Parameterization cost estimation method and device for aerospace project cost prediction

Abstract

The invention discloses a parameterization cost estimation method and device for aerospace project cost prediction. The method comprises the following steps: acquiring project historical data; processing the project historical data to obtain preprocessed project historical data and project feature information; based on the preprocessed project historical data and the project feature information, processing a parameterized cost processing model set to obtain a parameterized cost estimation model; and processing project parameter data by using the parameterized cost estimation model to obtain parameter sensitivity information. Aiming at the characteristics of small batch and multiple constraints of product data in the aerospace field, data cleaning based on noise filtering and unstructured text data analysis coding, adaptive algorithm selection of data feature engineering and a multi-dimensional model quality evaluation index system are integrated; and the high-reliability modeling requirement under the characteristics of small subsamples and multi-dimensional data of aerospace products is met.

Description

Technical Field

[0001] This invention relates to the field of aerospace project cost management, and in particular to a parametric cost estimation method and apparatus for aerospace project cost prediction. Background Technology

[0002] (1) Aerospace product data has characteristics such as multi-source heterogeneity (unstructured, high noise, etc.) and small sample size (small sample size). It lacks corresponding automated data cleaning and feature engineering capabilities. In particular, the processing of text data and high-dimensional data requires manual intervention, which can easily lead to data bias (such as misjudgment of missing values, inconsistent text encoding, etc.).

[0003] (2) Aerospace projects involve multiple disciplines and multiple stages. Traditional cost estimation methods rely on expert experience: ① The identification of cost drivers is too subjective and lacks objective means based on data; ② The selection of cost models depends on human experience and lacks an objective, automated recommendation mechanism linked to data features.

[0004] (3) Insufficient model accuracy and generalization ability: Traditional statistical models (such as multiple regression) are difficult to capture nonlinear relationships and multicollinearity problems in aerospace project data, and cannot dynamically adapt to changes in cost drivers of new models and new processes.

[0005] (4) The model evaluation index is too simple and ignores the stringent requirements of the project domain for interpretability and reliability; the parameter sensitivity analysis is not perfect and lacks quantitative impact assessment of cost-driven parameters, making it difficult to support parameter trade-offs and optimization decisions in the design phase.

[0006] In summary, existing cost estimation methods cannot address the high reliability modeling requirements of aerospace products with small prototypes and multi-dimensional data characteristics. There is an urgent need for a fully integrated parameterized cost estimation method that combines machine learning and large-scale modeling technologies to achieve a complete process from intelligent data preprocessing, data feature analysis, dynamic algorithm matching and adaptive modeling, to high-quality evaluation. This invention can be applied to the PACE (Performance and Cost Evaluation) parameterized cost estimation system. By combining machine learning and large-scale language models, it completes a technical closed loop from "data cleaning - data processing - feature analysis - cost modeling - parameter estimation - evaluation and verification," providing aerospace project designers and engineers with a convenient, fast, and integrated cost estimation method that allows for rapid estimation of target costs after the design scheme is finalized. Summary of the Invention

[0007] The technical problem this invention aims to solve is the issues of small sample data, high reliance on manual labor, insufficient accuracy, and non-standard processes in cost estimation for aerospace projects. The parametric cost estimation method and device provided by this invention for aerospace project cost prediction utilizes machine learning and large language modeling technology to propose a parametric cost modeling and cost estimation method that integrates automatic data cleaning, intelligent feature analysis, and adaptive modeling method recommendation and evaluation, thereby achieving high efficiency, accuracy, and interpretability in cost estimation throughout the entire aerospace project process.

[0008] To address the aforementioned technical problems, a first aspect of this invention discloses a parameterized cost estimation method for predicting the cost of aerospace projects, the method comprising:

[0009] S1, retrieve historical data for the project;

[0010] S2, process the project historical data to obtain preprocessed project historical data and project feature information;

[0011] S3, Based on the preprocessed project historical data and the project feature information, the parameterized cost processing model set is processed to obtain the parameterized cost estimation model;

[0012] S4. The parameterized cost estimation model is used to process the project parameter data to obtain parameter sensitivity information.

[0013] As an optional implementation, in the first aspect of the present invention, processing the project historical data to obtain preprocessed project historical data and project feature information includes:

[0014] S21, preprocess the project historical data to obtain preprocessed project historical data;

[0015] S22, Perform feature analysis on the preprocessed project data to obtain project feature information.

[0016] As an optional implementation, in the first aspect of the present invention, the preprocessing of the project historical data to obtain preprocessed project historical data includes:

[0017] S211, The project historical data is parsed and processed to obtain the project data field set;

[0018] The project field set includes several project data fields;

[0019] S212, Perform integrity judgment processing on any of the project data fields to obtain the project data field integrity judgment result;

[0020] If the integrity judgment result of the project data field is yes, execute S214;

[0021] When the integrity judgment result of the project data field is negative, the missing value processing of the project data field is performed to obtain the missing project data field;

[0022] S213, Update the project data field with the missing project data field;

[0023] S214, perform text judgment processing on the project data field to obtain the project data field type judgment result;

[0024] When the type determination result of the project data field is yes, the project data field is encoded to obtain the encoded project data field, and S215 is executed.

