An intelligent new product pricing method based on historical production data and cost model

Abstract

The application discloses a new product intelligent pricing method based on historical production data and a cost model, a structured job cost model library is established for historical products, and cost estimation of a new product no longer depends on manual item-by-item accounting, but automatically searches similar historical products through feature matching, combines parameter correction to quickly generate a cost base table, and greatly shortens a pricing cycle; a three-layer model of a product, a job and a cost element is adopted, and cost calculation granularity is sunk to a job activity level, so that cost estimation is no longer simple overall proportional scaling, but independently calculates for each job activity, and can accurately reflect differences between the new product and the historical product in specific process links; meanwhile, through multi-dimensional similarity matching and a multiple correction mechanism, it is ensured that the estimation result approximates to the actual cost.

Description

Technical Field

[0001] This invention belongs to the field of manufacturing cost management technology, and relates to a method for intelligent pricing of new products based on historical production data and cost models. Background Technology

[0002] In the scenario of new product pricing in the manufacturing industry, the core challenges faced by enterprises can be summarized into three levels: First, the quotation process is slow. Traditional product quotations rely heavily on the manual experience of senior engineers, requiring item-by-item calculation of material costs, processing time, and process fees. The quotation cycle for a complex product often takes several days or even weeks, making it difficult to adapt to the market demand for rapid response.

[0003] Secondly, the accuracy of price quotes is poor. Manual pricing relies on individual experience, and different quoters may have significant discrepancies in their estimates of the same product. This is especially true when dealing with new products that differ from existing products. The lack of systematic historical data makes price discrepancies even more pronounced, easily leading to lost orders due to overpricing or losses due to underpricing.

[0004] Third, the cost structure is opaque. Traditional quotations typically only provide a total price or highly consolidated category prices, making it difficult for clients to clearly understand the details and proportions of each cost component. Simultaneously, companies often struggle to accurately identify the source and justification of various costs during the quotation stage, hindering subsequent cost control and optimization.

[0005] Several cost estimation and pricing methods already exist in the prior art. For example, CN202210542760.X discloses a multi-dimensional target cost estimation method and system based on new products in the manufacturing industry. It uses a modular structure to estimate the costs of parts, components, and outsourcing lists separately. However, its cost model is a predefined general template and lacks systematic utilization of historical production data. CN202210963595.5 discloses a dynamic cost control system for the manufacturing process of large equipment products. It uses the proportional correction method and the key parameter correction method to generate pricing schemes for new products using historical product costs. However, its correction dimension is relatively simple and fails to decompose the product into fine-grained cost models according to work activities. CN202110824580.6 discloses a method for order splitting and pricing of customized log products. It achieves automatic pricing through product componentization and parameterization. However, this method relies on a predefined component formula library and is not adaptable to new product categories that lack historical component data.

[0006] Therefore, how to build a cost model that can adaptively match the characteristics of new products based on historical production data and using operational activities as cost collection units, and how to quickly and accurately output clearly structured quotation results according to the quotation template format required by customers, is an urgent problem to be solved. Summary of the Invention

[0007] In order to overcome the shortcomings of the prior art, this invention provides a method for intelligent pricing of new products based on historical production data and cost models.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: A smart pricing method for new products based on historical production data and cost models includes the following steps: S1. Construct a historical cost model library based on historical product data that the enterprise has completed production and delivery; S2. Configure template cost items according to the customer's quotation format, and establish mapping and aggregation rules between template cost items and cost elements; S3. Extract the feature vector of the new product and calculate the similarity with the features of each historical product in the historical cost model library to obtain a reference product set. S4. Summarize the estimated costs of the activities of the reference product set to form a new product activity cost table; S5. Fill the quotation template with the activity cost table, summarize and add the additional rates at each level, and then output a structured quotation.

