KR101322504B1 - System and method for evaluating economic and environmental impact assessment through estimating construction cost and co2 emission - Google Patents

Abstract

PURPOSE: A system for predicting construction costs and carbon dioxide emission amount in a construction step and a method thereof are provided to estimate the construction costs and the carbon dioxide emission amount in an initial construction step, thereby reducing time and effort required for a construction project. CONSTITUTION: A data server (110) stores construction project data, construction costs, and carbon dioxide emission amount data in a construction step. A data collection module (120) collects construction project data in the data server. A material construction amount prediction module (130) predicts material construction amount based on the construction project data. A construction costs and carbon dioxide emission amount calculation module (140) calculates carbon dioxide emission amount and construction costs based on the material construction amount and the construction project data. [Reference numerals] (100) System for predicting the costs of a construction work and CO_2 emissions at the stage of construction; (110) Data server; (120) Data collection module; (130) Material construction amount prediction module; (131) Similarity calculating unit; (132) Predicting value filtering unit; (133) Parameter optimizing unit; (140) Construction costs and carbon dioxide emission amount calculation module

Description

SYSTEM AND METHOD FOR EVALUATING ECONOMIC AND ENVIRONMENTAL IMPACT ASSESSMENT THROUGH ESTIMATING CONSTRUCTION COST AND CO2 EMISSION}

The present invention relates to a system and method for estimating construction costs and carbon dioxide (CO ₂ ) emissions at an early stage of a construction project. In particular, the present invention relates to a system and method for estimating construction cost and carbon dioxide emission in the construction stage by improving the accuracy of prediction by applying a combination of artificial neural network model, multiple regression analysis model, genetic algorithm based on a case-based reasoning model.

Due to the rapid growth and industrialization of cities all over the world, environmental problems such as global warming and resource depletion are occurring. As environmental issues arise, various efforts have been made in each country to reduce greenhouse gases since the 1997 Kyoto Climate Change Convention. It has declared he would either reduce CO ₂ emissions by 30% compared to the bar; if the country of greenhouse gas emissions projections (BAU Business-As-Usual) by 2020. We are working in various industries to reduce CO ₂ emissions. In particular, since the construction industry accounts for more than 30% of the CO ₂ emissions generated by all industries, efforts are being made to reduce the CO ₂ emissions of buildings through eco-friendly construction technology and eco-friendly building certification system.

Under these backgrounds, much attention and effort has been focused on the development of element technology for evaluating CO ₂ emissions from materials and equipment input at the construction stage of construction and predicting construction costs. Conventional technology for this has introduced a process for estimating the construction cost and CO ₂ emissions of the construction stage at or after the design drawings are drawn in detail. The construction cost and CO ₂ emission estimation technology in the construction stage used in the early stage of the construction project is based on the project characteristic-based classification system such as structural form, planar form, type, construction method, and strength. Project cost and CO ₂ emissions are estimated based on the project characteristic-based classification system.

Techniques for estimating construction costs and CO ₂ emissions generally apply various methodologies, such as case-based reasoning methodologies, statistical methodologies (eg multiple regression analysis), and machine learning methodologies (eg artificial neural networks).

Korean Patent Application Publication No. 10-2012-0024309, "A method for evaluating the economics of eco-friendly building technology," (a) provides for the construction, maintenance and demolition costs incurred during the life cycle of the target building. steps to calculate, (b) the step of estimating the environmental costs, including CO ₂ costs associated with the construction phase, operational phase and demolition stages of the target architecture against the green building technologies, (c) the step (a) and (b) Analyzing the whole life cycle cost of the target building for the green building technology based on the costs calculated in the step and (d) based on the analysis result of step (c) Evaluate against existing building technology corresponding to the technology.

However, such techniques remain at the level of simple prediction techniques and have a limit of prediction accuracy. In addition, to estimate the construction cost and CO ₂ emissions of the construction stage in the early stage of the construction project, to evaluate the adequacy of construction materials and construction methods, and to establish target values in terms of economic and environmental aspects of the project. There is a need to improve the prediction accuracy of systems and methods.

Economic and environmental evaluation system and method through estimating construction cost and CO ₂ emissions according to the present invention aims to solve the following problems.

First, the case-based model is basically used in the construction stage, which is the initial stage of the construction project, and the artificial neural network model, multiple regression analysis model, and genetic algorithm are combined to improve the accuracy of the construction cost and CO ₂ emission forecast.

Second, we estimate construction costs and CO ₂ emissions based on the information available in the early stages of construction projects, and improve work efficiency by reducing the time and effort required for construction projects. In addition, the results of the forecast will be used to assess the adequacy of construction materials and construction methods, and to establish target values in terms of economic and environmental aspects of the construction project.

Third, we will develop a model that can compare and evaluate various design alternatives by changing the characteristics of construction projects and select the optimal design alternatives.

The solution of the present invention is not limited to those mentioned above, and other solutions not mentioned can be clearly understood by those skilled in the art from the following description.

The construction cost and carbon dioxide emission prediction system in the construction phase according to the present invention is a data server, the construction server data and the construction data and CO ₂ emission data are stored in the construction stage, the construction project data collected from the data server And a collecting module unit.