[0025] If the result of the data field type determination for the project is negative, execute S216;

[0026] S215, Update the project data field with the encoded project data field;

[0027] S216, Perform data slicing and outlier processing on the project data fields to obtain preprocessed project data fields;

[0028] S217, the preprocessed project data fields are combined in sequence to obtain preprocessed project historical data.

[0029] As an optional implementation, in the first aspect of the present invention, the step of performing feature analysis on the preprocessed project data to obtain project feature information includes:

[0030] S221, The preprocessed project data is analyzed and processed to obtain weighted recommendation sequence data and project feature information;

[0031] S222, Perform feature analysis on the weighted recommendation sequence data and the project feature information to obtain high-influence parameter information;

[0032] S223, Based on the high-influence parameter information, the project characteristic information is determined.

[0033] As an optional implementation, in the first aspect of the present invention, the step of processing the parameterized cost processing model set based on the preprocessed project data and the project feature information to obtain a parameterized cost estimation model includes:

[0034] S31, the preprocessed project data is divided into a training dataset and a test dataset according to a set ratio;

[0035] S32, Based on the project feature information, determine the parameterized cost processing model set;

[0036] S33, Based on the project feature information, the parameterized cost processing model set is trained using the training dataset to determine the optimal parameterized cost processing model;

[0037] S34. Based on the project feature information, the optimal parameterized cost processing model is validated using the test dataset to obtain a parameterized cost estimation model.

[0038] As an optional implementation, in the first aspect of the present invention, the step of processing project parameter data using the parameterized cost estimation model to obtain parameter sensitivity information includes:

[0039] S41, retrieve project parameter data;

[0040] S42, Using the parameterized cost estimation model, cost prediction is performed on the project parameter data to obtain cost prediction information;

[0041] S43, The cost prediction information is analyzed and processed to obtain parameter sensitivity information.

[0042] As an optional implementation, in the first aspect of the present invention, the step of analyzing and processing the cost prediction information to obtain parameter sensitivity information includes:

[0043] S431, Perform single-factor analysis on the cost prediction information to obtain single-factor sensitivity information;

[0044] S432, Perform a two-factor analysis on the cost prediction information to obtain two-factor interaction effect information;

[0045] S433, The single-factor sensitivity information and / or the two-factor interaction effect information are comprehensively processed to obtain parameter sensitivity information.

[0046] The second aspect of this invention discloses a parametric cost estimation device for predicting the cost of aerospace projects, the device comprising: a historical data acquisition module, a preprocessing module, a model building module, and a data processing module;

[0047] The historical data acquisition module is used to acquire historical data of the project;

[0048] The preprocessing module is used to process the historical data of the project to obtain project feature information;

[0049] The model building module is used to process the parameterized cost processing model set based on the preprocessed project data and the project feature information to obtain a parameterized cost estimation model.

[0050] The data processing module is used to process project parameter data using the parameterized cost estimation model to obtain parameter sensitivity information.

[0051] A third aspect of this invention discloses a parameterized cost estimation device for predicting the cost of aerospace projects, the device comprising:

[0052] Memory containing executable program code;

[0053] A processor coupled to the memory;

[0054] The processor calls the executable program code stored in the memory to execute some or all of the steps in the parameterized cost estimation method for predicting aerospace project costs disclosed in the first aspect of the present invention.

[0055] The fourth aspect of the present invention discloses a computer-readable storage medium, characterized in that the computer-readable storage medium stores computer instructions, which, when invoked, execute some or all of the steps in the parameterized cost estimation method for predicting aerospace project costs disclosed in the first aspect of the present invention.

[0056] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0057] This invention employs a parametric cost modeling and cost estimation method for the early design phase of aerospace projects. Addressing the small-batch, multi-constraint characteristics of aerospace product data, it integrates data cleaning based on noise filtering and unstructured text data parsing and encoding, adaptive algorithm selection for data feature engineering, and a multi-dimensional model quality evaluation index system. This solves the high-reliability modeling requirements of aerospace products with small sample sizes and multi-dimensional data characteristics. It is applicable to the full lifecycle cost estimation, project budgeting, and intelligent solution selection for highly complex, small-batch products such as launch vehicles, spacecraft, and satellites. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1This is a schematic diagram illustrating an application scenario of the parameterized cost estimation device for aerospace project cost prediction provided in an embodiment of the present invention.

[0060] Figure 2 This is a flowchart illustrating a parameterized cost estimation method for predicting the cost of aerospace projects, as disclosed in an embodiment of the present invention.

[0061] Figure 3 This is a schematic diagram of the structure of a parameterized cost estimation device for predicting the cost of aerospace projects, as disclosed in an embodiment of the present invention.

[0062] Figure 4 This is a schematic diagram of another parametric cost estimation device for predicting the cost of aerospace projects disclosed in an embodiment of the present invention. Detailed Implementation

[0063] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0064] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0065] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0066] In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0067] It should be noted that since the method in this application embodiment is executed in a computer device, the processing objects of each computer device exist in the form of data or information, such as time, which is essentially time information. It is understood that if size, quantity, position, etc. are mentioned in subsequent embodiments, they are all corresponding data that exist so that the computer device can process them. Specific details will not be elaborated here.

