Demand forecasting method, demand forecasting device, and demand forecasting program
The demand forecasting device accurately predicts product demand changes by generating historical values and using generational change parameters to separate predicted demand, addressing the limitations of existing technologies and enhancing forecasting accuracy.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- HITACHI LTD
- Filing Date
- 2022-10-17
- Publication Date
- 2026-06-24
AI Technical Summary
Existing demand forecasting technologies, such as Patent Document 1, fail to accurately predict product demand changes due to the lack of consumer product usage data and do not provide specific details for generational changes, leading to inaccurate demand forecasting.
A demand forecasting device that includes a generation-coupled performance generation unit, a generation-coupled demand forecasting unit, and a generation-specific forecast demand separation unit, which generates historical values by combining actual demand values for older and new products, uses generational change parameters to forecast demand, and separates predicted demand using logistic curves.
Enables highly accurate demand forecasting for discontinued and new products by considering generational changes based on customer preferences, even without consumer product usage data, thereby optimizing sales opportunities and reducing inventory risks.
Smart Images
Figure 0007879778000001 
Figure 0007879778000002 
Figure 0007879778000003
Abstract
Description
Technical Field
[0001] The present invention relates to a demand prediction technology for predicting the demand for products based on the obsolescence of products and the generation change of new products.
Background Art
[0002] With product comparison sites, SNS, etc., consumers can easily search for and compare the design, functions, prices, and reviews of products on their own, and the options for consumers to choose products are diversifying. For consumers, which of the obsolete products and new products to choose is one of the important selection criteria.
[0003] Under such circumstances, businesses need to carry out the generation change of obsolete products and new products for each product group according to customer preferences so as to balance the acquisition of sales opportunities and the reduction of inventory risks.
[0004] As a prior art document for data mining for analyzing product transitions, there is Patent Document 1. Patent Document 1 describes querying product usage status data of legacy products and new products from a database as time series data, that the product usage status data represents the situations of a large number of consumers of legacy products and new products, determining the relationship between two time series data using a mathematical model, and estimating, determining, or predicting functions related to product transitions such as the value of product transitions and the transition period of product usage.
Prior Art Documents
Patent Documents
[0005]
Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0006] Patent Document 1 does not consider cases where product usage data for legacy and new products is not accessed. Therefore, its effect is limited to estimating, determining, or predicting functions related to product transition when product usage data for legacy and new products is accessed. Furthermore, since Patent Document 1 does not provide specific details regarding the predicted values, it is difficult to achieve highly accurate demand forecasting that takes generational changes into account.
[0007] Therefore, the present invention aims to provide a demand forecasting technology that can accurately predict the demand for discontinued and new products for each product group based on customer preferences, even when there is no data on consumer product usage. [Means for solving the problem]
[0008] To give one example, the present invention is: A demand forecasting device having a generation-coupled performance generation unit, a generation-coupled demand forecasting unit, and a generation-specific forecast demand separation unit performs the following: A demand forecasting method for forecasting demand for discontinued and new products, The aforementioned generation coupling performance generation unit, By combining the actual demand values for older products and the actual demand values for new products in a time series, we generate historical values for generationally combined demand. The aforementioned generation-coupled demand forecasting unit Actual values of generational coupling demand and the release of older models hand Using generational change parameters that include the release interval from one product to the release of a new product, and the transition period required for the generational change between older and newer products. By multiplying the predicted overall demand for the product during the aforementioned launch interval, which is a one-year cycle, by the annual market share, We forecast the demand for generational coupling, The aforementioned generation-specific forecast demand separation unit, Based on the forecast results for generational coupling demand, using generational change parameters By obtaining a logistic curve that has symmetry and approximates the generation ratio of the new product, and which passes close to the actual ratio of the actual value of the generation-bound demand and a hypothetical ratio that is symmetrical with the actual ratio, through regression calculation, the predicted ratios of the outdated product and the new product are obtained. The projected demand values for older and new products are separated. [Effects of the Invention]