In addition, based on the construction project data collected by the data collection module unit, the construction quantity prediction module by material, and the construction cost and material production, material transportation or including but pourable construction costs to calculate the CO ₂ emissions CO of one or more of the ₂ emissions from the construction site and the CO ₂ emission calculation module, material-specific construction volume predictions case-based reasoning model with respect to the material-specific construction volumes predicted data from the module unit, artificial The neural network model, multiple regression analysis model and genetic algorithm is characterized by the complex application.

In addition, the construction cost and CO2 emission prediction method in the construction stage according to the present invention is S10 that the construction project data and the test case data and search case data for the construction cost and CO ₂ emissions in the construction stage is input to the data server Step, step S20 in which case similarity calculation is applied to the test case data and search case data input from step S10 by the case similarity calculation unit of the construction quantity prediction module unit for each material, and the predicted value of the construction quantity prediction module unit by the material. In the filtering unit, in the step S30 of the case of the case similarity calculated in the step S20 is selected in the case selection range (RCS), in the predicted value filtering unit of the construction quantity prediction module for each material, step S20 for the case selected in step S30 Estimated Construction Volume by Material Predicted by Case Similarity And a step S40 which is filtered using the artificial neural network model and multiple regression model.

Further, in the parameter optimization unit of the construction quantity prediction module for each material , in step S50, the average construction accuracy of the construction quantity prediction value filtered for the material filtered in step S40 is calculated for the case selected in step S30, construction quantity prediction module for each material When the average prediction accuracy calculated in step S50 is less than the reference accuracy, the parameter optimization unit includes a step S60 in which dependent parameters affecting the construction quantity prediction value for each material are optimized through genetic algorithm.

In addition, the construction cost and CO ₂ emission calculation module Bouygues Construction project data and materials on the basis of specific construction volumes construction costs and materials CO ₂ emissions of one or more of the CO ₂ emissions of materials during transportation or construction site during production calculated in S40 step Comprising the step S70, characterized in that the step S50 is repeated from step S20 using an optimized dependent parameter.

The construction cost and CO ₂ emission prediction system and method at the construction stage by estimating construction cost and carbon dioxide emission have the following effects.

First, since construction cost and CO ₂ emissions are estimated based on the information available in the early stages of construction projects, work efficiency can be improved by reducing the time and effort required for construction projects.

Second, in the estimation of construction cost and CO ₂ emissions at the construction stage, the existing cases can be presented as evidence, so the basis for decision makers' clarification can be clarified, and the management index at construction stage can be clarified. Therefore, it is possible to set clear goals and actively encourage efforts to reduce greenhouse gases.

Third, in estimating the construction cost and CO ₂ emissions in the construction stage, the forecasting accuracy was greatly improved because the combination of the advantages of various methodologies was applied.

The effects of the present invention are not limited to those mentioned above, and other effects not mentioned can be clearly understood by those skilled in the art from the following description.

1 is a schematic block diagram showing the configuration of the construction cost and CO ₂ emission prediction system in the construction stage according to the present invention.
Figure 2 shows an example of the filtering process performed in step S40 according to the prediction value filtering unit of the construction quantity prediction module unit for each material according to the present invention or the method of the present invention.
Figure 3 is a flow chart schematically showing the order of the construction cost and CO2 emission prediction method in the construction stage according to the present invention.
Figure 4 is a flow chart mainly showing the optimization process using a case-based reasoning model and genetic algorithm in the construction cost and carbon dioxide emission prediction method in the construction stage according to the present invention.
5 and 6 are graphs showing the construction cost and CO ₂ emissions estimation results in the construction stage in the bar phase according to the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.

The terms first, second, A, B, etc., may be used to describe various components, but the components are not limited by the terms, but may be used to distinguish one component from another . For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component. And / or < / RTI > includes any combination of a plurality of related listed items or any of a plurality of related listed items.

As used herein, the singular " include "should be understood to include a plurality of representations unless the context clearly dictates otherwise, and the terms" comprises & , Parts or combinations thereof, and does not preclude the presence or addition of one or more other features, integers, steps, components, components, or combinations thereof.

In addition, in the present invention, the material refers to various materials such as soil, stone, water, cement, rebar, wood, formwork, concrete, etc. used in construction structures or building structures. It is desirable to predict. In the present invention, the main material can be determined as the main material reinforcement, formwork, concrete, and thus, the present invention will be described mainly with respect to the main material.

The present invention is not limited thereto, and one or more of the above materials can be determined as the main material. In the present invention, carbon dioxide and CO ₂ should be interpreted as having the same meaning.

Hereinafter, with reference to the drawings will be described in detail with respect to the construction cost and CO ₂ emissions prediction system 100 and method in the construction stage.

Before describing the drawings in detail, it is to be clarified that the division of constituent parts in this specification is merely a division by main functions of each constituent part. That is, two or more constituent parts to be described below may be combined into one constituent part, or one constituent part may be divided into two or more functions according to functions that are more subdivided. Each of the components to be described below may additionally perform some or all of the functions of other components in addition to the main functions of the components, and some of the main functions of each of the components are different. Of course, it may be carried out exclusively by.

Therefore, the existence of each component described through this specification should be functionally interpreted, and for this reason, the construction of the components according to the construction cost and CO ₂ emission prediction system 100 in the construction stage of the present invention is It should be clearly understood that the present invention may be different from FIG. 1 to the extent that the object of the present invention can be achieved.

