Analysis of the Mechanism Construction of Intelligent Construction Enabling Green Transformation in the Construction Industry and Its Influencing Factors on Sustainable Development Paths

At present, the concept of green, low carbon and energy saving is deeply rooted in people’s hearts, in order to realize green, low carbon and sustainable development in the pre-design stage of the building, and strive to improve the functionality and economy, it is necessary to apply BIM technology in the architectural design stage, and realize the visualization of architectural design. Relying on the advantages of digital technology, the visualization model of the building system is constructed, and through the functions of three-dimensional visualization, model simulation and information data sharing, the building environment, building planning layout and other information are analyzed in an integrated manner, so as to build up the data and information database in the building design stage, and provide information technology support for the building design. For example, with the help of BIM technology to simulate the external light, ventilation, humidity and temperature of the building, analyze the building space suitable for living, design green, low-carbon and environmentally friendly design solutions, effectively reduce the building carbon emissions, and contribute to environmental protection. At the same time, to promote civil engineering, electrical engineering and water supply and drainage engineering and other projects to carry out synchronous, special construction design, with the help of the Internet and other technical means to realize the design of each stage of the visualization, to ensure that the mid-term building construction technology of the advanced, scientific, maintenance of the later building management and operation, for the construction industry to lay a solid foundation for the green and low-carbon transformation.

The building construction environment is complex, and there are more dangerous factors. In recent years, with the accelerating process of intelligentization, industrialization, and greening, the requirements for building construction have been increasing. Following the development requirements of the time, BIM technology and virtual reality (VR) technology are applied to construction projects to achieve an all-round simulation of building design, construction, and operation management. Especially in the construction stage, in order to fully grasp the construction situation, the construction site is simulated in real life, so as to discover on-site construction problems in a timely manner and take countermeasures to prevent all kinds of problems from occurring in the construction site. At the same time, with the introduction of intelligent robots and other related technologies, the robot will be included in the construction site management, to achieve a similar “human eye” on-site monitoring status. For the construction of hazardous areas of the construction monitoring construction, can be directly remote control robot instead of manual completion, not only helps to improve construction safety, but also improve construction efficiency and construction quality. In addition, for the traditional building construction method of green, environmental protection is low, more and more of China’s projects using assembly building technology form, integration of information technology means (such as BIM technology), give full play to BIM technology in the production of prefabricated parts of assembly building, transportation and installation, in order to effectively save costs, reduce carbon emissions, strengthen environmental protection, in order to ensure the quality of engineering construction at the same time to meet the requirements of green construction, accelerate the advancement of construction efficiency and quality of construction. Meanwhile, it meets the requirements for green construction and accelerates the transformation of the construction industry towards green and low-carbon construction.

In a construction project, the project management team was faced with the challenge of ensuring that the building materials met the high standards of green building - to solve this problem, the project adopted a set of advanced intelligent identification system, which is specifically designed for the inspection and management of material quality - the core of the intelligent identification system is based on artificial intelligence image recognition technology and big data analysis. The system first scans the barcode or QR code of the material and automatically extracts detailed information about the material, including the manufacturer, production batch performance indicators and the corresponding environmental certification information - for example, for an energy efficient glass used in a construction project, the system is able to quickly verify whether its thermal conductivity meets the design standards to ensure that it has a good insulation performance - in addition, the intelligent identification system is equipped with a high-resolution camera to check the quality of incoming materials. In addition, the intelligent identification system is equipped with a high-resolution camera that takes photos of incoming materials and automatically detects whether there are any damages or defects on the surface of the materials through image analysis technology. During the early stages of the project, the system successfully identified a number of decorative stones with minor cracks on the exterior. Through timely feedback, the project team negotiated with the supplier to replace these substandard stones, avoiding potential safety hazards and rework costs.

Intelligent scheduling system adopts the most advanced artificial intelligence algorithm and integrates project management. It integrates several modules, such as project management, resource scheduling, and environment monitoring. Through real-time collection of construction site data (including temperature, humidity, personnel distribution, machinery and equipment use, etc.), the system can dynamically adjust the construction plan to optimize resource allocation. Through real-time collection of construction site data (including temperature, humidity, distribution of personnel, use of machinery and equipment, etc.), the system is able to dynamically adjust the construction plan and optimize the allocation of resources. In specific applications, the system first analyzes the overall construction plan of a construction project, identifying the peak energy consumption and key construction nodes.5 Then, through real-time monitoring of the construction team and machinery and equipment, the system is able to predict hot spots in the use of resources and adjust the allocation of personnel and equipment accordingly, avoiding the waste of resources. , avoiding wasteful resources. For example, for the underground garage construction phase of the project, by analyzing the construction tasks and site conditions, the intelligent scheduling system decided to carry out the concrete pouring work in the morning and evening when the temperature was lower to reduce the accelerated rate of concrete curing due to high temperatures. This adjustment not only improved the working performance of the concrete, but also reduced the need for additional water addition and re-pouring, which in turn reduced the consumption of water and carbon emissions. In addition, the system optimized the transportation plan for construction materials. By analyzing the distance between the construction site and the material supply location, the transportation routes, and the traffic conditions, the intelligent scheduling system arranged the best transportation time and batch, effectively reducing energy consumption and emissions during transportation. For example, during the transportation of materials, the system can reduce carbon emissions by about 20% per transportation through rational scheduling.