[0009] According to the present invention, it is possible to perform highly accurate demand forecasting that takes into account the generational change between discontinued and new products in product groups based on customer preferences. [Brief explanation of the drawing]
[0010] [Figure 1] This is a hardware configuration diagram of the demand forecasting device in Example 1. [Figure 2]This is a block diagram of the functional configuration of the demand forecasting device in Example 1. [Figure 3] This is a flowchart of the demand forecasting process in Example 1. [Figure 4] This figure shows the relationship between generation-coupled demand and generation-change parameters handled by the calculation unit of the demand forecasting device in Example 1. [Figure 5] This figure shows the relationship between the generational change parameters handled by the calculation unit of the demand forecasting device in Example 1 and the logistic curve. [Figure 6] This is a flowchart of the demand forecasting process in Example 2. [Figure 7] This is an example of revenue and demand forecast results obtained from the generation change parameters displayed by the display processing unit of the demand forecasting device in Example 2. [Modes for carrying out the invention]
[0011] Hereinafter, embodiments of the present invention will be described with reference to the drawings. [Examples]
[0012] Figure 1 is a hardware configuration diagram of the demand forecasting device in this embodiment. As shown in Figure 1, the demand forecasting device 100 has a CPU (computer) 101, which is a general information processing device, memory 102, storage device (storage unit) 105, communication unit 106, input device 103, and output device 104. The storage device 105 stores a processing program 107 and necessary related data 108. The storage device 105 has storage areas for the related data 108, including (1) actual product demand information, (2) lineage classification / generation notation rule information, (3) generation change parameter information, (5) actual product demand information by generation, (6) actual demand information by generation, (7) predicted demand information by generation, and (8) predicted demand information by generation. Details of this information will be described later.
[0013] The demand prediction device 100 realizes each function through software processing in which the CPU 101 loads the processing program 107 for realizing each function from the storage device 105 to the memory 102, interprets it, and executes it. Note that the input device 103 and the output device 104 may be externally connected. Further, the demand prediction device 100 may be configured as a demand prediction system that is composed of a server connected via a network and input / output devices and processes using a processing program installed in the server.
[0014] FIG. 2 is a functional configuration block diagram of the demand prediction device in the present embodiment. The demand prediction device 100 includes an operation processing unit 201 that receives operations by a user using an input device 103 such as a keyboard, a mouse, a touch panel, etc., a data input / output processing unit 202 that takes in input data regarding demand prediction obtained from a peripheral system or spreadsheet software in a TSV format or the like, and outputs intermediate data regarding demand prediction and the result of demand prediction as output data, an arithmetic unit 203 that performs processes such as combination, prediction, and separation of data regarding demand prediction, a storage processing unit 204 that stores the input data taken in, intermediate data for demand prediction, and the result of demand prediction in the storage device 105, and a display processing unit 205 that displays the result of demand prediction on an output device 104 such as a display or a projector and shows it to the user.
[0015] The arithmetic unit 203 includes a pedigree / generation information discrimination unit 211, a generation combination achievement generation unit 212, a generation combination demand prediction unit 213, and a generation-by-generation predicted demand separation unit 214 as processes performed by the CPU 101 according to the processing program 107. Here, the pedigree indicates the relationship between the phased-out products and the new products for each product group according to customer preferences. Details of these processes will be described later.
[0016] FIG. 3 is a processing flowchart of demand prediction in the present embodiment. In FIG. 3, first, in step S101, the user activates the demand prediction device 100 by the operation processing unit 201. Note that it may be activated by batch activation based on date and time specification or activation triggered by an event.
[0017] Next, in step S102, the demand prediction device 100 captures, by the data input / output processing unit 202, the input information of (1) product actual demand information, (2) pedigree classification / generation representation rule information, and (3) generation change parameter information, and stores it in the storage device 105 by the storage processing unit 204.