We will briefly explain case-based reasoning. Case-based reasoning is a technique for solving new problems based on past resolved cases, based on the assumption that similar problems have similar solutions and that problems that occur once occur frequently. It is easy to understand the result obtained by using the same method as human problem solving method, and it has the advantage of learning without any additional work besides storing the new case in the database. Case-based reasoning derives solutions through the Retrieve, Reuse, Revise, and Retain processes, and the accuracy of the system depends heavily on finding the best case for solving a new problem.

1) Retrieve

As a process of searching past cases with similar characteristics to new cases in the database, there are inductive searching method, knowledge-based searching method and nearest neighbor searching method. Inductive retrieval method has the advantage of faster case retrieval than the recent neighbor retrieval method because the goals are clearly organized and indexed by influencing factors on inductive results in the case data itself. On the other hand, if case data is missing, case inquiry becomes impossible at all, and the case has to be added to decision map whenever case increases. The knowledge-based retrieval method applies existing field knowledge in the process of retrieving the case in a similar way to the rule-based system. Lastly, Neighbor Lookup is a method of searching new cases and similar cases based on similarity measure among cases accumulated in the database. It is appropriate when the number of cases increases as it is appropriate when it is not necessary to focus on solving a specific problem. The time it takes to present a case increases, so it is suitable for a few cases. The recent neighbor lookup method, which is used as a representative case search method of case-based reasoning, refers to a method of retrieving similar cases by matching new and old cases by a certain similarity measure among stored cases.

2) Reuse

In order to reuse a case in which it is used to solve the problem, it is necessary to consider whether the case is suitable for solving a new problem.

3) Revise

After retrieving the most similar cases, the existing solution is modified to reflect the nature of the new problem. In the present invention, various methodologies described below are applied to increase the accuracy of energy usage predicted through the case-based reasoning model.

4) Retain

After the problem is solved, the process saves the case to a database so that it can be used for future problems. Regardless of the success of the generated case, the relevant information must be stored in a valid form for later use.

1 is a schematic block diagram showing the configuration of the economic and environmental evaluation system of a construction project by estimating the construction cost and CO ₂ emissions at the construction stage according to the present invention.

The construction cost and CO ₂ emission prediction system 100 at the construction stage according to an embodiment of the present invention includes a data server 110 in which construction project data, construction cost and CO ₂ emissions data are stored at the construction stage, Based on the construction project data collected by the data collection module unit 120 and the data collection module unit 120 to collect construction project data from the data server 110. Based on the projected construction quantity and construction project data, the construction quantity prediction module unit 130 for predicting construction quantity by material by applying a combination of case-based reasoning model, artificial neural network model, multiple rare analysis model and genetic algorithm It includes a construction cost and CO ₂ emissions calculation module unit 140 for calculating one or more CO ₂ emissions of the construction cost and material production, material transport or construction site CO ₂ emissions.

In the present invention, the construction project data is one of unit cost by material, material life, load of transportation equipment, fuel economy of transportation equipment, transportation distance, unit construction quantity, energy consumption per unit quantity of construction equipment or CO ₂ conversion coefficient per material It means more than.

In addition, a construction project means the kind, scale, construction method, and construction stage of a construction (building) to be constructed. In addition, the main material in the present invention means materials such as reinforcing bars, concrete, formwork, pipes, wood, stone.

Data server 110 is a server that stores similar conventional cases used in case-based reasoning. The construction cost and CO ₂ emissions at the construction stage are mainly derived from major works such as framing and finishing. The data server 110 of the present invention may be a separate server in which conventional materials are stored, or may be a separate device interworking with a server of an institution in which conventional materials are stored. In addition, the data server 110 stores construction project data, construction cost, and CO ₂ emission data at the construction stage.

The data server 110 stores the number of cases, case names, input attributes, output attributes, and the like. The scale of an attribute is also defined, including the nominal scale, the ratio scale, and the interval scale. A case-base is prepared in the case-based reasoning model through the data server 110.

The construction quantity prediction module unit 130 according to the present invention predicts the case similarity calculated by the case similarity calculator 131 and the case similarity calculator 131 that calculate the case similarity using a case-based reasoning model. Prediction value filtering unit 132 for filtering construction volume prediction values for each material by using an artificial neural network model and multiple regression analysis model, and parameter optimization unit for optimizing dependent parameters affecting construction material prediction values for each material by using a genetic algorithm ( 133).

The case similarity calculator 131 calculates case similarity by applying a case-based reasoning model to calculate attribute similarity and attribute weight, and finally to calculate case similarity. do.

Case-based reasoning is a methodology that derives Case Similarity through comparative analysis between cases in Test-Case and Case-Base based on project characteristics, and selects cases with high similarity in order. In the case-based reasoning model, a series of processes for calculating attribute similarity, attribute weight, and case similarity are applied, which can be represented by a simple matrix, as shown in Equation 1 below.

Here, AS denotes attribute similarity, AW denotes attribute weight, CS denotes case similarity, m denotes the number of cases, and n denotes the number of attributes.

The attribute similarity f _AS is calculated using the nearest neighbor retrieval method and is calculated using a function represented by Equation 2 below.

Here, f _AS is a function for calculating the degree of similarity property, _test AV _{_ case} the attribute values of the test cases, AV _{Retrieved _ case} indicates the value of the property of the search cases.

