Method and system for rapid cost evaluation in engineering change management

Abstract

The application provides a cost rapid evaluation method and system in engineering change management, relates to the technical field of engineering cost management, and comprises the following steps: acquiring an engineering type, establishing an engineering knowledge base; defining a change influence network, performing logical compilation and training on the change feature-variable-change influence network, and constructing an engineering evaluator; uploading an engineering change order to an engineering management system by an engineering party, triggering the engineering evaluator, performing change feature self-attention import and coding, performing attribution based on implicit variables and explicit variables, converting into a change influence network group, performing cost evaluation and comprehensive calculation, and determining a change evaluation order; and performing zero-knowledge proof on the cost rationality of the change evaluation order by an auditing party, and displaying on a terminal interface. The application solves the technical problem that, in the prior art, the change influence chain is long, the change influence analysis is not comprehensive, and the accuracy of cost evaluation is affected, and improves the accuracy of engineering change cost evaluation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of engineering cost management, and particularly relates to a cost rapid evaluation method and system in engineering change management. BACKGROUND

[0002] In engineering change management, the influence of changes is usually not limited to direct adjustment or replacement, but will be transmitted among multiple stages and different fields in the project life cycle, forming an influence chain with layer-by-layer association, so that the final cost impact of changes is difficult to fully sort out and accurately predict in a short time. At present, the existing cost evaluation method usually only analyzes the direct impact brought by changes, lacks the ability of systematic modeling and quantitative processing of indirect influence, transmission path and variable interaction, resulting in the problems of insufficient coverage and low accuracy of cost evaluation results.

[0003] In summary, the prior art has the technical problem that the influence analysis of changes is not comprehensive due to the long influence chain of engineering changes, thereby affecting the accuracy of cost evaluation. SUMMARY

[0004] The purpose of the present application is to provide a cost rapid evaluation method and system in engineering change management, to solve the technical problem in the prior art that the influence analysis of changes is not comprehensive due to the long influence chain of engineering changes, thereby affecting the accuracy of cost evaluation.

[0005] In view of the above problems, the present application provides a cost rapid evaluation method and system in engineering change management.

[0006] In a first aspect, the present application provides a cost rapid evaluation method in engineering change management, which is realized by a cost rapid evaluation system in engineering change management, wherein the cost rapid evaluation method in engineering change management comprises: obtaining an engineering type, and establishing an engineering knowledge base; defining a change influence network according to the engineering knowledge base, performing logical compilation and training with a change feature-variable-change influence network, and constructing an engineering evaluator, wherein the engineering evaluator is embedded and deployed in an engineering management system, and the change influence dimension at least includes materials, working hours and energy consumption; an engineering party uploads an engineering change sheet to the engineering management system through a first data port, so as to trigger the engineering evaluator with a term determination, perform change feature self-attention import and coding on the input side, perform attribution based on implicit variables and explicit variables, convert into a change influence network group, map the influence transmission of network nodes to perform cost evaluation and comprehensive calculation, and determine a change evaluation sheet; an auditing party performs zero-knowledge proof on the cost rationality of the change evaluation sheet according to a second data port, and displays on a terminal interface of the engineering management system.

[0007] Optionally, for the material dimension, a change impact feature is defined according to a knowledge part in the engineering knowledge base, wherein the change impact feature is determined at least according to a unit material cost and a utilization mode; a center node is defined by a material category, an extended node is defined by a material feature, an edge is defined by a cost relationship, and a first change impact network is constructed; and the first change impact network is added to the change impact network.

[0008] Optionally, according to the engineering knowledge base, an explicit variable set and an implicit variable set are defined, wherein each implicit variable is identified by a reasoning relationship; a first data port is deployed by a self-attention of a change feature, a second attribution node is deployed based on a variable definition of the explicit variable set and the implicit variable set, and a third conversion node is deployed based on a conversion of the change impact network; the first data port, the second attribution node, and the third conversion node are cascaded to construct the engineering evaluator.

[0009] Optionally, the engineering knowledge base is embedded into the engineering evaluator to determine an evaluation logic based on the cascaded nodes; an engineering record is called to integrate the engineering record based on the evaluation logic of the engineering change to determine sample data; and the engineering evaluator is supervised and trained to convergence according to the sample data.

[0010] Optionally, an engineering change order is uploaded to trigger the engineering evaluator by performing a term test, wherein a term is defined by a change semantic; the engineering change order is subjected to a self-attention recognition based on a change feature according to the first data port, and an engineering change feature is encoded and imported; a variable recognition measurement and attribution based on the engineering change feature are performed based on the second attribution node, and a change impact network conversion and a cost evaluation are performed based on the third conversion node to determine the change evaluation order.

[0011] Optionally, a first transmission mode and a second transmission mode are divided, wherein the first transmission mode is a symmetric impact transmission, and the second transmission mode is an asymmetric impact transmission; a symmetric determination is performed by whether an impact weight changes before and after a change; a change impact network group is mapped and transmission-identified based on the first transmission mode and the second transmission mode as constraints, and a cost evaluation based on impact transmission is performed.

[0012] Optionally, the change impact network includes a first network and a second network, wherein the first network is a network matrix before a change, and the second network is a network matrix after a change; the first network and the second network are mapped, a transmission mode determination and a transmission impact coefficient identification are performed for each mapping group, a cost evaluation and a comprehensive analysis under a transmission impact coefficient constraint for each mapping group are performed, and the change evaluation order is generated.

[0013] Optionally, a reference string is introduced, wherein the reference string is set by consensus of the engineering party and the auditing party; a trusted proof based on the engineering change order is set according to the reference string; the trusted proof is written into a second data port, and the change evaluation order is verified.

[0014] Optionally, engineering change data is generated based on the engineering change order and the change evaluation order; the engineering planning is replaced based on the engineering change data, and the terminal interface of the engineering management system is displayed.

[0015] In a second aspect, the application further provides a cost rapid evaluation system in engineering change management, which is used to execute the cost rapid evaluation method in engineering change management as described in the first aspect, wherein the cost rapid evaluation system in engineering change management comprises: a knowledge base establishment module, which is used to acquire an engineering type and establish an engineering knowledge base; an evaluator construction module, which is used to define a change influence network according to the engineering knowledge base, perform logical compilation and training based on the change influence network, and construct an engineering evaluator, wherein the engineering evaluator is embedded and deployed in an engineering management system, and the change influence dimension at least comprises material, working hours, and energy consumption; and an engineering evaluation module, which is used to upload an engineering change order to the engineering management system by the engineering party through a first data port, trigger the engineering evaluator by word entry, perform change feature self-attention import and coding on the input side, execute attribution based on implicit variables and explicit variables, convert into a change influence network group, perform influence transmission of network nodes, and perform cost evaluation and comprehensive calculation to determine a change evaluation order; and a zero-knowledge proof module, which is used to perform zero-knowledge proof of cost rationality of the change evaluation order by the auditing party according to a second data port, and display the terminal interface of the engineering management system.