[0018] Here, as data items (values), (1) product actual demand information has, for example, product model (e.g., PRD1-SPC14C, PRD1-SPC12D, PRD1-SPC14E), release week (e.g., week of 2021 / 07 / 12), demand week (e.g., week of 2022 / 07 / 04), actual demand quantity (e.g., 2,103 units), and so on.
[0019] (2) Pedigree classification / generation representation rule information is product model name rule information. For example, the pedigree classification 1 rule is a continuous character string from the first digit of the upper part of the product model (e.g., PRD1-SPC), the pedigree classification 2 rule is the numerical part excluding the last digit of the product model (e.g., 12 or more and 14 or less), and the generation representation rule is that the last digit of the product model is in ascending alphabetical order, and so on. That is, the pedigree classification 1 rule is a rule for specifying product groups for each customer preference such as the concept, function, and performance of the product (e.g., PRD1-SPC), the pedigree classification 2 rule is a rule for specifying product groups for each customer preference such as the size and weight of the product (e.g., 12 - 14), and the generation representation rule is a rule for expressing the relationship of the generation order such as phased-out products (e.g., D) and new products (e.g., E). By expressing the relationship between phased-out products and new products in the product group for each customer preference in the product model name rule in this way, it is possible to automatically aggregate the demand quantity, etc. for each product group for each customer preference by the product model, and the management man-hours can be reduced. Conversely, for flexible operation, the relationship between phased-out products and new products for each product group for each customer preference may be directly described in the master and aggregated. Also, a combination of rule-based and directly described ones may be used. Thereby, both reduction of management man-hours and flexibility can be achieved.
[0020] (3) Generational change parameter information includes, for example, the generational coupling type (e.g., PRD1-SPC:12-14:C->D->E), proposed parameters (e.g., P01), release interval (e.g., 52 weeks), and changeover period (e.g., 26 weeks).
[0021] Next, in step S103, the lineage / generation information discrimination unit 211 discriminates the lineage / generation information from (1) product actual demand information and (2) lineage classification / generation notation rule information. By adding this discriminated information to the data items of the (1) product actual demand information, (5) generation-specific product actual demand information is generated and stored in the storage device 105 by the storage processing unit 204. Here, the data items (values) of (5) generation-specific product actual demand information include, for example, product model (e.g., PRD1-SPC12D), lineage classification 1 (e.g., PRD1-SPC), lineage classification 2 (e.g., 12-14), generation (e.g., D), generation type (e.g., discontinued product (previous generation)), release week (e.g., week of 2021 / 07 / 12), demand week (e.g., week of 2022 / 07 / 04), and actual demand quantity (e.g., 2,103 units).
[0022] Then, in step S104, the generation-combined performance generation unit 212 generates (6) generation-combined performance demand information by aggregating the actual demand figures, for example, by adding up multiple generations of new and old products from (5) generation-specific product performance demand information for each lineage, and the storage processing unit 204 stores it in the storage device 105. Here, the data items (values) of (6) generation-combined performance demand information include, for example, the generation-combined type (e.g., PRD1-SPC:12-14:C->D->E), the demand week (e.g., the week of 2022 / 07 / 18), and the actual demand figures (e.g., 2,326 units).
[0023] Figure 4 shows the relationship between generation-coupled demand and generational change parameters handled by the calculation unit of the demand forecasting device in this embodiment. Generation-coupled demand 310, as the relationship between outdated products (1 generation prior) 311 and new products 312 in product groups for each customer preference, is as shown in the upper part of Figure 4. As shown in the figure, generation-coupled demand 310 has overlapping periods between generations, but compared to the actual demand for a single generation of products, it has a shape with multiple repeating peaks, resulting in time-periodic and stationary time series data. In addition to this, for example, time series data with even greater stationarity can be obtained by calculating the annual share of the number of units demanded by each generation in relation to the market as a whole product and combining them. When dealing with stationary time series data, parameters are selected using evaluation values such as AIC (Akaike Information Criterion) in a composite model of the time series forecasting method. In this way, while demand forecasting is difficult with a single peak, obtaining stationary time series data enables time series forecasting with high accuracy.