[0068] It should be noted that the artificial intelligence-related technologies that may be involved in this application will be briefly described. Artificial intelligence (AI) is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results. In other words, artificial intelligence is a comprehensive technology in computer science that attempts to understand the essence of intelligence and produce a new kind of intelligent machine that can react in a way similar to human intelligence. Artificial intelligence is the study of the design principles and implementation methods of various intelligent machines, enabling machines to have the functions of perception, reasoning, and decision-making.

[0069] Artificial intelligence (AI) is a comprehensive discipline encompassing a wide range of fields, including both hardware and software technologies. Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies primarily include computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0070] This application provides a parameterized cost estimation method, apparatus, computer equipment, and computer-readable storage medium for predicting the cost of aerospace projects, which will be described in detail below.

[0071] Please see Figure 1 , Figure 1This is a schematic diagram illustrating a scenario in which the parameterized cost estimation device for aerospace project cost prediction provided in this application embodiment is applied in an aerospace project management system. The aerospace project management system may include a computer device 100, which integrates the parameterized cost estimation device for aerospace project cost prediction. Figure 1 Computer equipment in the country.

[0072] In this embodiment, the computer device 100 can be a standalone server, a server network, or a server cluster. For example, the computer device 100 described in this embodiment includes, but is not limited to, a computer, a network host, a single network server, a set of multiple network servers, or a cloud server composed of multiple servers. The cloud server is composed of a large number of computers or network servers based on cloud computing.

[0073] It is understood that the computer device 100 used in the embodiments of this application can be a device that includes both receiving and transmitting hardware, that is, a device having receiving and transmitting hardware capable of performing bidirectional communication on a bidirectional communication link. Such a device may include: cellular or other communication devices having a single-line display, a multi-line display, or a cellular or other communication device without a multi-line display. Specifically, the computer device 100 may be a desktop terminal or a mobile terminal, and may also be one of a mobile phone, tablet computer, laptop computer, etc.

[0074] Those skilled in the art will understand that Figure 1 The application environment shown is merely one application scenario of the solution in this application and does not constitute a limitation on the application scenario of the solution in this application. Other application environments may include those that are more specific to this application. Figure 1 The number of computer devices shown is more or less, for example Figure 1 Only one computer device is shown in the diagram. It is understood that the system may also include one or more other services, which are not limited here.

[0075] In addition, such as Figure 1 As shown, the aerospace project management system may also include a memory 200 for storing historical project data, parameter sensitivity analysis data, and parameter cost optimization information.

[0076] It should be noted that, Figure 1The schematic diagram illustrating the application scenario of the parameterized cost estimation device for aerospace project cost prediction is merely an example. The simulation system and scenario described in this application embodiment are intended to more clearly illustrate the technical solutions of this application embodiment and do not constitute a limitation on the technical solutions provided in this application embodiment. As those skilled in the art will know, with the evolution of simulation control management systems and the emergence of new business scenarios, the technical solutions provided in this application embodiment are also applicable to similar technical problems.

[0077] This invention discloses a parametric cost estimation method and apparatus for aerospace project cost prediction. Addressing the characteristics of small-batch, multi-constraint aerospace product data, it employs a parametric cost model integrating data cleaning based on noise filtering and unstructured text data parsing and encoding, adaptive algorithm selection for data feature engineering, and multi-dimensional model quality evaluation indicators. This achieves parametric cost estimation processing and solves the high-reliability modeling requirements of aerospace products with small sample sizes and multi-dimensional data characteristics. Detailed explanations follow.

[0078] Example 1

[0079] Please see Figure 2 , Figure 2 This is a flowchart of a parameterized cost estimation method for aerospace project cost prediction, as disclosed in an embodiment of the present invention. Figure 2 The described parameterized cost estimation method for aerospace project cost prediction is applied in an aerospace project management system, such as a local server or cloud server within the system; this embodiment of the invention is not limited to this. Figure 2 As shown, the parameterized cost estimation method for predicting the cost of aerospace projects includes:

[0080] S1, retrieve historical data for the project;

[0081] It should be noted that the historical data of the project includes, but is not limited to, the number of equipment, equipment type, equipment capacity, fuel consumption, and unit price of equipment.

[0082] It should be noted that, in this embodiment, the project historical data is a space project dataset in CSV (Comma-Separated Values) format;

[0083] It should be noted that after the historical data of the project is successfully imported, the partitioning results of the candidate sets of independent and dependent variables will be automatically displayed, showing the current data dimensions;

[0084] It should be noted that, in this embodiment, the data dimensions include: airframe development cost, engine development cost, first airframe manufacturing cost, first engine manufacturing cost, fuel cost, maintenance cost, aircraft price, airframe price, engine price, cruise speed, maximum range, number of crew members, fuel type, side takeoff noise, supersonic detonation, displacement, fuel consumption rate, wingspan, cabin diameter, length, maximum takeoff weight, maximum structural weight, engine thrust, cruise altitude, equilibrium field length, lift-to-drag ratio in supersonic cruise, aircraft propellant mass ratio, takeoff thrust-to-weight ratio, effective dose rate, number of addressable routes, and number of aircraft sold.