[0016] One or more technical solutions provided in the application have at least the following beneficial effects:

[0017] By acquiring the engineering type, an engineering knowledge base is established; according to the engineering knowledge base, a change impact network is defined, a logic compilation and training are performed on a change feature-variable-change impact network, and an engineering evaluator is constructed, wherein the engineering evaluator is embedded in an engineering management system, and the change impact dimensions at least include materials, working hours and energy consumption; an engineering change order is uploaded to the engineering management system by an engineering party through a first data port, so that the engineering evaluator is triggered by a term determination, change feature self-attention import and coding are performed on the input side, attribution based on implicit variables and explicit variables is performed, a change impact network group is converted, influence transmission of network nodes is mapped, cost evaluation and comprehensive calculation are performed, and a change evaluation order is determined; a cost rationality of the change evaluation order is proved by zero knowledge by an auditing party through a second data port, and the change evaluation order is displayed on a terminal interface of the engineering management system. That is, by constructing a change impact network based on an engineering knowledge base, introducing an attribution mechanism of explicit and implicit variables, using an engineering evaluator to realize change feature recognition and influence transmission analysis, and combining zero knowledge proof for auditing verification, the accuracy and comprehensiveness of engineering change cost evaluation are improved.

[0018] The above description is only a summary of the technical solutions of the present application. In order to enable the technical means of the present application to be more clearly understood, and to be implemented according to the content of the specification, and in order to enable the above and other purposes, characteristics and advantages of the present application to be more apparent and easy to understand, the following specific embodiments of the present application are described. It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only exemplary, and other drawings can be obtained by those skilled in the art without creating laborious work on the basis of the provided drawings.

[0020] Figure 1 The flowchart of the cost rapid evaluation method in the engineering change management of the present application.

[0021] Figure 2 The structural schematic diagram of the cost rapid evaluation system in the engineering change management of the present application.

[0022] Explanation of reference signs: knowledge base establishment module 11, evaluator construction module 12, engineering evaluation module 13, zero knowledge proof module 14. DETAILED DESCRIPTION

[0023] The application provides a cost rapid evaluation method and system in engineering change management, solves the technical problem that the change influence analysis is not comprehensive due to the long influence chain caused by engineering change in the prior art, thereby affecting the accuracy of cost evaluation. By constructing a change influence network based on an engineering knowledge base, introducing an attribution mechanism of explicit and implicit variables, using an engineering evaluator to realize change feature recognition and influence transmission analysis, and combining zero-knowledge proof for audit verification, the accuracy and comprehensiveness of engineering change cost evaluation are improved.

[0024] Hereinafter, the technical solutions in the application will be described clearly and completely with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the application, rather than all the embodiments of the application. It should be understood that the application is not limited by the example embodiments described herein. Based on the embodiments of the application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the application. In addition, it should be noted that, for the convenience of description, only parts related to the application are shown in the drawings, not all.

[0025] Embodiment one, please refer to the attached Figure 1 The application provides a cost rapid evaluation method in engineering change management, wherein the cost rapid evaluation method in engineering change management is executed by a cost rapid evaluation system in engineering change management, and the cost rapid evaluation method in engineering change management specifically includes the following steps:

[0026] Obtain the engineering type and establish the engineering knowledge base.

[0027] Specifically, the engineering type, i.e. the specific category or field to which the engineering project belongs, such as building engineering, civil engineering, etc., is obtained by user selection, project metadata extraction (such as key word recognition in project description), etc. According to the engineering type of the current project being processed, the engineering knowledge base, i.e. the structured knowledge system constructed for a specific engineering type, is established, which contains various information related to engineering cost evaluation, such as bill of materials, unit cost, process node, equipment model, construction time standard, energy consumption data, process flow, logical relationship between variables, etc. For example, it is assumed that the engineering type is identified as a large infrastructure construction project, and accordingly the corresponding engineering knowledge base is constructed, including typical materials (such as steel bars, concrete, etc.), unit cost, construction process, unit working hours, typical equipment, energy consumption data, etc. The structured data is stored in categories according to the engineering type, and the association between them is established. Each process has corresponding working hours, materials and energy consumption data. The entire knowledge base forms a large, interconnected data network.

[0028] According to the engineering knowledge base, define the change impact network, compile and train the feature-variable-change impact network, and build an engineering evaluator, wherein the engineering evaluator is embedded in an engineering management system, and the change impact dimensions include materials, man-hours, and energy consumption.

[0029] Further, the application further includes the following steps: for the material dimension, according to the knowledge part in the engineering knowledge base, define the change impact feature, wherein the change impact feature is determined at least according to the unit material cost and the utilization mode; define the center node as the material category, define the extended node as the material feature, define the edge as the cost relationship, and build a first change impact network; and add the first change impact network into the change impact network.

[0030] Specifically, for the material dimension, according to the knowledge part in the engineering knowledge base, define the change impact feature, that is, extract the key information of the change involving the material from the engineering knowledge base, including the material category to which it belongs, the unit cost, the utilization mode, and the utilization rate, etc. The change impact feature refers to the key attribute or index used to describe how the change specifically affects the dimension under a certain dimension. For the material dimension, at least the unit material cost and the utilization mode determine the impact variable at the material level due to the engineering change, such as the unit material cost, the usage amount, the utilization mode, the utilization rate, etc. The unit material cost is the procurement or usage cost of a single material unit, such as 580 yuan per ton of asphalt. The utilization mode is the usage way of the material in the project implementation, such as the utilization rate under a certain utilization mode as a kind of change impact dimension, for example, process optimization, that is, changing from one utilization mode to another, which may exist in the change of material type, usage amount, utilization mode, utilization rate, etc. In short, extract the key information of the change involving the material from the engineering knowledge base, including the material category to which it belongs, the unit cost, the utilization mode, and the utilization rate, etc. Compare the changes of the corresponding features before and after the change, such as the difference in unit cost, whether the utilization mode is converted (such as from centralized construction to segmented construction). Extract the change impact feature that will have a significant impact on the cost.

[0031] According to the material category definition center node, material feature definition extension node, cost relationship definition edge, construct the first change impact network. Materials are classified as center nodes, such as steel, concrete, asphalt, etc. The center node is used as the anchor point of the network, linking all the feature attributes related to the material. For each specific material item under the category, extract its related features as extension nodes. Material features are cost and usage characteristics that describe specific materials, such as unit cost, utilization rate, usage amount, procurement method, storage method, etc., which are used to analyze their specific impact on cost. For each pair of nodes, an edge is established based on the cost transmission logic. Each edge can have a weight (e.g., the impact weight of unit cost on total cost is 0.8, and the impact weight of transportation method is 0.2). According to the center node, extension node, and edge, a complete first change impact network is formed. The first change impact network refers to the first sub-network constructed in the material dimension, which is a component of the entire change impact network.

