JPH02268396A - Commodity ordering quantity determining device - Google Patents

Abstract

PURPOSE:To easily determine the commodity ordering quantity of the next day and to reduce the burden of a staff in charge of ordering by determining the commodity ordering quantity of the next day by fuzzy inference according to the sales number of commodities on that day, sold-out time and the weather of the next day. CONSTITUTION:In a PLU file 4, store information such as the sales number, etc., for each commodity inputted from a POS terminal are stored. When one commodity is selected by an input device 1, a sales number X1 and sold-out time X2 relating to the commodity are read from the file 4 and inputted to a fuzzy controller 2. Further, rainfall probability X3 of the next date inputted from the input device 1 is inputted to the controller 2 and the fuzzy inference is executed according to these input values. Then, commodity ordering quantity (y) of the next day is outputted to a display device 3. Thus, the commodity ordering quantity of the next day can be easily determined and the burden of the staff in charge of ordering is reduced.

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Field of Industrial Application This invention relates to a product order amount determination device for determining the next day's product order amount in a store using fuzzy reasoning.

(b) Conventional technology In stores that handle fresh foods and fresh flowers, it is necessary for efficient management to match the daily sales volume with the daily order volume as much as possible. At large stores, the person in charge of ordering used information such as past sales volume and sell-out time to determine the next day's order amount based on experience and know-how.

(C) Problems to be solved by the invention However, if the next day's order quantity is determined based on the experience and know-how of the person in charge of ordering, a person with a wealth of experience will be needed as the person in charge of ordering. There was a problem that would require significant management responsibility.

The purpose of this invention is to easily determine the next day's product order amount by determining the next day's product order amount from today's product sales volume, sell-out time, and next day's weather using fuzzy reasoning. To provide a product order quantity determining device which can facilitate the selection of a person in charge of ordering and reduce the burden of responsibility of the person in charge of ordering.

(d) Means for Solving the Problems The product order amount determining device of the present invention includes an input device for inputting today's sales quantity of products, today's sold-out time of products, and the next day's weather, and a heka device. Fuzzy inference means performs fuzzy inference based on input from the computer and determines the order quantity of products for the next day.

(e) Effect In this invention, fuzzy inference is performed in the fuzzy inference means using today's sales quantity of the product, sold-out time, and next day's weather inputted from the input device as input values, and the order amount of the product for the next day is determined. be done.

The fuzzy inference means is composed of a fuzzy operation unit that performs fuzzy calculations and a defuzzify unit that performs definite value calculations, as is well known. The fuzzy calculation unit is equipped with a membership function generator that follows predetermined fuzzy rules, and calculates membership values for input variables, and sends inference values calculated based on the results to the defuzzification unit. and output it. This fuzzy rule is 1f(x, =A and xz=B...)t
It is expressed in the form of hen (y = Z), and (x+ = -A
and xz=B...) is the antecedent part, (y=Z)
is called the consequent.

FIG. 6 is a diagram for explaining a known method of outputting inference results according to the above-mentioned fuzzy rules.

Figures (A) and (B) show the membership functions corresponding to the two variables (XI, represents the membership function. Two membership functions for the antecedent part are shown here, but as the types of variables in the antecedent part increase, the number of membership functions will increase accordingly. In each figure, the horizontal axis represents the value of the variable, and the vertical axis represents the position (degree of affiliation) of the menax. Now, if the value of variable x1 in the first term of the antecedent is xl, then the degree of affiliation at that time. is 0.5 (same figure (A
)reference). Also, the value of variable x2 in the second item of the antecedent part is x
z', the degree of membership at that time is 0.3 (see (B) in the same figure).

In such a case, the fuzzy operation takes the smallest value among the degrees of membership. That is, in the above example, a degree of affiliation of 0.3 is selected. Next, the membership function corresponding to Z is cut off at the above membership degree of 0.3, and the lower trapezoidal part S
Find the center of gravity position y'. Then, this y' is output as the inference result.

The above inference is performed for one rule, but generally multiple rules are set. In this case, the inference results shown in FIG. 6(C) are output for each rule. Then, the trapezoid parts output for each rule are logically summed, and the center of gravity y of the logically summed part (the shaded area in FIG. 6(D)) is outputted as a determined logical value. In this way, the output value is determined so that the input value shown on the horizontal axis of the membership function in FIGS. 6(A) and 6(B) takes an intermediate value.

In the above logical method, the logical product operation rule (operation for selecting the smaller degree of membership) of the degree of belonging belonging to the antecedent part and the logical sum operation rule of the trapezoidal part for the consequent part are min-ma
They are called x-rules and are executed in the antecedent logical product circuit and the consequent logical sum circuit, respectively.

In this invention, the center of gravity y in FIG. 6(D) is output as the order amount of the next day's product.

(flEmbodiment FIG. 1 is a block diagram showing the configuration of a product order amount determining device which is an embodiment of the present invention.

The PLU file 4 stores in-store information such as the number of sold items for each product, which is input from a POS terminal (not shown). When one product is selected in the input device 1, the sales quantity and sell-out time for that product are read out from the PLU file 4, and the sales quantity x1 and sell-out time x
2 to the fuzzy controller 2. Also,
The next day's probability of precipitation x3 inputted from the input device 1 is inputted to the fuzzy controller 2. The fuzzy controller 2 performs fuzzy inference from these input values (x+ to x,) and outputs the next day's product order amount y to the display device 3. The display device 3 displays this content. FIG. 2 is a block diagram showing the configuration of a fuzzy controller forming a part of the product order amount determining device.