[0024] Returning to Figure 3, in step S105, the generation-coupled demand forecasting unit 213 uses the "launch interval" of the generation change parameter information (3) as the period and forecasts (7) generation-coupled forecast demand information from the generation-coupled actual demand information using a composite model of time series forecasting methods, and stores it in the storage device 105 by the storage processing unit 204. Here, the data items (values) of the generation-coupled forecast demand information (7) include, for example, the generation-coupled model (e.g., PRD1-SPC:12-14:C->D->E), parameter proposal (e.g., P01), demand week (e.g., week of 2022 / 08 / 08), and forecast demand quantity (2,681 units).
[0025] In the upper part of Figure 4, the proportion of each generation, such as the older product (previous generation) 311 and the new product 312, within the combined generation demand 310 is shown in the lower part of Figure 4, in the generation proportion 320. The period from the launch week of the older product (previous generation) 311 to the launch week of the new product 312 (e.g., 52 weeks) is called the launch interval 322, and the period from the launch of the new product 312 until the generational change with the older product (previous generation) 311 is completed (e.g., 26 weeks) is called the changeover period 323. The launch interval 322 and the changeover period 323 together are called the generational change parameter 321. The "launch interval" in the generational change parameter information (3) shown as the launch interval 322 within the generational change parameter 321 is used as the period for forecasting combined generation demand. In addition to the above, for example, one could determine the predicted market demand for the entire product, etc., by setting the cycle to one year (52 weeks), and then multiply that by the annual share of the previous year, etc., to predict the combined demand for each generation. By making predictions that take cycles into account in this way, it is possible to make highly accurate time series forecasts.
[0026] Next, in step S106 of Figure 3, the generation-specific forecast demand separation unit 214 separates the (7) generation-combined forecast demand information into (8) generation-specific product forecast demand information using the "transition period" of the (3) generation-transition parameter information, and stores it in the storage device 105 by the storage processing unit 204. Here, the data items (values) of the (8) generation-specific product forecast demand information include, for example, the generation-combined type (e.g., PRD1-SPC:12-14:C->D->E), the parameter proposal (e.g., P01), the product model (e.g., PRD1-SPC14E), the generation type (e.g., new product), the demand week (e.g., week of 2022 / 08 / 08), and the forecast demand quantity (e.g., 929 units).
[0027] For the generational change between the outdated product (1st generation) 311 and the new product 312, which is performed using the "generational change" parameter information (3) shown as the changeover period 323 (e.g., 26 weeks) in Figure 4, the generational ratio 320 of the new product 312 is approximated by a logistic curve y=f(x) with symmetry as shown in Figure 5. That is, in Figure 5, let's assume that the actual ratio 411, shown by the black circles, has three points: y=f(1) at x=1, y=f(2) at x=2, and y=f(3) at x=3. In contrast, taking into account the symmetry of the logistic curve, the provisional ratio 412, shown by the white circles, has three points: y=1-f(3) at x=24, y=1-f(2) at x=25, and y=1-f(1) at x=26. Here, the provisional ratio 412 can be expressed as y=1-f(c-x+1) at time x, where c is the value during the transition period 323 (e.g., 26 weeks). The predicted ratio 421 is obtained by using regression calculation to find a logistic curve y=f(x) that passes near the three points of the actual ratio 411 and the three points of the provisional ratio 412. After the transition period 323 (e.g., 26 weeks), from time x=27 onwards, the predicted ratio 421 is set to y=1. In the example in Figure 5, the predicted ratio 421 is used during the forecast period 420 (x≧4). The predicted number of units in demand for the new product 312 is calculated by multiplying the predicted ratio 421 by the generational combined demand 310. Furthermore, the predicted number of units in demand for the older product (1 generation prior) 311 is calculated by subtracting the predicted number of units in demand for the new product 312 from the generational combined demand 310.