[0032] The change impact dimension is the multiple aspects affected in engineering changes, including at least material, labor, and energy consumption. Each dimension has a direct or indirect impact on cost. According to the change impact dimension, the corresponding change impact network is constructed. For example, data and rules related to labor consumption are extracted from the engineering knowledge base, and the change impact features of the labor dimension are defined, including job type, labor quota, work environment, skill level requirements, etc. The job type or work unit is defined as the center node, the labor feature is defined as the extension node, and the relationship between labor increase or decrease is defined as the edge to construct the second change impact network. For the energy consumption dimension, relevant data is also extracted from the knowledge base, and the change impact features of the energy consumption dimension are defined, such as device type, device power, running time, energy type, etc. The device type is defined as the center node, the energy consumption feature is defined as the extension node, and the relationship between energy consumption increase or decrease is defined as the edge to construct the third change impact network of the energy consumption dimension. The independent impact networks (first change impact network, second change impact network, and third change impact network) constructed for the material, labor, and energy consumption dimensions are integrated to form a complete change impact network. The integration process is not just a simple physical superposition, but also requires defining the association between nodes in different dimensions, i.e., the association between nodes is established through cross variables to obtain the change impact network. The change impact network refers to the integration or parallel connection of sub-networks in multiple dimensions such as material, labor, and energy consumption to form a unified impact network for comprehensive cost evaluation, including multiple basic networks constructed in each dimension.

[0033] By constructing independent impact networks for material, labor, and energy consumption dimensions respectively, and integrating them into a complete change impact network, each dimension's network captures the specific impact chain within that dimension, and the integrated complete network reveals the cross-impact and chain reaction between different dimensions.

[0034] Further, the application further comprises the following steps: defining an explicit variable set and an implicit variable set according to the engineering knowledge base, wherein each implicit variable is identified with a reasoning relationship; deploying a first data port with self-attention of change characteristics, deploying a second attribution node based on variable definition of the explicit variable set and the implicit variable set, and deploying a third conversion node based on conversion of the change influence network; cascading the first data port, the second attribution node, and the third conversion node to construct the engineering evaluator.

[0035] Further, the application further comprises the following steps: embedding the engineering knowledge base into the engineering evaluator to determine the evaluation logic based on the cascaded nodes; calling engineering records to integrate the engineering records based on the evaluation logic of the engineering change to determine sample data; and supervising training of the engineering evaluator to convergence according to the sample data.

[0036] Specifically, according to the engineering knowledge base, it is defined in advance which variables are explicit (can be directly obtained) and which are implicit (need to be inferred). Each implicit variable has a clear reasoning rule. The explicit variable set is a variable that can be directly extracted and identified from the engineering change data, such as material name, change quantity, unit cost, etc. It has a clear feature boundary. In addition to the variables that only need to be identified, there are also some paper tube features in the engineering knowledge base that need to be analyzed by reasoning. Classification is performed to define the corresponding logical rules to obtain the implicit variable set. The implicit variable set is a variable that cannot be directly identified and needs to be inferred from existing knowledge. It has an impact on cost but is not directly presented. For example, the efficiency decline percentage of a certain process caused by the change, the actual utilization rate of a certain new material in a specific construction environment, and the increase in certain indirect management costs caused by the change, etc. reflect the indirect impact or deep correlation of the change, which is the soft part of the knowledge system and needs to be analyzed. The implicit variable is identified with a reasoning relationship, which means that when defining the implicit variable, the specific method and logical rule for deriving the implicit variable from other explicit variables or known information has been clearly specified. For example, the efficiency decline percentage = baseline efficiency - (actual completion amount / baseline completion amount baseline efficiency).

[0037] Three key processing nodes are set and cascaded in sequence, the first data port is specially designed to receive the original features describing the engineering change, and the self-attention mechanism is used to identify which parts of these features are the most critical, so as to extract the core elements of the change. The second attribution node receives the features processed by the self-attention, and according to the pre-defined explicit and implicit variable set and its reasoning relationship, the features are accurately mapped to specific engineering variables, and it is distinguished which are directly available data and which need further calculation or inference. The third conversion node receives the variable set from the second attribution node, and uses the complete change impact network model built before, according to the nodes (such as specific items of materials, labor, energy consumption) and edges (such as cost relationship, impact transmission path) defined in the network, to locate these variables in the change impact network, and simulate the propagation process of change impact in the network, and perform necessary conversion and aggregation. In other words, input the change single text or structured data into the first data port, identify the keywords through the self-attention mechanism, automatically extract and encode the features related to the change; map the features to the second attribution node, the explicit variables are directly read, and the implicit variables are obtained by reasoning; map the identified variables to the third conversion node, and perform change impact network conversion and cost evaluation through the change impact network.

[0038] Cascade the first data port, the second attribution node and the third conversion node to build an engineering evaluator. Cascade refers to connecting multiple processing modules or nodes in a certain order, the output of the previous module is used as the input of the next module, forming a processing pipeline, that is, the three nodes are connected in sequence as input, logic and evaluation, to form the overall evaluation process.

[0039] Embed the engineering knowledge base into the engineering evaluator, that is, embed the structured knowledge (material information, construction process, cost model, etc.) in the engineering knowledge base into the engineering evaluator, so that it serves as the knowledge base for reasoning, judgment, variable mapping, etc. during the evaluation process. The engineering evaluator is no longer an empty shell, but has a specific application scenario and data support. Embedding the engineering knowledge base into the evaluator clarifies the basic data and rules on which the evaluator operates, and determines the logical flow of internal data processing. That is, the specific rules and order of data flow and processing between the first data port, the second attribution node and the third conversion node inside the engineering evaluator, that is, the algorithm design of the complete process from how the data is input from the change features, through variable definition and attribution, to network conversion and cost calculation, and finally output the evaluation result.

[0040] The call engineering record, i.e. the detailed record of the actual changes in the project type in the past engineering projects and the final resulting costs, impacts, contains the comparison of the state before and after the change and the actual results, which is the real data source of the completed project. The evaluation logic of engineering change is the set of rules and processes that the engineering evaluator has determined, which describes how to start from the initial characteristics of engineering change, identify variables, apply reasoning relationships, and use change impact networks for conversion calculation to finally obtain the cost evaluation result. In simple terms, it is the way the evaluator works internally for calculation and reasoning. According to the evaluation logic of engineering change, the engineering record is integrated, i.e. the engineering record is cleaned, classified, and supplemented with variables according to the structure required by the evaluation logic, so that the data has the conditions to train the model. For example, using the first data port, extract keywords and feature information from the change description, such as pipe material change, equipment upgrade, construction method adjustment, etc.; using the second attribution node, attribute the features to explicit variables (such as material number, quantity) or implicit variables (such as construction efficiency decline, repeated operation risk); using the third conversion node, map the variables to the impact network of material, labor, energy consumption, etc. Thus, sample data is obtained, each sample contains input features (change content), variable information (explicit + implicit), and corresponding costs (target value).