[0085] S2, process the project historical data to obtain preprocessed project historical data and project feature information;

[0086] S3, Based on the preprocessed project historical data and the project feature information, the parameterized cost processing model set is processed to obtain the parameterized cost estimation model;

[0087] S4. The parameterized cost estimation model is used to process the project parameter data to obtain parameter sensitivity information.

[0088] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment addresses the small-batch, multi-constraint characteristics of aerospace product data. It integrates data cleaning based on noise filtering and unstructured text data parsing and encoding, data feature extraction to process historical project data, adaptive algorithm selection, and a multi-dimensional model quality evaluation index system. This solves the high reliability modeling requirements of aerospace products with small sample sizes and multi-dimensional data characteristics. It is suitable for full life cycle cost estimation, project budget preparation, and intelligent scheme selection for highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0089] Optionally, the processing of the project historical data to obtain preprocessed project historical data and project feature information includes:

[0090] S21, preprocess the project historical data to obtain preprocessed project historical data;

[0091] S22, Perform feature analysis on the preprocessed project data to obtain project feature information.

[0092] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment processes the project's historical data to obtain preprocessed project historical data and project feature information, providing data support for subsequent adaptive algorithm selection and multi-dimensional model quality evaluation. It solves the high reliability modeling requirements of aerospace products with small samples and multi-dimensional data features, and is applicable to the full life cycle cost estimation, project budget preparation, and intelligent scheme selection of highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0093] Optionally, the preprocessing of the project historical data to obtain preprocessed project historical data includes:

[0094] S211, The project historical data is parsed and processed to obtain the project data field set;

[0095] The project field set includes several project data fields;

[0096] S212, Perform integrity judgment processing on any of the project data fields to obtain the project data field integrity judgment result;

[0097] It should be noted that the integrity judgment process refers to performing a validation process on the project data fields according to the CSV format to obtain the integrity judgment result of the project data fields:

[0098] If the integrity judgment result of the project data field is yes, execute S214;

[0099] When the integrity judgment result of the project data field is negative, the missing value processing of the project data field is performed to obtain the missing project data field;

[0100] It should be noted that the missing value handling includes three modes: "deletion", "autofill" and "custom fill", which detect the missing ratio through the underlying algorithm and recommend a handling strategy.

[0101] It should be noted that the recommended processing strategies include, but are not limited to, prioritizing deletion for low missing values ​​and using KNN interpolation for high missing values;

[0102] It should be noted that in this embodiment, the "fuel consumption" field is automatically detected and identified to have a small number of missing values, and the "automatic filling" missing value processing method based on mean interpolation is recommended. The missing value filling model based on mean interpolation is used to fill the historical data of the project to obtain the missing project data field.

[0103] S213, Update the project data field with the missing project data field;

[0104] S214, perform text judgment processing on the project data field to obtain the project data field type judgment result;

[0105] It should be noted that the text judgment processing mentioned above refers to performing text judgment processing on the project data fields according to CSV format to obtain the project data field type judgment result:

[0106] When the type determination result of the project data field is yes, the project data field is encoded to obtain the encoded project data field, and S215 is executed.

[0107] It should be noted that the encoding process includes low-radix column encoding process and high-radix column encoding process;

[0108] It should be noted that in this embodiment, the text field "engine type" is automatically processed using one-hot encoding as a low cardinality column, while "sales region" is processed using improved Bayesian target encoding as a high cardinality column.

[0109] If the result of the data field type determination for the project is negative, execute S216;

[0110] S215, Update the project data field with the encoded project data field;

[0111] S216, Perform data slicing and outlier processing on the project data fields to obtain preprocessed project data fields;

[0112] It should be noted that the data slicing and outlier processing mentioned above include both data slicing processing and outlier processing.

[0113] It should be noted that the data slicing process supports text-based label column filtering and allows users to customize filtering conditions, such as classifying aerospace products as satellites, launch vehicles, etc.; the outlier processing integrates algorithms such as Isolation Forest, Z-score (a standard score that measures the position of a data point relative to the population mean), IQR (interquartile range, an indicator used to describe the dispersion of data distribution), DBSCAN (Density-Based Spatial Clustering of Applications with Noise, a density-based clustering algorithm that can discover clusters of arbitrary shapes and process noisy data), Mahalanobis distance to automatically detect and identify outliers, and visualize abnormal data through box plots and scatter matrices;

[0114] It should be noted that in this embodiment, an improved random forest algorithm model is used to detect outliers at a pollution ratio of 0.05, identify the corresponding outliers and delete the fields, and provide outlier visualization analysis through variable distribution box plots and scatter matrices.