The fuzzy controller 2 includes a fuzzy calculation section 40 and a defuzzify section 41. Fuzzy calculation unit 40
outputs the inference result xi for each rule according to the fuzzy rules shown in FIG. Each fuzzy calculation unit is provided for each rule, and the inference results of a total of eight fuzzy calculation units 40 are output in parallel to the defuzzification unit 41. For example, the fuzzy unit calculation unit 40 located at the top in FIG.
x,=PS/PM and xz=Ps/r'M an
d x,=ZR/N5) then (y=PS).

FIG. 3(A) shows the configuration of the fuzzy calculation section. The fuzzy operation unit 40 has four general-purpose membership function generators 50 to 53, and the membership function generator 5
For each of 0 to 53, the sales quantity x1 and the corresponding label PS/PM, the sold-out time x2 and the corresponding label PS/PM, the probability of precipitation X and the corresponding label Z
Label PS corresponding to R/NS and product order 1) y
is done manually. Each membership function generator generates a membership function corresponding to its label. That is, in the membership generator 50, the PS and PM membership functions shown in FIG. 5(A) are generated, and in the membership function generator 51, the PS or PM membership functions shown in FIG. 5(B) are generated.
membership function occurs. Further, the membership generator 52 generates a ZR or NS membership function shown in FIG. 5(C). The membership generator 53 generates the PS membership function shown in FIG. 5(D). The membership functions shown in FIGS. 5(A) to 5(D) are predetermined based on the sales performance and experience of the store.

The outputs of the membership function generators 50 to 52, that is, the degree of membership of each term in the antecedent part of the fuzzy rule, are output to the antecedent part AND circuit 54, and
The smaller degree of affiliation is selected according to the 1ni rule.

The result is sent to the consequent AND circuit 55. The consequent AND circuit 50 applies the inference result from the antecedent AND circuit 54 to the membership function output from the membership function generator 53 to generate a head as shown in FIG. 6(C). Performs a cut (takes logical product) and outputs the trapezoidal part as the inference result.

FIG. 3(B) is a diagram showing the configuration of the defuzzifier 41. As shown in the figure, the defuzzify unit 41 includes a logical sum circuit (consequent logical sum circuit) 60 and a definite value calculation circuit 6
1. The OR circuit 60 is mini-max
This is the part that calculates the max rule of the rules, and the trapezoidal outputs (inference results) from each of the eight fuzzy calculation units 40 are logically summed to form a hunting area as shown in FIG. 6(D).

The fixed value calculation circuit 61 determines the center of gravity position from this area and outputs the fixed value of the product order type.

The specific operation of the product order amount determining device will be explained below.

Assuming that today's sales quantity xl, sold-out time xt, and next day's probability of precipitation The combination of and the fuzzy operation output are as follows.

■x, :BM=0.8 Xz :PS=0.8 Now, select PS from the fuzzy rule as the label corresponding to the consequent (see Figure 4), and set its membership function to 0.
．． Cut the head at 8.

■xl :PM=0.8 x, :PM=0. 2 X3:ZR=1 In this combination, the antecedent AND circuit has the minimum value 0
, 2, and the consequent logical product circuit truncates the membership function of the label PS corresponding to the consequent at 02.

■x, :PL=0.2 x, :PS=0.8 x3 :ZR=1 In this combination, the antecedent AND circuit is 0. 2, and the consequent logical product circuit truncates the membership function of the label PM corresponding to the consequent by 0.2.

■x, :PL=0.2 Xz: PM=0.2 x, :ZR=1 In this combination, the antecedent AND circuit is 0. 2, and the consequent logical product circuit truncates the membership function of the label PM corresponding to the consequent by 0.2.

As a result of the above, as shown in Figure 5 (D), the labels PS, PM
The membership function corresponding to is truncated, and the trapezoidal part shown by hunting in the figure is output as the inference result. The defuzzifier executes the max rule and logically ORs the plurality of trapezoidal parts. After this, the center of gravity of the trapezoidal part is determined and the center of gravity position y shown in FIG. 5(D) is outputted as the next day's product order quantity. Is displayed.

In this way, the fuzzy controller uses fuzzy reasoning to determine the next day's order amount according to membership functions and fuzzy rules determined based on past sales performance and experience, so even an inexperienced ordering person can The amount of products to be ordered for the next day can be easily determined according to the displayed contents.

fg) Effects of the invention According to this invention, it is possible to easily determine the order quantity of products for the next day without relying on the experience and know-how of the person in charge of ordering, so even a person with little experience can easily carry out ordering work. This makes it easier to select the person in charge of ordering, and
This has the advantage of reducing the responsibility of the person in charge of ordering.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the configuration of a product order amount determination device that is an embodiment of the present invention, FIG. 2 is a diagram showing the configuration of a fuzzy controller that constitutes a part of the product order amount determination device, and FIG. A) and CB) are diagrams showing the configurations of the fuzzy calculation section and defuzzify section, respectively, of the same fuzzy controller, FIG. 4 is a diagram showing the fuzzy rules in the same fuzzy controller, and FIGS. 5 (A) to (D) are It is a diagram showing membership functions in the same fuzzy controller. Moreover, FIGS. 6(A) to 6(D) are diagrams for explaining a known fuzzy inference method. 1-human power device, 2-fuzzy controller, 3-display device.

Claims (1)

[Claims] (1) Number of products sold today, sold-out time of today's products,
and an input device for inputting the next day's weather, and fuzzy inference means for performing fuzzy inference based on the input from the input device and determining the order amount for the next day's products.