[0041] Supervised training of the engineering evaluator using sample data, by continuously adjusting the model parameters, so that the engineering evaluator can accurately predict the costs that may be generated by the change according to the input change description, until the prediction error of the engineering evaluator reaches a stable and acceptable level (i.e. convergence). That is, the process of training the model using sample data with correct answers (i.e. actual cost changes in sample data). The engineering evaluator makes predictions based on input features, then compares the predicted results with the correct answers, calculates the error, and adjusts the parameters inside the engineering evaluator according to the error in order to be more accurate next time.

[0042] The integrated sample dataset is divided into training set, validation set and test set. The training set is used for learning model parameters, the validation set is used for monitoring the training process, adjusting hyperparameters (such as learning rate, batch size, etc.), preventing overfitting, and the test set is used for final evaluation of the trained model performance. Assuming there are 1000 sample data, it is divided into 700 training set, 150 validation set and 150 test set. Assign initial values to the connection weights, bias and other parameters inside the engineering evaluator, usually use random initialization. Randomly select one or a small batch of samples (for example, 32 samples each time) from the training set. Input the input features (change features) of each sample into the first data port of the engineering evaluator. After the features are processed by the self-attention mechanism, they are passed to the second attribution node, which defines and reasons variables based on the explicit and implicit variable sets. Then, the data flows to the third conversion node, which performs conversion calculation based on the change impact network. Finally, the evaluator outputs a predicted cost change value. Compare the predicted cost change value output by the engineering evaluator with the actual cost change value (label) recorded in the sample data for that sample, calculate the difference between the two, i.e. loss, the commonly used loss function for regression problems (such as cost prediction) is mean square error. For example, the predicted value is 52,000 yuan, and the actual value is 50,000 yuan, the mean square error loss is (5.2-5.0)²=0.04. Calculate the gradient of the loss function with respect to all parameters inside the engineering evaluator (i.e. the degree of influence of parameter change on loss), usually use backpropagation algorithm to realize. Using optimization algorithms (such as Adam, SGD, etc.) to update the parameters inside the evaluator according to the calculated gradient and the preset learning rate. The goal is to reduce the loss function. Every certain number of iterations (or a round, i.e. once through the entire training set), evaluate the performance of the model on the validation set (calculate the loss and possible evaluation indicators such as R square value). If the performance on the validation set starts to decline or stagnates, it may mean that the model has started to overfit, and the training can be stopped early or the strategy can be adjusted. Repeat the process of forward propagation, loss calculation, backpropagation and parameter update until the loss on the validation set reaches a satisfactory low level and does not decrease significantly over multiple rounds, i.e. the model has converged. Use the test set to evaluate the generalization ability of the finally trained model, and determine the trained engineering evaluator. The input of the engineering evaluator includes change features, and the output includes predicted cost.

[0043] The logic compilation and training of the change feature-variable-change impact network refers to deploying a first data port according to the change feature, deploying a second attribution node according to the variable, and deploying a third conversion node according to the change impact network, cascading the first data port, the second attribution node and the third conversion node to form an engineering evaluator. The relationship between the change feature, the variable and the impact network is structured and integrated, so that it can be used as an input and inference logic execution framework of a machine learning model. According to the three-layer relationship structure, the model is trained, and the engineering evaluator is supervised learning using sample data, so that it can automatically evaluate the cost when receiving new change information.

[0044] The trained engineering evaluator is embedded in the engineering management system, and the logic compilation and training of the change feature-variable-change impact network are performed to build the engineering evaluator, wherein the engineering evaluator is embedded in the engineering management system, so as to become a component of the engineering management system, and is used for evaluating the uploaded engineering change order.

[0045] The engineering party uploads the engineering change order to the engineering management system through the first data port, triggers the engineering evaluator by word item detection, performs change feature self-attention import and coding on the input side, performs attribution based on implicit variables and explicit variables, converts into a change impact network group, maps the influence transmission of the network nodes to perform cost evaluation and comprehensive calculation, and determines the change evaluation order.

[0046] Further, the application further includes the following steps: uploading an engineering change order, triggering the engineering evaluator by performing word item detection, wherein the word item is defined based on change semantics; based on the first data port, performing self-attention recognition based on change features on the engineering change order, and importing and coding the engineering change features; based on the second attribution node, performing variable recognition measurement and attribution based on the engineering change features, and based on the third conversion node, performing change impact network conversion and cost evaluation to determine the change evaluation order.

[0047] Further, the application further includes the following steps: dividing a first transmission mode and a second transmission mode, wherein the first transmission mode is symmetric influence transmission, and the second transmission mode is asymmetric influence transmission; determining symmetry based on whether the influence weight before and after the change changes; mapping and transmitting the change impact network group based on the first transmission mode and the second transmission mode as constraints, and performing cost evaluation based on influence transmission.

[0048] In particular, the engineering party uploads the engineering change order to the engineering management system through the first data port. The engineering party is the main body of the construction of the engineering project, including the contractor, the design unit and other relevant parties participating in the engineering change activity. The first data port is a port deployed according to the self-attention of the change characteristics. The engineering change order is a file or document submitted by the engineering party to describe the content that needs to be changed in the project, such as material replacement, construction sequence adjustment, and extension of construction period. Using natural language processing technology, the engineering change order is semantically analyzed to identify keywords and core descriptions, and matched according to a pre-set vocabulary to trigger the subsequent change evaluation process. The change semantics is the meaning of the actual engineering change content described in the change order, which not only includes specific engineering terms, but also includes words describing the change action, as well as prefixes or suffixes that may imply the nature of the change. The text content of the engineering change order is scanned to match these predefined word entries and rules. For example, if the change content involves material replacement, the keywords steel and cement are identified as word entries and trigger the evaluation logic related to materials.

[0049] According to the word entries identified from the engineering change order, standardized information related to the word entries is automatically generated, and the engineering evaluator is automatically triggered. Through the first data port, the engineering change order is identified based on the self-attention of the change characteristics, and the key change items in the change description (such as material code, part, unit, etc.) are identified. That is, using the engineering change order information received through the first data port, the self-attention based on the change characteristics is identified, the relevance between different parts of the change order is focused on, and which parts are the core features of the change are identified. For example, the information of material replacement, equipment model change and construction stage adjustment in the change content is identified. The self-attention mechanism enables the model to automatically focus on key features and ignore irrelevant information (such as project background description, etc.). The identified engineering change characteristics are converted into coded form, usually into a digital vector form that the model can understand. For example, the change type is material model replacement, which is converted into coded form as 1000, and the new and old model comparison is converted from ABC to DEF, which is converted into coded form as [0.3, 0.7], indicating the performance difference between the two models.