[0115] S217, the preprocessed project data fields are combined in sequence to obtain preprocessed project historical data;

[0116] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment preprocesses the project's historical data to obtain preprocessed project historical data, providing data support for subsequent feature analysis and adaptive algorithm selection, and multi-dimensional model quality evaluation. It solves the high reliability modeling requirements of aerospace products with small sample sizes and multi-dimensional data characteristics, and is applicable to the full life cycle cost estimation, project budget preparation, and intelligent scheme selection of highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0117] Optionally, the missing value processing of the project data field to obtain the missing project data field includes:

[0118] S2121, Perform missing value detection on the project data fields to obtain the field missing rate value;

[0119] It should be noted that the expression for the missing rate value of the field is as follows:

[0120]

[0121] Where A represents the missing rate value of the field; Nb represents the standard length of the project field; N A The actual length of the project field;

[0122] It should be noted that the standard length Nb of the project field is obtained based on the type of the project field;

[0123] S2122, Determine whether the missing rate value of the field is greater than the missing rate threshold, and obtain the missing rate determination result;

[0124] It should be noted that the missing rate threshold is set to A = 0.3.

[0125] If the missing rate determination result is negative, delete the project data field and execute S2121;

[0126] If the missing rate determination result is yes, execute S2123;

[0127] S2123, perform missing value filling processing on the project data field to obtain the missing project data field;

[0128] It should be noted that in this embodiment, the "fuel consumption" field is automatically detected and identified to have a small number of missing values. The missing values ​​are filled using a mean interpolation missing value filling model to obtain the missing item data field.

[0129] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment processes the missing values ​​of the project data fields to obtain missing project data fields, providing data support for subsequent data preprocessing, feature analysis and processing, adaptive algorithm selection, and multi-dimensional model quality evaluation. It solves the high reliability modeling requirements of aerospace products with small samples and multi-dimensional data characteristics, and is applicable to the full life cycle cost estimation, project budget preparation, and intelligent scheme selection of highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0130] Optionally, the expression for the missing value imputation model based on mean interpolation is:

[0131]

[0132] in, This represents the filling result for missing values ​​in group j; J represents the mean of non-missing values ​​in the j-th item field; J represents the set of item fields. I(j) represents the total mean of non-missing values ​​in the project field set; α represents the correction coefficient; I(j) i =j) represents the group identifier function; i represents the index of the item field;

[0133] It should be noted that the correction coefficient α takes values ​​in the range of [0, 1].

[0134] It should be noted that in this embodiment, the correction coefficient α is set to α = 0.3;

[0135] It should be noted that the value is 1 if sample i belongs to group j, and 0 otherwise;

[0136] It should be noted that the missing value imputation model based on mean interpolation is used to reduce the interference of population heterogeneity through group division.

[0137] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment uses the mean interpolation missing value filling model to process the missing values ​​of the project's historical data, obtaining missing value processed historical data. This provides data support for subsequent data preprocessing, feature analysis and processing, adaptive algorithm selection, and multi-dimensional model quality evaluation. It solves the high reliability modeling requirements of aerospace products with small sample sizes and multi-dimensional data characteristics. It is applicable to the full life cycle cost estimation, project budget preparation, and intelligent scheme selection of highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0138] Optionally, the encoding process for the project data field to obtain the encoded project data field includes:

[0139] S2141, Obtain the cardinality column of the project data field;

[0140] S2142, determine whether the cardinality column is greater than the cardinality threshold, and obtain the cardinality column judgment result;

[0141] It should be noted that, in this embodiment, the base threshold is set to 3;

[0142] If the result of the determination of the cardinality column is negative, execute S2143;

[0143] If the determination result of the cardinality column is yes, execute S2144;

[0144] S2143, The project data field is encoded using a low-radix column encoding processing model to obtain the encoded project data field;

[0145] It should be noted that the low cardinality column encoding processing means that when the cardinality column is ≤3, the low cardinality column encoding processing model is used to perform low cardinality column encoding processing on the project data field;

[0146] S2144, Perform high cardinality column encoding on the project data field to obtain the encoded project data field;

[0147] It should be noted that the high cardinality sequence encoding processing means that when the cardinality sequence is >3, the optimal encoding method is recommended by algorithms such as Bayesian target encoding, frequency encoding, label encoding and effect encoding, and the class discrimination is evaluated by combining the entropy value and F statistic. This embodiment does not limit this.

[0148] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment encodes the project data fields to obtain coded project data fields, providing data support for subsequent data preprocessing, feature analysis, adaptive algorithm selection, and multi-dimensional model quality evaluation. This solves the high reliability modeling requirements of aerospace products with small samples and multi-dimensional data characteristics, and is applicable to the full life cycle cost estimation, project budget preparation, and intelligent scheme selection of highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0149] Optionally, the low-radix sequence encoding processing expression is:

[0150]

[0151] in, This represents the hierarchical encoding vector for category n; This represents the parent class of category n at level m; This represents the one-hot encoded vector of the m-th layer; M represents the total number of layers; β mThis represents the weight coefficient of the m-th layer;

[0152] It should be noted that the length of the encoding vector is the sum of the one-hot encoding lengths of each level;

[0153] It should be noted that the higher the level, the larger the weighting coefficient.