[0009] Furthermore, in step S1, the method for constructing the historical cost model library includes: decomposing historical products into several parts / work units, determining the various work activities that the parts / work units undergo from raw material input to warehousing, establishing a work cost card for each work activity, the work cost card including cost elements, abstracting the cost elements into parameterized formulas, and storing the work cost cards and parameterized formulas according to a three-dimensional index of product type, structural level, and work activity to form a historical cost model library.

[0010] Furthermore, in step S2, the template cost items include the top-level total price, the intermediate-level cost categories, and the bottom-level detailed items.

[0011] Furthermore, in step S3, feature vectors are extracted based on the technical parameters and process requirements of the new product.

[0012] Furthermore, in step S3, the similarity calculation is based on structural similarity, process similarity, and parameter similarity. When the similarity of a certain dimension does not meet the threshold but the overall similarity does, the cost element corresponding to that dimension is marked as an element to be corrected.

[0013] Furthermore, in step S4, the activity cost table is structured with products, components, parts, and activities as the row dimension and each cost element as the column dimension, forming a cost matrix.

[0014] Furthermore, in step S5, based on the mapping and aggregation rules in step S2, corresponding data is extracted or aggregated from the activity cost table and populated into each cost item of the quotation template.

[0015] In summary, the advantages of this invention are: This invention establishes a structured activity cost model library for historical products, so the cost estimation of new products no longer relies on manual item-by-item calculation. Instead, it automatically retrieves similar historical products through feature matching and quickly generates a cost table by combining parameter correction, thus significantly shortening the quotation cycle.

[0016] By adopting a three-tiered model of product, operation, and cost elements, the cost calculation granularity is reduced to the operational level. This makes cost estimation no longer a simple overall scaling, but an independent calculation for each operational activity, which can accurately reflect the differences between new products and historical products in specific process links. At the same time, through multi-dimensional similarity matching and multiple correction mechanisms, the estimation results are ensured to be close to the actual cost.

[0017] With activity-based costing as its core, each quoted amount can be traced back to specific activities and cost elements. Whether for internal audits or presentations to clients, it can display a complete cost traceability chain from the total price to the detailed items and from the detailed items to the activities, making the cost composition of the quote clear at a glance.

[0018] Through a feedback learning mechanism, actual production data of new products are continuously fed back into the historical cost model library, so that the cost model is continuously calibrated and optimized as product data accumulates, and the accuracy of quotations continues to improve over time.

[0019] The customer quotation template is decoupled from the underlying cost model. The same cost table can output quotations with different structures and granularities according to the quotation format requirements of different customers, which improves the system's versatility and applicability. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the intelligent pricing method for the new product of the present invention. Detailed Implementation

[0021] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0022] This invention provides a method for intelligent pricing of new products based on historical production data and cost models, comprising the following steps: Step S1: Construct a model based on the three elements of product, operation, and cost. The model is based on historical product data of the enterprise's completed production and delivery. Perform the following processing on each historical product: S1.1 Based on the bill of materials and process route of historical products, the products are decomposed into several intermediate products / components from top to bottom according to their manufacturing process, until the parts or work units that cannot be further divided. S1.2. For each part or work unit, identify its complete manufacturing operation chain, which includes all the operation activities that the part / work unit goes through from raw material input to finished product and warehousing, including but not limited to: blanking, forming, machining, heat treatment, surface treatment, assembly, and inspection. S1.3 Establish an activity cost card for each activity. The activity cost card records the cost drivers and corresponding cost element data for the activity. The cost elements include at least: direct material costs, direct labor costs, equipment usage costs, tooling and mold allocation costs, energy consumption costs, and outsourced processing costs. Direct material costs are calculated by multiplying the standard usage of materials consumed in the operation by the unit price of the materials; direct labor costs are calculated by multiplying the standard working hours of the operation by the hourly rate; equipment usage costs are calculated by multiplying the equipment operating time by the equipment depreciation and maintenance costs per unit time; tooling and mold allocation costs are calculated by dividing the total cost of tooling and molds by the expected output over the life cycle and then multiplying by the output of this batch. S1.4. Abstract each cost element into a parameterized cost calculation formula. Each formula includes basic parameter variables and corresponding influence coefficients. The basic parameter variables describe the physical / process characteristics of the product (such as material grade, specifications, weight, and processing accuracy level). The influence coefficients describe the relationship between external conditions and costs (such as material utilization rate, labor cost rate, and equipment depreciation rate). S1.5 Store the activity cost cards and parameterized formulas at each level according to the three-dimensional index of product type, structural level, and activity to form a historical cost model library.