[0050] The encoded engineering change feature is imported into the second attribution node, and the engineering change feature is finely variable-identified and attributed by the second attribution node to ensure that each change feature can be mapped to the correct variable and the variables are measured. Detailed analysis and quantification of the engineering change feature identifies the variable corresponding to each feature, and according to the explicit variable set and implicit variable set defined from the engineering knowledge base, it is analyzed which specific engineering variables the currently identified change feature may be associated with. Explicit variables can usually be directly obtained from the change description, and implicit variables need to be determined by reasoning. The second attribution node measures the identified variables (quantifies the degree of influence, such as calculating the added concrete volume and steel weight), and accurately attributes them to the corresponding engineering dimensions (materials, labor, energy consumption) and specific cost components. For example, material replacement features are attributed to material cost variables, and construction delays are attributed to labor and labor cost variables.

[0051] The variables attributed by the second attribution node are mapped to the third conversion node for change impact network conversion. That is, after processing by the second attribution node, variables such as material unit price, labor, and equipment usage are mapped to the corresponding change impact network. The change impact network includes multiple sub-networks (such as materials, labor, energy consumption, etc.), and there is a certain association relationship between these sub-networks, and the impact is transmitted from one sub-network to another. The variable enters the node of the network and transmits the impact through the edge.

[0052] The first transmission mode and the second transmission mode are divided, the first transmission mode is symmetric impact transmission, and the second transmission mode is asymmetric impact transmission. When the variables before and after the change have equal or small changes in the impact degree, it is called symmetric impact transmission. That is, if the impact weight of node A on node B is the same as the impact weight of node B on node A or the change trend is consistent, then this impact transmission relationship is called symmetric. In simple terms, the degree of mutual influence is equivalent. Asymmetric impact transmission is opposite to symmetric impact transmission, when the impact weight of the variables before and after the change on the cost changes greatly, that is, the impact brought by the change is not symmetric in the pre and post stages, it is called asymmetric impact transmission. For example, the impact of node A on node B is large, but the impact of B on A is small or almost none.

[0053] According to whether the influence weight value before and after the change changes, it is determined whether to perform symmetric influence transmission or asymmetric influence transmission. In the change influence network, the weight value of the edge represents the influence degree of a change feature on subsequent cost calculation. The mutual influence weight value between nodes in the change influence network is checked. If the influence degree before and after the change is small or basically consistent, the symmetric influence transmission mode is adopted. If the influence caused by the change has significant changes, the asymmetric influence transmission mode is adopted. For example, a certain device in the change is a first device before the change, and a second device needs to be used after the change. For the transition adjustment of the part of the change content, the first device has a first influence degree in the engineering stage, and the second device has a second influence degree in the engineering stage after the change. If the first influence degree and the second influence degree are inconsistent, it is asymmetric influence transmission, and the proportion of the corresponding change variable needs to be increased when subsequent comprehensive measurement is performed.

[0054] With the first and second transmission modes as constraints, the mapping and transmission of the identification of the entire change impact network group are performed, and each change variable is mapped from one sub-network (such as a material network or a working hour network) to other sub-networks in the change impact network group according to the impact degree of the change. According to the transmission mode (symmetric or asymmetric), the cost change in each network is calculated according to the weight change before and after the change. In the symmetric impact mode, the impact degree is transmitted to the target node, and the original weight proportion is maintained. In the asymmetric impact mode, the weight transmission is adjusted according to the change amplitude, and the nodes with a larger impact degree obtain more cost changes. The final change cost is calculated according to the nodes and edges of each sub-network. Through the calculation of the impact transmission, a change evaluation sheet is generated, which lists the impact of the change on each dimension (material, working hour, energy consumption, etc.) and the overall cost. The change evaluation sheet is the final output based on the change impact network transmission and contains a detailed report of the specific impact (such as cost, time, etc.) of the change on the project. For example, the change characteristics (concrete grade change, addition of cast-in-place pile) are identified, and the relevant variables are determined through the second attribution node: first concrete consumption, second concrete reduction, cast-in-place pile material consumption, cast-in-place pile construction working hour, and related equipment (such as a drilling machine) usage time, etc. The impact degree of the original second concrete on the cost before the change (assuming 0.1, as it is a standard material), and the impact degree of the first concrete on the cost after the change (assuming 0.15, as the grade is increased, and the unit price and consumption may increase slightly). These two impact degrees are inconsistent (0.1 is not equal to 0.15), and therefore belong to asymmetric impact. Similarly, the newly added cast-in-place pile has an impact degree of 0 before the change and a significantly increased impact degree (assuming 0.2) after the change, which is also a typical asymmetric impact. For nodes with asymmetric impact, their weights are increased when the total cost is calculated. Assuming that the preliminary calculation of the cost change of each part is: the first concrete increases the cost by 5000 yuan, the second concrete reduces the cost by 3000 yuan, the cast-in-place pile increases the cost by 80000 yuan, and the related working hour and equipment increase the cost by 15000 yuan. According to the asymmetric adjustment of the weight, a higher weight coefficient (assuming 1.2 and 1.1 respectively) is given to the cast-in-place pile and the first concrete, and a lower coefficient (assuming 0.9) is given to the second concrete reduction part, and the calculation is re-calculated as 103800 yuan. A change evaluation sheet is output, which clearly indicates that the change is expected to result in an increase of about 103800 yuan in the total cost, and details the cost change of each sub-item (material, working hour, equipment, etc.).

[0055] By distinguishing between symmetric and asymmetric influence transmission and adjusting the weight in the comprehensive calculation accordingly, the cost assessment can better reflect the actual complex and nonlinear chain reaction of changes in the engineering system, especially when dealing with elements with significant changes in influence degree, making the assessment results more accurate and reliable. This allows the engineering assessor to better adapt to different types and complexities of engineering changes, rather than being a one-size-fits-all calculator. It improves the adaptability and robustness of the model by adjusting the calculation strategy based on specific circumstances.

[0056] Further, the application further comprises the following steps: the change influence network comprises a first network and a second network, wherein the first network is the network matrix before the change, and the second network is the network matrix after the change; mapping the first network and the second network, performing transmission mode judgment and transmission influence coefficient identification for each mapping group, performing cost assessment and comprehensive analysis under the constraint of transmission influence coefficient for each mapping group, and generating the change assessment form.

[0057] Specifically, the change influence network represents the relationship and influence path between various variables (such as materials, labor hours, energy consumption, etc.) during the engineering change process. The nodes of the network represent variables, and the edges represent the transmission path of influence. The change influence network comprises a first network and a second network. The first network is the network structure before the change, which records all variables and their mutual influence relationship before the change, serving as a benchmark network. The second network is the network structure after the change, which records the changes in variables and their influence relationship after the change, reflecting the changes brought about by the change. Mapping the first network and the second network, the same nodes in the pre-change network and the post-change network are mapped one by one, and the relationship between them is connected.