[0154] It should be noted that the low cardinality column encoding process preserves the hierarchical relationship between categories by weight superposition, solves the problem of category isolation in traditional one-hot encoding, and improves the model's ability to capture semantic associations;

[0155] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment utilizes the low cardinality column encoding process to encode the project's historical data, obtaining coded project data fields. This provides data support for subsequent data preprocessing, feature analysis, adaptive algorithm selection, and multi-dimensional model quality evaluation. It addresses the high reliability modeling requirements of aerospace products with small sample sizes and multi-dimensional data characteristics. It is applicable to the full life-cycle cost estimation, project budget preparation, and intelligent scheme selection for highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0156] Optionally, the step of performing feature analysis on the preprocessed project data to obtain project feature information includes:

[0157] S221, The preprocessed project data is analyzed and processed to obtain weighted recommendation sequence data and project feature information;

[0158] S222, Perform feature analysis on the weighted recommendation sequence data and the project feature information to obtain high-influence parameter information;

[0159] It should be noted that in this embodiment, the core cost drivers are identified by comprehensively analyzing the correlation, sensitivity, PCA (Principal Component Analysis) weights and VIF (variance inflation factor) values ​​of the basic features. These are then supplemented by descriptive statistics, correlation coefficient matrices, heatmaps, and variable relationship visualization interfaces to determine the high-impact parameter information.

[0160] It should be noted that, in this example, the high-impact parameters include "year of production", "model", "number of engines", and "engine type".

[0161] S223, Based on the high-influence parameter information, determine the project characteristic information;

[0162] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment performs feature analysis on the preprocessed project data to obtain project feature information, providing data support for subsequent adaptive algorithm selection and multi-dimensional model quality evaluation. It solves the high reliability modeling requirements of aerospace products with small samples and multi-dimensional data features, and is applicable to the full life cycle cost estimation, project budget preparation, and intelligent scheme selection of highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0163] Optionally, the step of processing the parameterized cost processing model set based on the preprocessed project data and the project feature information to obtain a parameterized cost estimation model includes:

[0164] S31, the preprocessed project data is divided into a training dataset and a test dataset according to a set ratio;

[0165] S32, Based on the project feature information, determine the parameterized cost processing model set;

[0166] It should be noted that the parameterized cost processing model set includes, but is not limited to, ordinary least squares model, PLS (Partial Least Squares Regression) power regression model, PLS linear regression model, power function model, exponential function model, random forest regression model and XGBoost (an improved version of gradient boosting tree that can train models faster and more efficiently) regression model, and this embodiment does not impose any restrictions.

[0167] S33, Based on the project feature information, the parameterized cost processing model set is trained using the training dataset to determine the optimal parameterized cost processing model;

[0168] It should be noted that this embodiment uses cross-validation and the optuna (a Python library for hyperparameter optimization) optimization method based on Bayesian optimization to compare the performance of different models on the training dataset and determine the optimal parameterized cost processing model.

[0169] It should be noted that in this embodiment, the analysis of the current VIF value shows high collinearity, and based on the underlying model selection principle, three variable fitting methods are automatically recommended: PLS linear regression, random forest regression, and XGBoost regression.

[0170] It should be noted that in this embodiment, the PLS linear regression model is selected to train the automatically partitioned 80% training set to obtain the optimal model equation. After training, the R2 reaches 0.9914. The optimal model equation is the optimal parameterized cost processing model.

[0171] S34. Based on the project feature information, the optimal parameterized cost processing model is verified using the test dataset to obtain the parameterized cost estimation model.

[0172] It should be noted that for ordinary least squares models, PLS power regression models, PLS linear regression models, power function models, and exponential function models, model coefficient testing and feature testing techniques are used to conveniently check the significance and stability of the coefficients and the above models; for random forest regression models and XGBoost regression models, model feature testing techniques are used to conveniently check the significance and stability of the features.

[0173] It should be noted that in this embodiment, model coefficient tests and feature tests were conducted using bootstrap sampling analysis. The significance and stability of the coefficients met the requirements, and the model was evaluated.

[0174] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment processes the parameterized cost processing model set based on the preprocessed project data and the project feature information to obtain the parameterized cost estimation model. It solves the high reliability modeling requirements under the small sample and multi-dimensional data characteristics of aerospace products, and is applicable to the whole life cycle cost estimation, project budget preparation, and intelligent scheme selection of highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0175] Optionally, the step of processing the project parameter data using the parameterized cost estimation model to obtain parameter sensitivity information includes:

[0176] S41, retrieve project parameter data;

[0177] S42, Using the parameterized cost estimation model, cost prediction is performed on the project parameter data to obtain cost prediction information;

[0178] S43, The cost prediction information is analyzed and processed to obtain parameter sensitivity information.

[0179] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment uses the parameterized cost estimation model to process project parameter data, obtain parameter sensitivity information, and obtain optimized cost parameters for aerospace products with multi-sample and multi-dimensional data characteristics. It solves the high reliability modeling requirements of aerospace products with small sample sizes and multi-dimensional data characteristics, and is applicable to the full life cycle cost estimation, project budget preparation, and intelligent scheme selection of highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0180] Optionally, the step of analyzing and processing the cost prediction information to obtain parameter sensitivity information includes:

[0181] S431, Perform single-factor analysis on the cost prediction information to obtain single-factor sensitivity information;

[0182] It should be noted that the single-factor analysis refers to quantifying the impact of changes in the high-impact parameters on costs, and outputting sensitivity rankings and asymmetric analysis.