[0023] Step S2: Configure the customer quotation template S2.1. Based on the customer's quotation format requirements, define the cost item hierarchy of the quotation template, including the top-level total price, the middle-level cost categories (such as material costs, processing costs, management fees, and profits), and the bottom-level detailed items (such as the usage and unit price of each specification of material, and the labor hours and rates of each process). S2.2 Establish a mapping table between template cost items and the cost elements, so that each template item corresponds to one or more activity cost elements with aggregation rules (such as summation, weighted average). S2.3 Configure the calculation logic for each cost item in the template, including the reference path of cost elements, aggregate functions, and additional rates (such as management fee rate, tax rate, and target profit margin).

[0024] Step S3: New product feature extraction and historical product matching S3.1 Obtain technical parameters and process requirements information for new products, including at least: product type, main materials and specifications, dimensions, weight, precision level, estimated output, and expected delivery cycle. S3.2 Extract the feature vector of the new product, wherein the feature vector is composed of the quantitative indicators and non-quantitative classification labels in the above-mentioned technical parameters; S3.3 Calculate the similarity between the feature vector of the new product and the features of each historical product in the historical cost model library, and select the top N historical products with similarity exceeding a preset threshold as the reference product set; the similarity calculation comprehensively considers three dimensions: structural similarity, process similarity, and parameter similarity. Structural similarity is determined by calculating the tree edit distance between the new product BOM tree and the historical product BOM tree; process similarity is determined by calculating the Jaccard similarity coefficient between the new product production chain set and the historical product production chain set; parameter similarity is determined by calculating the weighted Euclidean distance of each key specification index; the overall similarity is the weighted sum of the above three dimensions. S3.4 When the similarity of a certain dimension does not meet the threshold but the overall similarity does, mark the cost element corresponding to that dimension as an element to be corrected.

[0025] Step S4: Cost Estimation S4.1 For new products and reference products in the reference product set that are highly matched at the operational activity level (i.e., the operational activity types are the same and the process parameter deviations are within the preset range), the cost data in the historical operational cost card is directly used, and only the quantifiable parameters such as material usage and working hours caused by specification differences are proportionally corrected. S4.2 For the part of the work activity with the same type but the process parameter deviation exceeds the preset range, the key parameter correction method is adopted: extract the cost calculation formula related to the deviation parameter in the work activity, substitute the parameter value corresponding to the new product into the formula to recalculate, and keep the original value of the remaining unchanged parameters. S4.3 For new operations that do not have corresponding work activities in the reference product set, such as new products that adopt new processes that are different from all historical products, the process decomposition estimation method is adopted: the new operation is further decomposed into atomic processes, and the unit cost of each atomic process is calculated based on its process characteristics (such as cutting volume, welding length, spraying area) and basic rates (such as unit cutting cost, unit welding cost, unit spraying cost), and then summarized into the estimated cost of the new operation. S4.4. The estimated costs of the above-mentioned activities are summarized from bottom to top according to the product structure hierarchy to form the activity cost table of the new product. The activity cost table is in the form of products, components, parts and activities as the row dimension and each cost element as the column dimension, forming a complete cost matrix.