[0028] Thus, according to this embodiment, even when consumer product usage data is unavailable, it is possible to make highly accurate demand forecasts for older and newer products, taking into account generational changes, by using generational coupling demand and generational change parameters for each customer preference. [Examples]
[0029] This embodiment describes an example in which, in addition to demand forecasting as in Embodiment 1, the generational change parameters are optimized.
[0030] Figure 6 is a flowchart of the demand forecasting process in this embodiment. In Figure 6, the same processes as in Figure 3 are denoted by the same reference numerals, and their explanations are omitted.
[0031] In this embodiment, the calculation unit 203 in Figure 2 performs processing as a generation change parameter evaluation unit 215 (not shown) in addition to that of Embodiment 1. Furthermore, the storage device 105 stored by the storage processing unit 204 has storage areas for (4) intergenerational trade-off information and (9) generation change parameter evaluation information (not shown), in addition to that of Embodiment 1.
[0032] In Figure 6, in step S502, between steps S102 and S103 of Example 1, the demand forecasting device 100 receives input information of (4) intergenerational trade-off information via the data input / output processing unit 202 and stores it in the storage device 105 via the storage processing unit 204. Here, the data items (values) of (4) intergenerational trade-off information include, for example, the generational coupling type (e.g., PRD1-SPC:12-14:C->D->E), the intergenerational demand ratio (e.g., 1:0.8), and the intergenerational profit ratio (e.g., 1:1.2). The intergenerational demand ratio and intergenerational profit ratio represent the estimated relative ratio of demand and profitability between the older product and the new product. In addition to these, for example, the estimated prices of the older product and the new product, and the estimated compensation costs may also be included.
[0033] In step S503, for each of the proposed parameters of the generation change parameter information (3), the three steps of Step S105, Step S106 of Example 1, and Step S504, which will be described below, are repeated.
[0034] In step S504, the profitability of the generational change parameters is evaluated as (9) generational change parameter evaluation information from (4) intergenerational trade-off information evaluation logic and (8) generational product forecast demand information, and stored in the storage device 105 by the memory processing unit 204. Here, the data items (values) of (9) generational change parameter evaluation information include, for example, the generational coupling type (e.g., PRD1-SPC:12-14:C->D->E), recommended parameter proposal (e.g., P01), product type of the older product (1 generation ago) (e.g., PRD1-SPC12D), predicted number of units of the older product (1 generation ago) (e.g., 51,524 units), evaluation value of the older product (1 generation ago) (e.g., 52pt), product type of the new product (e.g., PRD1-SPC14E), predicted number of units of the new product (e.g., 25,034 units), evaluation value of the new product (e.g., 30pt), etc. As an evaluation logic, for example, the predicted number of units for older and new products is calculated by multiplying the predicted demand for each product by the (8) predicted demand for each generation of product information, using the intergenerational demand ratio (e.g., 1:0.8) from the (4) intergenerational trade-off information. The evaluation value is then calculated by multiplying each predicted number by the intergenerational revenue ratio (e.g., 1:1.2) from the (4) intergenerational trade-off information. In addition to this, evaluation may also be performed using, for example, the revenue of older products (e.g., a value calculated from the formula "(price - compensation cost) × predicted demand"), the revenue of new products (e.g., a value calculated from the formula "price × predicted demand").
[0035] In step S505, the (9) generational change parameter evaluation information and the (8) generational product forecast demand information for that parameter, which maximizes revenue among the parameter proposals for (3) generational change parameter information, are output and stored in the storage device 105 by the storage processing unit 204. These outputs are passed to peripheral systems and spreadsheet software by the data input / output processing unit 202. The output data stored in the storage device 105 can be accessed by the user by specifying the product model using the input device 103. For example, as shown in Figure 7, the parameter proposal (proposal A) 600 that maximizes revenue, and the forecast results for the generational change parameters 321, revenue 601, generational combined demand 310, outdated product 311, and new product 312 at that time are displayed on the output device. Here, time-series graphs of actual and projected demand for each of the following categories—combined generation demand 310 (e.g., PRD1-SPC:12-14:C->D->E), older models (previous generation) 311 (e.g., PRD1-SPC12D), and new products 312 (e.g., PRD1-SPC14E)—are displayed side-by-side with the time axis aligned. This allows users to understand the past trends, future outlook, generational shift patterns, and profitability of each of these categories.