[0183] It should be noted that the sensitivity refers to the degree of influence of high-influence parameters on the percentage change of the target variable when the same percentage change occurs.

[0184] It should be noted that the single-factor analysis process refers to performing a single high-influence parameter analysis on all high-influence parameters to obtain single-factor sensitivity information;

[0185] It should be noted that the single-factor sensitivity information refers to the sensitivity ranking of the high-influence parameter;

[0186] S432, Perform a two-factor analysis on the cost prediction information to obtain two-factor interaction effect information;

[0187] It should be noted that the two-factor analysis means providing a heatmap of the interaction effect between any two parameters, as well as suggestions for adjusting the interaction effect, effect intensity classification, and domain experience references to assist users in adjusting design parameters.

[0188] It should be noted that the two-factor analysis refers to the analysis of two high-influence parameters among all high-influence parameters using a heatmap of the interaction effect between two parameters, or suggestions for adjusting the interaction effect, or graded effect intensity, or domain experience reference, to obtain information on the two-factor interaction effect.

[0189] It should be noted that the two-factor interaction effect refers to the degree of influence of adjusting two parameters simultaneously on the percentage change of the target variable;

[0190] S433, The single-factor sensitivity information and / or the two-factor interaction effect information are comprehensively processed to obtain parameter sensitivity information;

[0191] It should be noted that the comprehensive processing refers to sorting the single-factor sensitivity information and the two-factor interaction effect information according to the parameter sensitivity values ​​to obtain parameter sensitivity information;

[0192] It should be noted that in this embodiment, the validated PLS linear regression model was solidified into a parameterized cost estimation model. Input parameters included: maximum structural weight 105,564 lbs, engine thrust 100,221 lbs, wingspan 56 ft, supersonic cruise time 3292 seconds, number of addressable routes 17.50, and number of sold aircraft 136. The final predicted airframe development cost was $970 million. Single-factor analysis with a parameter variation range of 10% was used for parameter sensitivity analysis to obtain a parameter sensitivity ranking, i.e., analyzing which parameter changes (including positive and negative changes) had a significant impact on cost, thereby dynamically adjusting design parameters.

[0193] As can be seen, the parameterized cost estimation method for aerospace project cost prediction described in this embodiment analyzes and processes the cost prediction information to obtain parameter sensitivity information, and uses the parameterized cost estimation model to obtain optimized cost parameters for aerospace products with multi-sample and multi-dimensional data characteristics. This solves the high reliability modeling requirements of aerospace products with small sample sizes and multi-dimensional data characteristics, and is applicable to the full life cycle cost estimation, project budget preparation, and intelligent scheme selection of highly complex, small-batch products such as launch vehicles, spacecraft, and satellites.

[0194] Example 2

[0195] Please see Figure 3 , Figure 3 This is a schematic diagram of a parametric cost estimation device for aerospace project cost prediction, as disclosed in an embodiment of the present invention. Figure 3 The described device can be applied to aerospace project management systems, such as local servers or cloud servers used in aerospace project management systems, and the embodiments of the present invention are not limited thereto. Figure 3 As shown, the device includes: a historical data acquisition module 101, a preprocessing module 102, a model building module 103, and a data processing module 104;

[0196] The historical data acquisition module 101 is used to acquire historical data of the project;

[0197] The preprocessing module 102 is used to process the historical data of the project to obtain project feature information;

[0198] The model building module 103 is used to process the parameterized cost processing model set based on the preprocessed project data and the project feature information to obtain a parameterized cost estimation model.

[0199] The data processing module 104 is used to process project parameter data using the parameterized cost estimation model to obtain parameter sensitivity information.

[0200] The device provided in this embodiment has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the parameterized cost estimation method for aerospace project cost prediction described in the aforementioned embodiment one.

[0201] As can be seen, the simulator simulation interconnection system based on data subscription described in this embodiment uses a highly simulated full-state training simulation environment and a data subscription method for interconnection and communication. It supports the systematic learning of trainees on the entire process of target equipment operation, improves the timeliness of data communication and the simulation degree of the training simulation environment, meets the needs of high-volume data communication, and solves the problem of insufficient practical teaching and training equipment in the simulation training process.

[0202] Example 3

[0203] Please see Figure 4 , Figure 4 This is a schematic diagram of another parametric cost estimation device for aerospace project cost prediction disclosed in an embodiment of the present invention. Figure 4 The described device can be applied to aerospace project management systems, such as local servers or cloud servers used in aerospace project management systems, and the embodiments of the present invention are not limited thereto. Figure 4 As shown, the device may include:

[0204] Memory 201 storing executable program code;

[0205] Processor 202 coupled to memory 201;

[0206] The processor 202 calls the executable program code stored in the memory 201 to execute the steps in the parameterized cost estimation method for predicting aerospace project costs described in Embodiment 1.