[0058] According to the influence degree of the mapped pre-change and post-change networks, it is determined whether to use symmetric transmission mode or asymmetric transmission mode. When the pre-change and post-change influences change little, the transmission influence path is relatively stable, and the influence degree changes little, symmetric influence transmission is selected. When the pre-change and post-change influences differ greatly, the transmission influence path changes significantly, and the influence degree fluctuates greatly with the degree of change, asymmetric influence transmission is selected. The degree of change is calculated, and an influence coefficient is given. Each node's influence path is marked with a transmission influence coefficient to quantify the degree of change in the transmission process. The transmission influence coefficient represents the influence intensity of a change item on subsequent nodes. For example, the material cost change coefficient is 1.0 (main change), and the labor cost change coefficient is 0.8 (secondary change, because part of the labor is reduced and part is increased).

[0059] According to the transmission influence coefficient and the mapping relationship in the network, the specific influence of the change item on each cost is calculated, including material cost, man-hour cost, energy consumption cost, etc. For example, the difference between the unit price of the new material and the unit price of the original material is calculated, and then the total material cost change is calculated; according to the workload change brought by the change, the increase or decrease of labor cost is calculated; according to the change of equipment power and working time, the influence of energy consumption on project cost is calculated.

[0060] After the influence transmission and cost analysis, a change evaluation sheet is generated, which lists the specific influence of each change item on project materials, cost, time, etc. It not only contains the total cost change, but also contains the cost contribution analysis of each mapping group (i.e. each main influence aspect). For example, a change sheet describes the replacement of the insulation material of a certain pipe from rock wool (first network node) to rubber plastic (second network node), and determines that the change of material leads to the change of working hours, which is a non-symmetrical influence transmission (because the cutting and installation processes are different, the working hour influence weight changes). Calculate the transmission influence coefficient, the material cost change coefficient is 1.0 (main change), and the working hour cost change coefficient is 0.8 (secondary change, because part of the working hours is reduced and part is increased). Material part 500 yuan (remove) 1.0+800 yuan (add) 1.0=300 yuan; Working hours part: increase working hours 50 people / hour 250 yuan / hour=12500 yuan, comprehensive cost influence=300 yuan+12500 yuan=12800 yuan. By comparing the two network states before and after the change, and analyzing the mapping relationship and influence transmission between nodes, the direct and indirect influence brought by the change is fully captured, especially the complex influence across different engineering dimensions.

[0061] The audit party performs zero-knowledge proof on the cost reasonableness of the change evaluation sheet according to the second data port, and displays it on the terminal interface of the engineering management system.

[0062] Further, the application further includes the following steps: introducing a reference string, wherein the reference string is a consensus setting of the engineering party and the audit party; setting a trusted proof based on the engineering change sheet according to the reference string; writing the trusted proof into the second data port to verify the change evaluation sheet.

[0063] Further, the application further includes the following steps: generating engineering change data with the engineering change sheet and the change evaluation sheet; replacing the engineering planning with the engineering change data, and displaying it on the terminal interface of the engineering management system.

[0064] Specifically, the reference string is introduced, which is agreed upon by the engineering party and the audit party, contains the key information of the change content and its evaluation conditions, and is used to ensure the accuracy, integrity and consistency of the data as the basis for trusted proof. That is, in the change evaluation process, the engineering party and the audit party jointly agree on a set of reference strings, which contains the key content of the change order, such as change items, pre-change and post-change data, and preset cost limit conditions. The reference string is used to verify the cost evaluation as a benchmark to ensure that the evaluation data is within a reasonable range and meets the agreement of both parties. The reference string includes the change type, the cost data before the change, and the cost limit condition after the change (such as a maximum allowed cost increase of 10%).

[0065] According to the reference string, combined with the detailed data in the change order (such as material unit price, man-hour, etc.), a trusted proof is generated, which exists in the form of a string, verifying whether each cost calculation in the change evaluation order is reasonable. The trusted proof is a data verification mechanism based on the reference string, which is used to verify the cost data in the change evaluation order, and its role is to ensure that the evaluation data after the change is not tampered with or has not too much error, while protecting data privacy. For example, assuming that the reference string shows that the raw material cost is 5000 yuan, and the material cost after the change should not exceed 5500 yuan (i.e. an increase of 10%), the trusted proof will ensure that the material cost in the change evaluation order does not exceed 5500 yuan.

[0066] The trusted proof is written into the second data port, and the proof is received through the second data port and the verification process begins. All data in the change evaluation order is checked against the preset conditions (such as cost limit conditions) defined in the reference string to ensure that the evaluation data is within a reasonable range. If the data in the evaluation order exceeds the preset conditions, a warning will be triggered or corrected to ensure the fairness and transparency of the change evaluation. Through the second data port, the change evaluation order is verified to check whether all cost items meet the standards in the trusted proof. After the verification is passed, the change evaluation order will be marked as verified and returned to the engineering party and the audit party for further decision-making. If the verification fails, an error prompt is given, requiring the relevant parties to recheck the change data. For example, assuming that the reference string in the construction engineering includes: the change type is material replacement (such as replacing the first type of cement with the second type), the cost before the change is 400 yuan / cubic meter for the first type of cement, the cost after the change is 500 yuan / cubic meter for the second type of cement, and the cost limit is not more than a 10% price increase. The cost of concrete before the change is 400 yuan / cubic meter x 50 cubic meters = 20000 yuan, and the cost of concrete after the change is 500 yuan / cubic meter x 50 cubic meters = 25000 yuan, the price increase is (25000-20000) / 20000=25% (exceeding the cost limit, triggering a warning).

[0067] Zero-knowledge proof is a cryptographic technique that allows one party (prover) to prove to another party (verifier) that a certain statement is true without revealing any other information that could help prove the statement, that is, zero-knowledge proof is used to prove that the cost assessment result of the change assessment sheet is reasonable, without revealing the specific cost detail data. The core of zero-knowledge proof is that the auditor does not need to obtain the specific cost data of the change assessment sheet, but verifies whether the data meets the pre-set rationality conditions (such as cost increase, budget range, etc.) through encryption algorithm. For example, the auditor only needs to verify whether the cost increase of a certain change is within 10%, without seeing the actual cost number. After zero-knowledge proof, an engineering change data is generated, including change content (such as material replacement, time extension, etc.), cost, time, resources before and after the change, verification results (whether it meets the rationality condition) of each change, etc.