[0026] Step S5: Cost Adjustment and Optimization S5.1 Check the elements marked in step S3.4 that need to be corrected, and combine the industry median data or the latest quotation from the supplier to manually or automatically correct the cost elements with large deviations. S5.2. In view of the scale effect of production, based on the estimated production of new products, the fixed cost elements such as equipment allocation and tooling allocation are amortized and adjusted according to the production volume. S5.3. For delivery cycle requirements, apply a corresponding cycle correction factor to operations involving expedited production or outsourcing to in-house manufacturing.

[0027] Step S6: Generate a quotation based on the customer's template S6.1 Based on the mapping relationship between template cost items and cost elements established in step S2, extract or aggregate the corresponding data from the activity cost table and fill it into each cost item of the quotation template. S6.2. Based on the hierarchical structure defined in the template, summarize the details from the bottom layer upwards to generate the intermediate cost categories and the top total price. S6.3. According to the additional fee rate rules configured in the template, add management fees, taxes and profits to the corresponding intermediate or top layer to form the final quotation amount; S6.4 Output a structured quotation, which should simultaneously present: detailed amounts and percentages of each cost item expanded according to the template hierarchy, the traceability path of each cost item to the underlying operational activities, and a summary of key indicators compared with historical products.

[0028] Step S7: After the new product is actually produced and delivered, compare the actual cost data of each operation with the estimated data in steps S4 and S5, calculate the estimated deviation rate of each operation, mark the operation with the deviation rate exceeding the preset learning threshold as the item to be calibrated, update the influence coefficient in the corresponding parameterized cost calculation formula, and add the complete cost data of the new product to the historical cost model library in the format of step S1 for reference optimization of subsequent quotations.

[0029] Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort should fall within the scope of protection of the present invention.

Claims

1. A method for intelligent pricing of new products based on historical production data and cost models, characterized in that, Includes the following steps: S1. Construct a historical cost model library based on historical product data that the enterprise has completed production and delivery; S2. Configure template cost items according to the customer's quotation format, and establish mapping and aggregation rules between template cost items and cost elements; S3. Extract the feature vector of the new product and calculate the similarity with the features of each historical product in the historical cost model library to obtain a reference product set. S4. Summarize the estimated costs of the activities of the reference product set to form a new product activity cost table; S5. Fill the quotation template with the activity cost table, summarize and add the additional rates at each level, and then output a structured quotation.

2. The intelligent pricing method for new products based on historical production data and cost models according to claim 1, characterized in that, In step S1, the method for constructing the historical cost model library includes: decomposing historical products into several parts / work units, determining the various work activities that the parts / work units undergo from raw material input to warehousing, establishing a work cost card for each work activity, the work cost card including cost elements, abstracting the cost elements into parameterized formulas, and storing the work cost cards and parameterized formulas according to a three-dimensional index of product type, structural level, and work activity to form a historical cost model library.

3. The intelligent pricing method for new products based on historical production data and cost models according to claim 1, characterized in that, In step S2, the template cost items include the top-level total price, the intermediate-level cost categories, and the bottom-level detailed items.

4. The intelligent pricing method for new products based on historical production data and cost models according to claim 1, characterized in that, In step S3, feature vectors are extracted based on the technical parameters and process requirements of the new product.

5. The intelligent pricing method for new products based on historical production data and cost models according to claim 1, characterized in that, In step S3, the similarity calculation is based on structural similarity, process similarity, and parameter similarity. When the similarity of a certain dimension does not meet the threshold but the overall similarity does, the cost element corresponding to that dimension is marked as an element to be corrected.

6. The intelligent pricing method for new products based on historical production data and cost models according to claim 1, characterized in that, In step S4, the activity cost table is structured with products, components, parts, and activities as the row dimension and each cost element as the column dimension, forming a cost matrix.

7. The intelligent pricing method for new products based on historical production data and cost models according to claim 1, characterized in that, In step S5, based on the mapping and aggregation rules in step S2, corresponding data is extracted or aggregated from the activity cost table and populated into each cost item of the quotation template.

Images