Attribute similarity is defined as a continuous scale, such as a scale scale or interval scale, and is calculated with the value expressed in Equation 2 when the attribute is above a minimum criterion for scoring attribute similarity (MCAS) and is less than MCAS. If 0 is determined. On the other hand, if the attribute is nominal and the attribute value is the same, the attribute similarity is 1; if the attribute value is not the same, the attribute similarity is 0.

Attribute weights are determined using genetic algorithms. Genetic algorithm is an optimization algorithm that is accumulated based on the principle of survival of the fittest and natural selection, which is the law of natural evolution. For genetic algorithms, there are four stages of fitness evaluation, reproduction, breeding and mutation. The fitness evaluation is a process of evaluating fitness in a population and selecting individuals according to the result, and then in reproduction, the individual with high suitability is selected and transferred to the next generation. Based on the intersection of two parental chromosomes, they are combined with each other, create a new individual, and then create a new individual from a mutation. As the evolution evolves, the individuals change into high-fitting individuals and converge to the optimal solution.

Case similarity is calculated through Equation 3 below. The weighted attribute similarity is derived by multiplying the attribute weight and the attribute similarity, and the case similarity is calculated when the accumulated sum is divided by the attribute weight sum.

f _CS is a function for calculating case similarity, f _AS is a function for calculating attribute similarity, f _AW is a function for calculating attribute weight, and n is the number of attributes.

As will be described later, the MCAS is set in the range of 1 to 100% using the genetic algorithm in the parameter optimizer 133, and the number of search cases is in the range of 1 to 100% using the genetic algorithm in the parameter optimizer 133. It is determined by the case selection range (RCS) which is a specific area to have, the attribute weight (RAW) is optimized to have a range of 0.01 ~ 1.00 using a genetic algorithm in the parameter optimizer 133.

In the previous studies, the case-based reasoning method can predict new values based on historical data, and predictive performance can be improved according to the continuous scale of the database. However, in spite of these advantages, the prediction accuracy is somewhat lower than that of the artificial neural network model. In the present invention, to compensate for the shortcomings of the case-based reasoning method, the effective prediction range is set using the predicted values through the artificial neural network model and the multiple regression analysis model, and the prediction values are filtered based on the estimated values. We want to improve the prediction accuracy.

2 illustrates an example of a filtering process performed in step S40 according to the predicted value filtering unit 132 of the construction quantity prediction module unit for each material according to the present invention or the method of the present invention.

The prediction value filtering unit 132 filters the construction quantity prediction value for each material by using an effective prediction range where a value predicted by applying an artificial neural network model and a value predicted by applying a multiple regression analysis model intersect.

The value PR _ANN predicted by applying the neural network model is a range expressed by Equation 4 below.

Here, PV _ANN is the predicted value of the artificial neural network model, SER _ANN is the standard error rate of the artificial neural network model. The equation description for the standard error rate is described later in the description of Prediction Accuracy.

On the other hand, the value (PR _MRA ) predicted by applying the multiple regression analysis model is a range expressed by Equation 5 below.

Where PV _MRA is the predicted value of the multiple regression model and SER _MRA is the standard error rate of the neural network model.

The effective prediction range (CRMA) of the present invention is a range expressed by Equation 6 below.

Here, PR _ANN is a value predicted by applying an artificial neural network model, PR _MRA is a value predicted by applying a multiple regression analysis model.

Furthermore, the effective prediction range (CRMA) of the present invention is a filtered value (CRMA *) as shown in Equation 7 below using a tolerance error prediction range (TRCRMA). In addition, it is preferable that the effective prediction range CRMA is a range filtered using a valid prediction range having a tolerance using the tolerance prediction range. Equation 7 below expresses this.

The filtered value CRMA * means a filtering range to which the tolerance prediction range TRCRMA is applied to the effective prediction range CRMA.

The TRCRMA is set in the range of 1 to 100% using a genetic algorithm in the parameter optimizer 133.

The construction cost and CO ₂ emission calculation module 140 applies the unit cost for each material to the construction quantity for each material predicted by the construction quantity prediction module 130 for each material. Calculate Cost.

Where EMQ is the estimated quantity of material, UC is the unit cost, and n is the number of materials such as concrete, rebar and formwork.

In addition, the construction cost and CO ₂ emission calculation module unit 140 using the CO ₂ conversion coefficient for each material for the construction quantity for each material predicted by the construction quantity prediction module 130 for each material, the following equation 9 To calculate the CO ₂ emissions from the production of materials.

Here, CM is the amount of CO ₂ generated during material production, EMQ is the estimated amount of material, CCF is the CO ₂ conversion factor, n is the number of materials such as concrete, rebar, formwork.

In addition, the construction cost and CO ₂ emission calculation module unit 140, the construction load for each material predicted by the construction quantity prediction module unit 130 for each material, the loading load of the transportation equipment, fuel efficiency, transportation distance and CO of the transportation equipment _{2 Using the} conversion factor, Equation 10 below calculates CO ₂ emissions during material transportation.

Where CT is the amount of CO ₂ generated during transportation of materials, EMQ is the estimated volume of material, ELC is the loading capacity of transportation equipment, TD is the one-way transportation distance, EFE is the fuel efficiency of transportation equipment, CCF is CO ₂ conversion factor, n is concrete The number of materials, such as rebar and formwork.