[0207] Example 4

[0208] This invention discloses a computer-readable storage medium storing a computer program for electronic data interchange, wherein the computer program causes a computer to execute the steps in the parameterized cost estimation method for predicting aerospace project costs described in Embodiment 1.

[0209] Example 5

[0210] This invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to perform the steps in the parameterized cost estimation method for predicting aerospace project costs described in Embodiment 1.

[0211] The device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0212] Through the detailed description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0213] Finally, it should be noted that the parameterized cost estimation method and apparatus for aerospace project cost prediction disclosed in the embodiments of the present invention are merely preferred embodiments of the present invention and are only used to illustrate the technical solutions of the present invention, not to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A parameterized cost estimation method for predicting the cost of aerospace projects, characterized in that, The method includes: S1, retrieve historical data for the project; S2, process the project historical data to obtain preprocessed project historical data and project feature information; S3, Based on the preprocessed project historical data and the project feature information, the parameterized cost processing model set is processed to obtain the parameterized cost estimation model; S4. The parameterized cost estimation model is used to process the project parameter data to obtain parameter sensitivity information.

2. The parameterized cost estimation method for aerospace project cost prediction according to claim 1, characterized in that, The process of processing the project's historical data to obtain preprocessed project historical data and project feature information includes: S21, preprocess the project historical data to obtain preprocessed project historical data; S22, Perform feature analysis on the preprocessed project data to obtain project feature information.

3. The parameterized cost estimation method for aerospace project cost prediction according to claim 2, characterized in that, The preprocessing of the project's historical data to obtain preprocessed project historical data includes: S211, The project historical data is parsed and processed to obtain the project data field set; The project field set includes several project data fields; S212, Perform integrity judgment processing on any of the project data fields to obtain the project data field integrity judgment result; If the integrity judgment result of the project data field is yes, execute S214; When the integrity judgment result of the project data field is negative, the missing value processing of the project data field is performed to obtain the missing project data field; S213, Update the project data field with the missing project data field; S214, perform text judgment processing on the project data field to obtain the project data field type judgment result; When the type determination result of the project data field is yes, the project data field is encoded to obtain the encoded project data field, and S215 is executed. If the result of the data field type determination for the project is negative, execute S216; S215, Update the project data field with the encoded project data field; S216, Perform data slicing and outlier processing on the project data fields to obtain preprocessed project data fields; S217, All the preprocessed project data fields are combined in sequence to obtain the preprocessed project historical data.

4. The parameterized cost estimation method for aerospace project cost prediction according to claim 2, characterized in that, The process of performing feature analysis on the preprocessed project data to obtain project feature information includes: S221, The preprocessed project data is analyzed and processed to obtain weighted recommendation sequence data and project feature information; S222, Perform feature analysis on the weighted recommendation sequence data and the project feature information to obtain high-influence parameter information; S223, Based on the high-influence parameter information, the project characteristic information is determined.

5. The parameterized cost estimation method for aerospace project cost prediction according to claim 1, characterized in that, The step of processing the parameterized cost processing model set based on the preprocessed project historical data and the project feature information to obtain a parameterized cost estimation model includes: S31, the preprocessed historical data of the project is divided into a training dataset and a test dataset according to a set ratio; S32, Based on the project feature information, determine the parameterized cost processing model set; S33, Based on the project feature information, the parameterized cost processing model set is trained using the training dataset to determine the optimal parameterized cost processing model; S34. Based on the project feature information, the optimal parameterized cost processing model is validated using the test dataset to obtain a parameterized cost estimation model.

6. The parameterized cost estimation method for aerospace project cost prediction according to claim 1, characterized in that, The process of using the parameterized cost estimation model to process project parameter data and obtain parameter sensitivity information includes: S41, retrieve project parameter data; S42, Using the parameterized cost estimation model, cost prediction is performed on the project parameter data to obtain cost prediction information; S43, The cost prediction information is analyzed and processed to obtain parameter sensitivity information.

7. The parameterized cost estimation method for aerospace project cost prediction according to claim 6, characterized in that, The analysis and processing of the cost prediction information to obtain parameter sensitivity information includes: S431, Perform single-factor analysis on the cost prediction information to obtain single-factor sensitivity information; S432, Perform a two-factor analysis on the cost prediction information to obtain two-factor interaction effect information; S433, The single-factor sensitivity information and / or the two-factor interaction effect information are comprehensively processed to obtain parameter sensitivity information.

8. A parameterized cost estimation device for predicting the cost of aerospace projects, characterized in that, The device includes: a historical data acquisition module, a preprocessing module, a model building module, and a data processing module; The historical data acquisition module is used to acquire historical data of the project; The preprocessing module is used to process the historical data of the project to obtain project feature information; The model building module is used to process the parameterized cost processing model set based on the preprocessed project data and the project feature information to obtain a parameterized cost estimation model. The data processing module is used to process project parameter data using the parameterized cost estimation model to obtain parameter sensitivity information.

9. A parameterized cost estimation device for predicting the cost of aerospace projects, characterized in that, The device includes: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the parameterized cost estimation method for predicting aerospace project costs as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when invoked, execute the parameterized cost estimation method for predicting aerospace project costs as described in any one of claims 1-7.

Images