Drones are used to capture high-definition video and photos of construction sites, and this real-time data is transmitted to the project management center, enabling the management team to monitor construction progress and site safety in real time. For example, the drone was used to fly and photograph the main construction areas of a construction project at regular intervals each day. By comparing and analyzing the video data from consecutive days, the project management team was able to identify deviations in construction progress and potential safety hazards, such as construction activities that were not executed according to plan or improperly stacked construction materials.

At the same time, several environmental monitoring sensors were installed at the construction site to monitor environmental parameters such as air quality, noise level, temperature, and humidity in real-time. The data from these sensors is also transmitted back to the project management center in real-time to ensure that construction activities do not have an unacceptable impact on the surrounding environment. In one specific case, the sensors detected that the noise level in a construction area exceeded the specified standard, and the project team immediately took vibration mitigation measures, such as adjusting the construction time and adding soundproofing facilities, to minimize the impact on the surrounding residents.

If there are p independent variables used to explain dependent variable Y: X₁, X₂, ⋯, X_p, then the formula between them is (1)Y=β0+β1X1+β2X2+⋯+βpXp+ε

where ε is the error term and β₀, β₁, ⋯, β_p is the unknown parameter. There are n sets of observations (xi1,xi2,⋯,xip,yi) , i = 1, 2, ⋯, n for the independent variable X₁, X₂, ⋯, X_p, which satisfy (2)yi=β0+β1xi1+β2xi2+⋯+βpxip+ε

We must make certain settings for the variables in Eq. First of all, in the multivariate linear equation, the independent variable x₁, x₂, ⋯, x_p is not a random variable. Also the error terms ε_i, i = 1, 2, ⋯, n are random variables, which are generated by a combination of factors. Since the error random variable ε is random, the dependent variable Y is also random. We first assume that the error terms ε_i, i = 1, 2, ⋯, n satisfy the following assumptions, i.e. (3)E(εi)=0 var(εi)=σ2 cov(εiεj)=0,i≠j

We set the dependent variable about the independent variable in the form (1).

The regression model is now written in matrix form: (4)( y1 y2 ⋮ yn)=( 1 x11 ⋯ x1p 1 x21 ⋯ x2p ⋮ ⋮ ⋮ ⋮ 1 xn1 ⋯ xnp)( β1 β2 ⋮ βp)+( ε1 ε2 ⋮ εn)

Noting the column vectors or matrices in (4) as y, X, β and ε in turn, the matrix form of (4) can again be written as: (5)y=Xβ+ε

where y is the observed variable in n × 1 dimensions; X is the known matrix in n × (p + 1) dimensions: β is the unknown parameter vector in (p + 1) × 1 dimensions; and ε is the random error vector. Markov’s assumption that is: (6)E(ε)=0,cov(ε)=σ2In

In summary, the linear regression model is: (7)y=Xβ+ε,E(ε)=0,cov(ε)=σ2In

For regression analysis, a question that must be investigated is how the parameters should be estimated.

For the problem of how to estimate the parameter vector β using the observations of the independent and dependent variables, the method usually used is Gaussian least squares. The idea is that, based on the regression model described above, we can know that Xβ is the predicted value of the dependent variable obtained by the model’s prediction of multiple independent variables in the training sample, but in reality, we know that the value of the dependent variable is y. So the distance between the actual y and our prediction of Xβ looks like this: (8)Q(β)=‖y−Xβ‖2=(y−Xβ)T(y−Xβ)

Then find β such that Q(β) is minimized. Equation (8) is now expanded as (9)Q(β)=yTy−2yTXβ+βTXTXβ

Take the partial derivative with respect to β so that its value is zero to obtain the system of equations: (10)XTXβ=XTy

Solution for: (11)β^=(XTX)−XTy

Here (XTX)− is any generalized inverse of X^TX. According to the extremum theory of calculus, β^ is just a stationary point of the function Q(β) . We also need to show that β^ does minimize Q(β) . For any β there is: (12)Q(β) = ‖y−Xβ‖2=‖y−Xβ^+X(β^−β)‖2 = ‖y−Xβ^‖2+(β^−β)TXTX(β^−β) +2(β^−β)TXT(y−Xβ^)

Since β^ satisfies Eq. (10), XT(y−Xβ^)=0 , follows: (13)‖y−Xβ‖2=‖y−Xβ^‖2+(β^−β)TXTX(β^−β)

Again, since X^TX is a positive definite array, it follows that (14)Q(β)=‖y−Xβ‖2≥‖y−Xβ^‖2=Q(β^)

This shows that β^ does minimize Q(β) . (15)β^=(XTX)−1XTy

At this point, β^ is called the least squares estimate of β.