In addition, the construction cost and CO ₂ emission calculation module 140 applies the energy consumption per unit quantity of the construction equipment, CO ₂ conversion factor to the construction volume for each material predicted by the construction quantity prediction module unit 130 for each material By using Equation 11 below, the CO ₂ emission at the construction site is calculated.

Here, CC is the amount of CO ₂ generated at the construction site, EMQ is the estimated amount of material per unit, ECUQ is the energy consumption per unit of the construction equipment, CCF is the CO ₂ conversion factor, n is the number of materials, such as concrete, rebar, formwork.

Hereinafter will be described the construction cost and CO ₂ emissions prediction method in the construction stage according to the present invention. The overlapping contents of the construction cost and CO ₂ emission forecast in the above-described construction stage will be briefly described, and the focus will be on the key parts of the forecasting method.

Figure 3 is a flow chart schematically showing the order of the construction cost and CO2 emission prediction method in the construction stage according to the present invention.

Construction method cost and carbon dioxide emission prediction method in the construction stage according to the present invention is a step S10 inputting the construction project data and the test case data and search case data for the construction cost and CO ₂ emissions in the construction phase, the data server, The case similarity calculation unit of the construction quantity prediction module for each material includes a step S20 in which case similarity is calculated by applying a case-based reasoning model to the test case data and the search case data input from the step S10.

In addition, the construction cost and carbon dioxide emission prediction method in the construction stage according to the present invention is the case belonging to the case selection range (RCS) of the case similarity calculated in the step S20 in the prediction value filtering unit of the construction quantity prediction module for each material In step S30 , the construction value prediction module for each material , the prediction value filtering unit, the artificial neural network model and multiple regression analysis for the construction material prediction value predicted by the case similarity calculated in step S20 for the case selected in step S30 Step S40 that is filtered using the model.

In addition, the construction cost and carbon dioxide emission prediction method in the construction step according to the present invention is the parameter optimization unit of the construction quantity prediction module unit for each material, in the construction quantity prediction value for each material filtered in step S40 for the case selected in step S30 S50 step of calculating the average prediction accuracy for.

Furthermore, when the parameter optimization unit of the construction material quantity prediction module for each material is less than the reference accuracy, the average prediction accuracy calculated in step S50, the dependent parameter affecting the construction material quantity prediction value for each material includes a step S60 is optimized through the genetic algorithm.

In addition, the construction cost and CO ₂ emission calculation module Bouygues Construction project data and materials on the basis of specific construction volumes construction costs and materials CO ₂ emissions of one or more of the CO ₂ emissions of materials during transportation or construction site during production calculated in S40 step Calculating step S70.

Steps S20 to S50 are repeatedly performed using the optimized dependent parameter. This process is illustrated in detail in FIG.

In step S20, the case similarity represented by Equation 3 is calculated using the attribute similarity and the attribute weight.

f _AS is a property similarity calculated using a nearest-neighbor retrieval method, which is represented by Equation 2 above.

In step S40, the construction quantity prediction value for each material is filtered using an effective prediction range where the value predicted by applying the neural network model and the value predicted by applying the multiple regression analysis model intersect.

The value predicted by applying an artificial neural network model (PR _ANN ), the value predicted by applying a multiple regression analysis model (PR _MRA ) is expressed by the equation shown in Equation 4 above. Further, the effective prediction range CRMA * filtered using the effective prediction range CRMA and the tolerance error valid prediction range TRCRMA is also expressed by the equations (5) through (7).

In step S50, the average prediction accuracy is an average of prediction accuracy for each case, and the prediction accuracy for each case is calculated using Equation 12 below.

Here, f _PA is a function for calculating prediction accuracy, and f _SER is a function for calculating standard error rate. f _SER is expressed by Equation 13 below.

Here, the actual value of the construction volume for each material that is the AV dependent variable, PV is the predicted value of the dependent variable, m is the number of cases.

Dependent parameters optimized through genetic algorithm at step S60 Minimum reference value (MCAS) for determining attribute similarity used for case similarity calculation in step S20, range of attribute weights (RAW) used for case similarity calculation in step S20, case selection range (RCS) in step S30, S40 At least one of the effective predicted range (CRMA) where the value predicted by applying the neural network model and the predicted value by applying the multiple regression model intersects or the tolerance predicted effective range (TRCRMA) with the tolerance added to CRMA. Can be.

Step S70 is the same as Equation 8 to Equation 11 described in the above-described construction cost and CO ₂ emission calculation module 140.

Figure 4 is a flow chart mainly showing the optimization process using a case-based reasoning model and genetic algorithm in the construction cost and CO ₂ emission prediction method in the construction stage according to the present invention.

The block labeled (1) is the process of calculating the case similarity, the block (2) is the process of selecting and filtering the predicted value as the dependent variable, and the block (3) is the process of optimizing the dependent parameter through the genetic algorithm. Means.

In the block marked (1), two optimization parameters are applied to calculate case similarity. Two optimization parameters are the Minimum Criterion for scoring the Attribute Similarity (MCAS) and the Range of the Attribute Weight (RAW).

 In the block denoted by (2), two optimization parameters are applied to improve prediction accuracy and are used as a filtering device. Two optimization parameters to improve prediction accuracy are the range of the case selection (RCS) and the tolerance range of the cross-range of the results of neural networks and multiple regression analysis. between MRA and ANN; TRCRMA).