[0036] Thus, according to this embodiment, in addition to demand forecasting as in Example 1, it is possible to evaluate the generational change between outdated and new products and optimize the generational change parameters.
[0037] As illustrated above, the present invention enables demand forecasting that takes into account the generational change between discontinued and new products in product groups based on customer preferences, thereby enabling the acquisition of sales opportunities and the reduction of inventory risk. Therefore, appropriate resource utilization is possible, and the present invention contributes to waste reduction, particularly in relation to SDG 12, Responsible Consumption and Production, in order to achieve the Sustainable Development Goals (SDGs).
[0038] Furthermore, the present invention is not limited to the embodiments described above, and various modifications are included. For example, the embodiments described above are explained in detail to make the present invention easier to understand, and are not necessarily limited to those having all the configurations described. Also, some of the configurations of one embodiment may be used in other embodiments. It is possible to substitute the configuration of one embodiment with that of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace parts of the configuration of each embodiment with those of other embodiments. [Explanation of symbols]
[0039] 100: Demand forecasting device, 101: CPU (computer), 103: Input device, 104: Output device, 105: Memory device, 107: Processing program, 108: Related data, 201: Operation processing unit, 202: Data input / output processing unit, 203: Calculation unit, 204: Memory processing unit, 205: Display processing unit, 211: Genealogy / generation information discrimination unit, 212: Generation-linked actuals generation unit, 213: Generation-linked demand forecasting unit, 214: Generation-specific forecasted demand separation unit, 215: Generation change parameter evaluation unit, 310: Generation-linked demand, 311: Outdated products, 312: New products, 320: Generation ratio, 321: Generation change parameters, 322: Release interval, 323: Change period, 411: Actual ratio, 412: Tentative ratio, 420: Forecast period, 421: Forecast ratio, 600: Proposed parameters, 601: Revenue
Claims
1. A demand forecasting method for forecasting demand for outdated products and new products, which is performed by a demand forecasting device having a generation-coupled performance generation unit, a generation-coupled demand forecasting unit, and a generation-specific forecast demand separation unit, The generation-combined demand generation The generational coupling demand forecasting unit uses the actual generational coupling demand and generational change parameters, which include the release interval from the release of the older product to the release of the new product and the changeover period required for the generational change between the older product and the new product, to forecast the generational coupling demand by multiplying the forecast demand for the entire product in the release interval, which is a one-year cycle, by the annual share. A demand forecasting method characterized in that the generation-specific forecast demand separation unit, using the generation change parameter, uses the forecast results of the combined generation demand to obtain, through regression calculation, a logistic curve that is symmetrical and approximates the generation ratio of the new product, and that passes close to the actual ratio of the actual value of the combined generation demand and a provisional ratio that is symmetrical with the actual ratio, thereby obtaining the forecast ratios of the older product and the new product, and separating them into the forecast demand values of the older product and the new product.
2. In the demand forecasting method described in claim 1, A demand forecasting method characterized in that the generational change parameter evaluation unit of the demand forecasting device evaluates the revenue of the outdated product and the new product based on the demand values of the outdated product and the new product in multiple proposed generational change parameters and the relationship between generational trade-offs related to revenue, and extracts the generational change parameter that maximizes the overall revenue.
3. In the demand forecasting method according to claim 1 or 2, The lineage / generation information discrimination unit of the demand forecasting device generates generational product performance demand information having product naming rules that identify product groups according to customer preferences, including product specifications. A demand forecasting method characterized in that the generation-combined demand generation generation unit generates the generation-combined demand value by combining the actual demand value of the outdated product and the actual demand value of the new product using the generation-specific product demand information.