[0068] Based on the verified engineering change data, the existing engineering plan is updated and replaced. For example, according to the cost assessment after the change, the budget of the project is updated; if the change leads to the extension of the construction period, the schedule of the project is updated; if the change affects the resource allocation (such as materials, labor, etc.), the resource plan is automatically adjusted. The updated engineering plan, cost, time, etc. information will be displayed on the terminal interface of the engineering management system, for project managers, auditors or other relevant personnel to view. The terminal interface will display the updated data after the change, visual planning diagram, budget details, etc.

[0069] Through zero-knowledge proof, the auditor can verify the rationality of the change assessment sheet without viewing the actual sensitive data, which not only ensures the credibility of the data, but also protects the privacy. Zero-knowledge proof ensures that the data of the change assessment sheet meets the pre-set rationality conditions, such as cost increase limit, avoiding manual intervention and data bias, and improving the fairness of the assessment. The automatically generated change data and the updated engineering plan ensure the transparency of project decision-making, and project managers, auditors, etc. can clearly see the specific impact of the change on the project. Through the display of the terminal interface, the change information of the project is updated in real time, helping the project team to make decisions quickly.

[0070] In summary, the cost rapid assessment method in the engineering change management provided by the present application has the following beneficial effects:

[0071] By acquiring the engineering type, an engineering knowledge base is established; according to the engineering knowledge base, a change impact network is defined, and a change feature-variable-change impact network is logically compiled and trained to construct an engineering evaluator, wherein the engineering evaluator is embedded in an engineering management system, and the change impact dimensions at least include materials, man-hours, and energy consumption; an engineering party uploads an engineering change order to the engineering management system through a first data port, so that the engineering evaluator is triggered by a term test, change feature self-attention import and coding are performed on the input side, attribution based on implicit variables and explicit variables is executed, a change impact network group is converted to map the influence transmission of network nodes to perform cost evaluation and comprehensive calculation, and a change evaluation order is determined; an auditing party performs zero-knowledge proof on the cost reasonableness of the change evaluation order according to a second data port, and the change evaluation order is displayed on a terminal interface of the engineering management system. That is, by constructing a change impact network based on an engineering knowledge base, introducing an attribution mechanism of explicit and implicit variables, using an engineering evaluator to realize change feature recognition and influence transmission analysis, and combining zero-knowledge proof for auditing verification, the accuracy and comprehensiveness of engineering change cost evaluation are improved.

[0072] In the second embodiment, based on the same inventive concept as the cost rapid evaluation method in the engineering change management in the first embodiment, the present application also provides a cost rapid evaluation system in engineering change management. Please refer to the attached Figure 2 The cost rapid evaluation system in engineering change management comprises:

[0073] A knowledge base establishment module 11 is configured to acquire an engineering type and establish an engineering knowledge base; an evaluator construction module 12 is configured to define a change impact network according to the engineering knowledge base, logically compile and train a change feature-variable-change impact network, and construct an engineering evaluator, wherein the engineering evaluator is embedded in an engineering management system, and the change impact dimensions at least include materials, man-hours, and energy consumption; an engineering evaluation module 13 is configured to trigger the engineering evaluator by a term test to perform change feature self-attention import and coding on the input side, execute attribution based on implicit variables and explicit variables, convert to a change impact network group, map the influence transmission of network nodes to perform cost evaluation and comprehensive calculation, and determine a change evaluation order; and a zero-knowledge proof module 14 is configured to perform zero-knowledge proof on the cost reasonableness of the change evaluation order by an auditing party according to a second data port, and display the change evaluation order on a terminal interface of the engineering management system.

[0074] Further, the evaluator construction module 12 in the cost rapid evaluation system in the engineering change management is further used to: for the material dimension, define the change impact feature according to the knowledge part in the engineering knowledge base, wherein the change impact feature is determined at least according to the unit material cost and the utilization mode; define the center node with the material category, define the extended node with the material feature, define the edge with the cost relationship, and construct the first change impact network; and add the first change impact network into the change impact network.

[0075] Further, the evaluator construction module 12 in the cost rapid evaluation system in the engineering change management is further used to: define the explicit variable set and the implicit variable set according to the engineering knowledge base, wherein each implicit variable is identified with a reasoning relationship; deploy the first data port with the self-attention recognition of the change feature, deploy the second attribution node with the variable definition based on the explicit variable set and the implicit variable set, and deploy the third conversion node based on the conversion of the change impact network; cascade the first data port, the second attribution node, and the third conversion node to construct the engineering evaluator.

[0076] Further, the evaluator construction module 12 in the cost rapid evaluation system in the engineering change management is further used to: embed the engineering knowledge base into the engineering evaluator to determine the evaluation logic based on the cascaded nodes; call the engineering record to integrate the engineering record based on the evaluation logic of the engineering change to determine the sample data; and supervise and train the engineering evaluator to convergence according to the sample data.

[0077] Further, the engineering evaluation module 13 in the cost rapid evaluation system in the engineering change management is further used to: upload the engineering change order to trigger the engineering evaluator by performing the term detection, wherein the term is defined with the change semantics; perform the self-attention recognition based on the change feature on the engineering change order according to the first data port to encode and import the engineering change feature; perform the variable recognition measurement and attribution based on the engineering change feature based on the second attribution node, perform the change impact network conversion and cost evaluation based on the third conversion node, and determine the change evaluation order.

[0078] Further, the engineering evaluation module 13 in the cost rapid evaluation system in the engineering change management is further used to: divide the first transmission mode and the second transmission mode, wherein the first transmission mode is the symmetric impact transmission, and the second transmission mode is the asymmetric impact transmission, and the symmetry is determined according to whether the impact weight changes before and after the change; map and transmit the change impact network group according to the constraint of the first transmission mode and the second transmission mode to perform the cost evaluation based on the impact transmission.

[0079] Further, the engineering evaluation module 13 in the cost rapid evaluation system in the engineering change management is further used for: the change influence network contains a first network and a second network, wherein the first network is a network matrix before the change, and the second network is a network matrix after the change; mapping the first network and the second network, performing transfer mode determination and transfer influence coefficient identification for each mapping group, performing cost evaluation and comprehensive analysis under the transfer influence coefficient constraint for each mapping group, and generating the change evaluation form.

[0080] Further, the zero-knowledge proof module 14 in the cost rapid evaluation system in the engineering change management is further used for: introducing a reference string, wherein the reference string is a consensus setting of an engineering party and an auditing party; setting a trusted proof based on the engineering change form according to the reference string; writing the trusted proof into a second data port, and verifying the change evaluation form.

[0081] Further, the zero-knowledge proof module 14 in the cost rapid evaluation system in the engineering change management is further used for: generating engineering change data by using the engineering change form and the change evaluation form; and replacing engineering planning by using the engineering change data, and displaying the engineering planning on a terminal interface of the engineering management system.

