1 to 4. The standard inventory adjustment means creates a plurality of adjustment plans for adjusting a standard inventory quantity of products randomly extracted from a plurality of products, and calculates the standard inventory quantity from among the prepared plurality of adjustment plans. Choose the adjustment proposal to adjust the
The standard stock quantity calculation system according to claim 5 or 6. An order probability density function generation step of generating an order probability density function that is a probability density function representing the likelihood of the number of orders received based on order acceptance actual data;
A standard inventory quantity calculation step of calculating a standard inventory quantity that realizes a service rate equal to or higher than a target service rate based on the order acceptance probability density function generated in the order acceptance probability density function generation step;
Base stock quantity calculation method. On the computer
An order probability density function generation step of generating an order probability density function that is a probability density function representing the likelihood of the number of orders received based on order acceptance actual data;
A standard inventory quantity calculation step of calculating a standard inventory quantity that realizes a service rate equal to or higher than a target service rate based on the order acceptance probability density function generated in the order acceptance probability density function generation step;
A program to run a program.

Images