4. In the demand forecasting method described in claim 2, A demand forecasting method characterized in that the display processing unit of the demand forecasting device displays the generation change parameter, the expected revenue, and, as a forecast result, graphs of the actual and forecast values of the generation-combined demand, the demand for the outdated product, and the demand for the new product.
5. A demand forecasting device that forecasts demand for discontinued and new products, A generation-combined demand generation unit generates generation-combined demand data by combining the actual demand data for the discontinued product and the actual demand data for the new product in a time series. A generational coupling demand forecasting unit predicts the generational coupling demand by multiplying the predicted demand for the entire product in the one-year cycle of the release interval by the annual share, using the actual value of the generational coupling demand, the release interval from the release of the older product to the release of the new product, and generational change parameters including the replacement period required for the generational change between the older product and the new product. A demand forecasting device characterized by having a generational forecast demand separation unit that, based on the forecast results of the generational combined demand, uses the generational change parameter to obtain a logistic curve that is symmetrical and approximates the generational ratio of the new product, and that passes close to the actual ratio of the actual value of the generational combined demand and a provisional ratio that is symmetrical with the actual ratio, thereby obtaining the forecast ratios of the older product and the new product, and separating them into forecast demand values for the older product and the new product.
6. In the demand forecasting device according to claim 5, A demand forecasting device characterized by having a generational change parameter evaluation unit that evaluates the revenue of the outdated product and the new product based on the predicted demand values of the outdated product and the new product in multiple proposed generational change parameters and the relationship of intergenerational trade-offs related to revenue, and extracts the generational change parameter that maximizes the overall revenue.
7. In the demand forecasting device according to claim 5 or 6, It has a lineage / generation information discrimination unit that generates generational product performance demand information having product naming rules that identify product groups according to customer preferences, including product specifications, The generation-combined performance generation unit is a demand forecasting device characterized in that it combines the actual demand values for the outdated product and the actual demand values for the new product using the generation-specific product performance demand information to generate the actual demand values for the generation-combined demand.
8. In the demand forecasting device according to claim 6, A demand forecasting device characterized by having a display processing unit that displays a graph of the generational change parameters, the expected revenue, and the actual and predicted values of the generational change demand, the demand for the outdated product, and the demand for the new product, as forecast results.
9. A demand forecasting program that uses a computer to predict demand for discontinued and new products, The steps include generating a time-series combined historical demand value for the older product and the new product, and The steps include: predicting the generational demand by multiplying the predicted demand for the entire product in the one-year cycle by the annual share, using the actual value of the generational demand, the release interval from the release of the older product to the release of the new product, and generational change parameters including the replacement period required for the generational change between the older product and the new product; A demand forecasting program characterized by causing the computer to perform the following steps: using the generational change parameters from the forecast results of the generational combined demand, to obtain by regression calculation a logistic curve that has symmetry and approximates the generational ratio of the new product, which passes close to the actual ratio of the actual value of the generational combined demand and a provisional ratio that is symmetrical with the actual ratio, thereby obtaining the forecast ratios of the outdated product and the new product, and separating them into forecast demand values for the outdated product and the new product.
10. In the demand forecasting program described in claim 9, A demand forecasting program characterized by causing the computer to perform the step of evaluating the revenue of the outdated product and the new product based on the predicted demand values of the outdated product and the new product in multiple proposed generational change parameters and the relationship between generational trade-offs related to revenue, and extracting the generational change parameters that maximize overall revenue.
11. In the demand forecasting program according to claim 9 or 10, A step of generating generational product performance demand information that has product naming rules to identify product groups for each customer preference, including product specifications, A demand forecasting program characterized by causing the computer to perform the step of generating a combined generational demand value by combining the actual demand value of the discontinued product and the actual demand value of the new product using the generational product demand information.
12. In the demand forecasting program described in claim 10, A demand forecasting program characterized by causing the computer to perform the step of displaying a graph of the generational change parameters, the expected revenue, and the actual and predicted values of the generational change demand, the demand for the older product, and the demand for the new product, as forecast results.