[0082] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The foregoing Figure 1 The cost rapid evaluation method and specific examples in the engineering change management in the first embodiment are also applicable to the cost rapid evaluation system in the engineering change management in the embodiment. The cost rapid evaluation system in the engineering change management in the embodiment can be clearly understood by the person skilled in the art through the foregoing detailed description of the cost rapid evaluation method in the engineering change management. Therefore, for the sake of brevity of the specification, the cost rapid evaluation system in the engineering change management in the embodiment will not be described in detail.

[0083] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0084] Obviously, for those skilled in the art, some improvements and modifications can be made to the present application without departing from the principles of the present application, and these improvements and modifications also fall within the protection scope of the present application.

Claims

1. A method for rapid cost evaluation in engineering change management, characterized by, The method comprises the following steps: acquiring an engineering type and establishing an engineering knowledge base; defining a change impact network based on the engineering knowledge base, performing logical compilation and training on a change feature-variable-change impact network, and constructing an engineering evaluator, wherein the engineering evaluator is embedded in an engineering management system, and the change impact dimension at least includes materials, man-hours, and energy consumption; uploading an engineering change order to the engineering management system through a first data port by an engineering party, triggering the engineering evaluator through term indexing, performing change feature self-attention import and coding on the input side, performing attribution based on implicit variables and explicit variables, converting into a change impact network group, mapping network node impact transmission for cost evaluation and comprehensive calculation, and determining a change evaluation order; an audit party performs zero-knowledge proof on the cost reasonableness of the change evaluation order based on a second data port, and displays the change evaluation order on the terminal interface of the engineering management system; the first data port is specially used to receive original features describing engineering changes, and uses a self-attention mechanism to identify which parts of the features are most critical, thereby extracting the core elements of the changes; the second attribution node receives the features processed by the self-attention, accurately maps the features to specific engineering variables based on a predefined set of explicit and implicit variables and their reasoning relationships; the identified variables are mapped to the third conversion node, and the change impact network conversion and cost evaluation are performed through the change impact network.

2. The method for rapid cost evaluation in engineering change management according to claim 1, wherein, Defining a change impact network based on the engineering knowledge base comprises: for the material dimension, defining a change impact feature based on the knowledge part in the engineering knowledge base, wherein the change impact feature is determined based on at least unit material cost and utilization mode; defining a center node based on the material category, defining an extended node based on the material feature, and defining an edge based on the cost relationship to construct a first change impact network; adding the first change impact network to the change impact network.

3. The method for rapid cost evaluation in engineering change management according to claim 1, wherein, Performing logical compilation and training on a change feature-variable-change impact network to construct an engineering evaluator comprises: defining a set of explicit variables and a set of implicit variables based on the engineering knowledge base, wherein each implicit variable is identified with a reasoning relationship; deploying a first data port based on self-attention identification of change features, deploying a second attribution node based on variable definition of the set of explicit variables and the set of implicit variables, and deploying a third conversion node based on conversion of the change impact network; cascading the first data port, the second attribution node, and the third conversion node to construct the engineering evaluator.

4. The method for rapid cost evaluation in engineering change management according to claim 3, wherein, Constructing the engineering evaluator comprises: embedding the engineering knowledge base in the engineering evaluator to determine the evaluation logic based on the cascaded nodes; calling engineering records to integrate the engineering records based on the evaluation logic of engineering changes to determine sample data; supervised training the engineering evaluator based on the sample data until convergence.

5. The method for rapid cost evaluation in engineering change management according to claim 3, wherein, Determining a change evaluation order comprises: uploading an engineering change order to trigger the engineering evaluator through term indexing, wherein the terms are defined based on change semantics; performing self-attention identification of the engineering change order based on change features through the first data port, and performing coding import of engineering change features; Based on the second home node, a variable identification measure based on the engineering change feature is performed, and a change impact network conversion and cost evaluation based on the third conversion node is performed to determine the change evaluation sheet.

6. The method for rapid cost evaluation in engineering change management according to claim 5, wherein, The change impact network conversion and cost evaluation based on the third conversion node comprises: Divide the first transmission mode and the second transmission mode, wherein the first transmission mode is a symmetric influence transmission, and the second transmission mode is an asymmetric influence transmission, and the symmetric judgment is performed according to whether the influence weight before and after the change is changed; Map and transmit the identification of the change impact network group with the first transmission mode and the second transmission mode as constraints, and perform the cost evaluation based on the influence transmission.

7. The method for rapid cost assessment in engineering change management according to claim 6, wherein, The change impact network comprises a first network and a second network, wherein the first network is a network matrix before the change, and the second network is a network matrix after the change; Map the first network and the second network, perform transmission mode judgment and transmission influence coefficient identification for each mapping group, perform cost evaluation and comprehensive analysis under the constraint of transmission influence coefficient for each mapping group, and generate the change evaluation sheet.

8. The method for rapid cost assessment in engineering change management of claim 1, wherein, According to the second data port, the cost rationality of the change evaluation sheet is proved by zero knowledge, comprising: Introducing a reference string, wherein the reference string is a consensus setting of the engineering party and the audit party; According to the reference string, a trusted proof based on the engineering change sheet is set; The trusted proof is written into the second data port, and the change evaluation sheet is verified.

9. The method for rapid cost assessment in engineering change management of claim 1, wherein, Display on the terminal interface of the engineering management system, comprising: Generate engineering change data with the engineering change sheet and the change evaluation sheet; Replace the engineering plan with the engineering change data, and display on the terminal interface of the engineering management system.

10. A system for rapid cost evaluation in engineering change management, characterized by, The steps of the cost rapid evaluation method in the engineering change management of any one of claims 1 to 9, the cost rapid evaluation system in the engineering change management comprises: A knowledge base establishment module for obtaining an engineering type and establishing an engineering knowledge base; An evaluator construction module for defining a change impact network based on the engineering knowledge base, logically compiling and training a change feature-variable-change impact network, and constructing an engineering evaluator, wherein the engineering evaluator is embedded in an engineering management system, and the change impact dimension at least includes materials, working hours, and energy consumption; An engineering evaluation module for an engineering party to upload an engineering change sheet to an engineering management system through a first data port, trigger the engineering evaluator with a word entry, import and encode the change feature from the attention on the input side, perform attribution based on implicit variables and explicit variables, convert to a change impact network group, map the influence transmission of the network nodes to perform cost evaluation and comprehensive calculation, and determine a change evaluation sheet; The first data port specially receives original features describing engineering changes, and uses a self-attention mechanism to identify which parts of the features are most critical, thereby refining the core elements of the change; The second attribution node receives the features processed by the self-attention, and accurately maps the features to specific engineering variables according to a predefined set of explicit and implicit variables and their reasoning relationships; The identified variable is mapped to a third conversion node, and the change impact network conversion and cost evaluation are performed through change impact network modification; The zero-knowledge proof module is configured to perform zero-knowledge proof on the cost rationality of the change evaluation module by the second data port, and display the zero-knowledge proof on a terminal interface of the engineering management system.

Images