In the block labeled (3), the tolerance for minimum criteria (MCAS) for calculating attribute similarity scores, range of attribute weighting (RAW), range of case selection (RCA), and cross range of artificial neural network and multiple regression analysis results. Range (TRCRMA) is driven in conjunction with genetic algorithms and contributes to improved prediction accuracy.

The case-base is established when the method of estimating construction costs and CO ₂ emissions at the construction stage begins. This means that a case is constructed in the data server 110 in the construction cost and CO ₂ emission prediction system 100 at the construction stage. In step S60 of determining whether the average of the prediction accuracy (APA) is less than the reference accuracy, the process of calculating the average of the prediction accuracy and the number of prediction cases (NPC), the average of the prediction accuracy (APA) and the prediction case A process of determining whether the number NPC exceeds 95% of the total cases and a process of determining whether the APA satisfies the maximum value (each step is indicated by a curve in FIG. 4). In FIG. 4, a genetic algorithm was used to allow APA to have a maximum value when the NPC was 95% or more of all cases, and if the APA had a maximum value, the dependent algorithm was newly generated if the APA had no maximum value. If the maximum value is reached, the process is terminated.

5 and 6 are graphs showing the construction cost and CO ₂ emissions estimation results in the construction stage in the bar phase according to the present invention. As shown, Figures 5 and 6 show some differences in the target and actual values of construction costs and CO ₂ emissions, depending on the final decision maker's criteria (e.g. Best Estimate, Low Estimate and High Estimate). Suggests that is occurring.

These results may vary as the properties of various factors, such as the projected value of construction volume by material, type of material and equipment, unit cost, and CO ₂ conversion factor, are changed. Therefore, economic and environmental analysis should be conducted in consideration of the characteristics of the construction project, and the target criteria for the construction cost and CO ₂ emissions of the project should be established based on the results.

The embodiments and the drawings appended hereto are merely to clearly illustrate some of the technical ideas included in the present invention, and may be easily inferred by those skilled in the art within the scope of the technical ideas included in the specification and drawings of the present invention. It will be apparent that both the modified examples and the specific embodiments are included in the scope of the present invention.

100: Construction cost and CO ₂ emission prediction system at the construction stage
110: data server 120: data collection module
130: construction quantity prediction module unit by material 131: case similarity calculation unit
132: prediction value filtering unit 133: parameter optimization unit
140: construction cost and CO ₂ emissions calculation module

Claims (31)

A data server storing construction project data, construction cost and CO ₂ emission data in the construction stage;
A data collection module unit configured to collect construction project data from the data server;
Based on the construction project data collected by the data collection module unit Construction quantity prediction module unit for predicting the construction quantity by material; And
Comprising the construction materials by volume and construction projects based on the construction cost data and materials during production, materials during transportation or construction costs to calculate the CO ₂ emissions of one or more of the CO ₂ emissions of the construction site and CO ₂ emissions calculation module unit,
Construction cost in the construction stage characterized in that the combination of case-based reasoning model, artificial neural network model, multiple regression analysis model and genetic algorithm to the construction volume data forecast by the construction volume prediction module by material and CO2 Emission Prediction System.
The method of claim 1, wherein the construction project data,
Construction stage characterized by at least one of unit cost by material, material life, loading load of transportation equipment, fuel economy of transportation equipment, transportation distance, unit construction volume, energy consumption per unit quantity of construction equipment, or CO ₂ conversion factor by material Construction cost and CO2 emission forecasting system.
3. The method of claim 2,
The construction quantity prediction module unit for each material,
A case similarity calculator for calculating case similarity using a case-based reasoning model;
A prediction value filtering unit for filtering the construction quantity prediction value for each material predicted through the case similarity calculated by the case similarity calculator using an artificial neural network model and a multiple regression analysis model; And
And a parameter optimization unit for optimizing the dependent parameter affecting the construction material quantity prediction value for each material by using a genetic algorithm.
The method of claim 3,
The case similarity calculation unit,
The construction cost and carbon dioxide emission prediction system at the construction stage, characterized by calculating the case similarity represented by the following equation using the attribute similarity (f _AS ) and the attribute weight (f _AW ).

(Where f _CS is a function that computes case similarity, f _AS is a function that computes attribute similarity, f _AW is a function that computes attribute weight, n is the number of attributes)
5. The method of claim 4,
The property similarity (f _AS ) is a property similarity calculated using a recent neighbor lookup method, it is expressed by the following equation, the construction cost and carbon dioxide emission prediction system in the construction stage.

(Where AV _{test _ case} is the attribute value of the test case, AV _{Retrieved _ case} is the attribute value of the search case, and MCAS is the minimum reference value for determining attribute similarity)
The method of claim 5,
The MCAS is set to a range of 1 to 100% using a genetic algorithm in the parameter optimizer, and the number of search cases is a specific region having a range of 1 to 100% using a genetic algorithm in the parameter optimizer. Determined by the (RCS), the property weight is optimized by the parameter optimizer using a genetic algorithm to have a range of 0.01 ~ 1.00 in the construction stage construction cost and carbon dioxide emission prediction system.
The method of claim 3,
The prediction value filtering unit applies a neural network model to filter the prediction value for each construction material by using an effective prediction range where the predicted value intersects with the predicted value by applying a multiple regression analysis model. And carbon dioxide emission prediction system.
The method of claim 7, wherein
The value predicted by applying the artificial neural network model (PR _ANN ) is a construction cost and carbon dioxide emission prediction system in the construction stage, characterized in that the range expressed by the following equation.

Where PV _ANN is the predicted value of the neural network model and SER _ANN is the standard error rate of the neural network model.
The method of claim 7, wherein
The value predicted by applying the multiple regression analysis model (PR _MRA ) is a construction cost and carbon dioxide emission prediction system in the construction stage, characterized in that the range expressed by the following equation.

Where PV _MRA is the predicted value of the multiple regression model and SER _MRA is the standard error rate of the neural network model.
The method of claim 7, wherein
The effective prediction range (CRMA) is a construction construction cost and carbon dioxide emission prediction system in the construction phase, characterized in that the range expressed by the following equation.

(Here, PR _ANN is a value predicted by using an artificial neural network model, PR _MRA is a value predicted by applying a multiple regression analysis model)
The method of claim 10,
The effective prediction range is a construction cost and carbon dioxide emission prediction system in the construction stage, characterized in that the filter value (CRMA *) filtered using the following formula (TRCRMA).

12. The method of claim 11,
The TRCRMA is a construction construction cost and carbon dioxide emission prediction system in the construction stage, characterized in that the parameter is set in the range of 1 ~ 100% using a genetic algorithm.
According to claim 2, The construction cost and CO ₂ emissions calculation module unit,
Construction construction cost in the construction stage by calculating the construction cost by applying the unit cost for each material to the construction quantity for each material predicted by the module prediction quantity by material, as shown in the following equation CO2 Emission Prediction System.

Where EMQ is the estimated quantity of material, UC is the unit cost, and n is the number of materials such as concrete, rebar and formwork.
The method of claim 13, wherein the construction cost and CO ₂ emissions calculation module unit,
In the construction quantity prediction module part by material With respect to the projected construction volume by material, using the CO ₂ conversion coefficient for each material, it is characterized by calculating the CO ₂ emissions during material production by the following equation Construction cost and CO2 emission prediction system at the construction stage.

Where CM is the amount of CO ₂ generated during material production, EMQ is the estimated volume of material, CCF is the CO ₂ conversion factor, and n is the number of materials such as concrete, rebar and formwork.
The method of claim 14, wherein the construction cost and CO ₂ emissions calculation module unit,
With respect to the material-specific construction amount predicted material specific construction amount prediction in the module unit, transport loading capacity of the equipment, the fuel economy of the transport equipment, transport distance, and by applying the CO ₂ conversion factors, during transport material by the following equation under the CO ₂ emissions To calculate the Construction cost and CO2 emission prediction system at the construction stage.

Where CT is the amount of CO ₂ generated during transportation of materials, EMQ is the estimated volume of material, ELC is the loading capacity of the transport equipment, TD is the one-way transport distance, EFE is the fuel economy of the transport equipment, CCF is the CO ₂ conversion factor, and n is Number of materials such as concrete, rebar and formwork)
The method of claim 15, wherein the construction cost and CO ₂ emissions calculation module unit,
For the construction volume forecast by the module, the energy consumption per unit volume of the construction equipment and the CO ₂ conversion factor are applied to calculate the CO ₂ emissions at the construction site. As a feature Construction cost and CO2 emission prediction system at the construction stage.

Where CC is the amount of CO ₂ generated at the construction site, EMQ is the estimated quantity of material per unit, ECUQ is the energy consumption per unit of construction equipment, CCF is the CO ₂ conversion factor, and n is the number of materials such as concrete, rebar and formwork.
S10 step of inputting the construction project data and the test case data and search case data for the construction cost and CO ₂ emissions in the construction phase in the data server;
Step S20 of calculating the case similarity by applying a case-based reasoning model to the test case data and the search case data inputted from the step S10 in the case similarity calculator of the construction quantity prediction module for each material;
A step S30 of selecting a case belonging to a case selection range (RCS) among cases in which case similarity is calculated in step S20 in the prediction value filtering unit of the construction quantity prediction module for each material;
In the prediction value filtering unit of the construction volume prediction module by material, the construction construction material prediction value predicted through the case similarity calculated in the step S20 for the case selected in the step S30 is filtered using the artificial neural network model and the multiple regression analysis model. S40 step;
A step S50 of calculating, by the parameter optimization unit of the construction material quantity prediction module for each material, an average prediction accuracy of the construction material quantity prediction values filtered in the step S40 for the case selected in the operation S30;
A step S60 of optimizing the dependent parameter affecting the construction material quantity prediction value according to the genetic algorithm when the parameter optimization unit of the construction material quantity prediction module for each material is less than the reference accuracy; And
Construction costs and CO ₂ emissions calculation module additional construction project data and based on a material-specific construction volume calculated in S40 calculating construction costs and materials CO ₂ emissions of one or more of the CO ₂ emissions of materials during transportation or construction site during production Including the S70 stage,
Method of predicting the construction cost and carbon dioxide emissions in the construction stage, characterized in that the step S20 is repeatedly performed using the optimized dependent parameter.
The method of claim 17, wherein the construction project data,
Unit cost by material, material life, load capacity of transportation equipment, fuel economy of transportation equipment, transportation distance, unit A method of predicting construction cost and carbon dioxide emission at a construction stage, characterized in that at least one of construction quantity, energy consumption per unit quantity of construction equipment, or CO ₂ conversion factor.
The method of claim 18, wherein the step S20,
A method of predicting construction cost and carbon dioxide emission at a construction stage, comprising calculating the case similarity represented by the following equation using the attribute similarity and the attribute weight.

(Where f _CS is a function that computes case similarity, f _AS is a function that computes attribute similarity, f _AW is a function that computes attribute weight, n is the number of attributes)
20. The method of claim 19,
The f _AS is a property similarity calculated using the recent neighbor lookup method, it is expressed by the following equation, the construction cost and CO2 emission prediction method in the construction stage.

(Where AV _{test _ case} is the attribute value of the test case, AV _{Retrieved _ case} is the attribute value of the search case, and MCAS is the minimum reference value for determining attribute similarity)
The method of claim 18, wherein the step S40,
Cost of construction work in the construction stage, characterized in that the construction quantity prediction value for each material is filtered using the effective prediction range that the value predicted by applying the artificial neural network model and the value predicted by applying the multiple regression analysis model intersect. How to predict CO2 emissions.
The method of claim 21,
Value predicted by applying the artificial neural network model (PR _ANN ) is a construction construction cost and carbon dioxide emission prediction method in the construction stage, characterized in that the range expressed by the following equation.

Where PV _ANN is the predicted value of the neural network model and SER _ANN is the standard error rate of the neural network model.
The method of claim 21,
The value predicted by applying the multiple regression analysis model (PR _MRA ) is a construction cost and carbon dioxide emission prediction method in the construction stage, characterized in that the range expressed by the following equation.

Where PV _MRA is the predicted value of the multiple regression model and SER _MRA is the standard error rate of the neural network model.
19. The method of claim 18,
The effective prediction range (CRMA) is a construction construction cost and carbon dioxide emission prediction method in the construction phase, characterized in that the range expressed by the following equation.

(Here, PR _ANN is a value predicted by using an artificial neural network model, PR _MRA is a value predicted by applying a multiple regression analysis model)
25. The method of claim 24,
The effective prediction range is a construction cost and carbon dioxide emission prediction method in the construction stage, characterized in that the filter value (CRMA *) filtered using the following formula (TRCRMA).

19. The method of claim 18,
In the step S50, the average prediction accuracy is an average of prediction accuracy for each case, and the prediction accuracy for each case is calculated by using the following equation. .

(Where f _PA is a function for calculating prediction accuracy, f _SER is a function for calculating standard error rate,

Where AV is the actual value of construction volume by material, the dependent variable, PV is the predicted value of the dependent variable, and m is the number of cases)
19. The method of claim 18,
The dependent parameter optimized through the genetic algorithm in step S60,
Minimum reference value (MCAS) for determining attribute similarity used for case similarity calculation in step S20, range of attribute weights used for case similarity calculation in step S20, and case selection range (RCS) in step S30 ), The effective prediction range (CRMA) where the value predicted by applying the artificial neural network model and the value predicted by applying the multiple regression analysis model or the allowable error effective prediction range by adding a tolerance to the CRMA in step S40 ( TRCRMA) is one or more of the construction cost and CO2 emission prediction method in the construction stage.
The method of claim 18, wherein the step S70,
The construction cost and carbon dioxide emission prediction method at the construction stage, characterized in that for calculating the construction cost for each material predicted in step S40, by applying the unit cost for each material as shown in the following equation.

Where EMQ is the estimated quantity of material, UC is the unit cost, and n is the number of materials such as concrete, rebar and formwork.
The method of claim 28, wherein the step S70,
For the construction volume for each material predicted in step S40, using the CO ₂ conversion coefficient for each material, it is characterized by calculating the CO ₂ emissions during production of the material by the following equation Method of predicting construction cost and carbon dioxide emission at the construction stage

Where CM is the amount of CO ₂ generated during material production, EMQ is the estimated volume of material, CCF is the CO ₂ conversion factor, and n is the number of materials such as concrete, rebar and formwork.
The method of claim 29, wherein the step S70,
In step S40 With respect to the projected construction volume by material, by applying the loading load of the transportation equipment, fuel economy, transportation distance and CO ₂ conversion factor of the transportation equipment, it is characterized by calculating the CO ₂ emissions during transportation of materials by the following equation Method of predicting construction cost and carbon dioxide emission at the construction stage

Where CT is the amount of CO ₂ generated during transportation of materials, EMQ is the estimated volume of material, ELC is the loading capacity of the transport equipment, TD is the one-way transport distance, EFE is the fuel economy of the transport equipment, CCF is the CO ₂ conversion factor, and n is Number of materials such as concrete, rebar and formwork)
The method of claim 30, wherein the step S70 ,
In step S40 With respect to the projected construction volume for each material, the energy consumption per unit volume of construction equipment, CO ₂ conversion factor is applied, and the following equation is used to calculate the CO ₂ emissions at the construction site. Method of predicting construction cost and carbon dioxide emission at the construction stage

Where CC is the amount of CO ₂ generated at the construction site, EMQ is the estimated quantity of material per unit, ECUQ is the energy consumption per unit of construction equipment, CCF is the CO ₂ conversion factor, and n is the number of materials such as concrete, rebar and formwork.

Images