An extremely important metric in regression analysis is the coefficient of determination R², which measures how well the independent variable x₁, x₂, ⋯, x_p fits the dependent variable y. R² is the proportion of the sum of squared deviations explained by the model to the total sum of squared deviations, i.e. (16)R2=∑i=1n(y^i−y¯)2∑i=1n(yi−y¯2)

Obviously 0 ≤ R² ≤ 1, the larger its value, the better the model fit.

The stepwise regression method is a method that combines the characteristics of the stepwise exclusion method and the stepwise introduction method. Its basic principle is: starting from an explanatory variable, the regression equation is introduced from the largest to the smallest depending on the significance of the effect of the explanatory variable on the explained variable; at the same time, the explanatory variables are selected into the regression equation one by one. If it is discovered that an explanatory variable introduced earlier diminishes in importance later due to the introduction of other explanatory variables, it can be removed from the regression equation at any time. The introduction of a variable or the exclusion of a variable is a step in the stepwise regression, and at each step a significance test is performed to ensure that only significant variables are included in the regression equation before each introduction, and this process is repeated until there are no insignificant variables to be excluded from the equation, and there are no significant variables to be introduced into the regression equation.

Multiple stepwise regression is an important analytical method in linear regression analysis, which is mainly used to solve the problem of how to select explanatory variables when there are many explanatory variables in the regression model, so that the regression equation contains all the explanatory variables that have a significant impact on the explanatory variables but does not contain the explanatory variables that have no significant impact. Multivariate stepwise regression is a regression approach designed to solve this problem. The commonly used selection methods are stepwise selection-in method and stepwise exclusion method, and both methods get the same result in most cases, i.e., the explanatory variables included in the final obtained regression model are the same (Liu Lixiang, 2015). According to the m explanatory variables {x1,x2,…,xm} screened by the information gain value, this paper uses the stepwise selection method to analyze, and its main idea is to introduce the regression equation one by one among all the explanatory variables considered according to the size of their contribution to the explanatory variables, and the variables that have been introduced into the regression equation may also lose their importance after the introduction of the new variables and need to be eliminated from the regression equation. The introduction of a variable or the exclusion of a variable from the regression equation is subjected to a F-test to ensure that the regression equation contains only those variables that have a significant effect on the explanatory variables before the introduction of the new variable, and that those that are not significant have been excluded, as follows:

Step 1: A m univariate regression equation is established with y each of the m explanatory variables x₁, x₂, ⋯, x_m and the value of the F test statistic for the corresponding regression coefficient of each explanatory variable is computed, denoted as F₁₁, F₁₂, ⋯, F_1m and set F1i=max{F11,F12,⋯,F1m} to a critical value F_α(1, n − 2) for a pre-given level of significance α. If F_1i > F_α(1, n − 2) so, its corresponding variable xi1 is introduced into the regression model.

Step 2: A subset of the selected explanatory variables {xi1} and the remaining explanatory variables x_j(1 ≤ j ≤ m, j ≠ i) are used to set up m − 1 binary linear regression equations with the explanatory variables Y, and the values of the F statistics are also computed and denoted as F₂₁, F₂₂, ⋯F_2i−1, F_2i+1, ⋯F_2m. Set F2i=max{F21,F22,⋯F2i−1,F2i+1,⋯F2m} , for the given level of significance F_α(1, n − 3), if F_2i > F_α(1, n − 3), then the corresponding variables xi2 are introduced into the regression equation. At this point a subset of the variables has been selected to be {xi1,xi2} introduced. If F_2i < F_α(1, n − 3) then the introduction of the variables is terminated.

Step 3: Repeat the method in step 2 with the selected subset of explanatory variables {xi1,xi2} until no variable can be introduced into the equation, i.e., the explanatory variables introduced have good explanatory effect on the explanatory variables, and the explanatory variables selected by the stepwise regression method will be labeled as {xi1,xi2,……,xiq} .

Step 4: After multiple stepwise regression, the following equation is obtained: (17)Y=β0+β1xi1+β2xi2+⋯+βpxiq

Where, {β0,β1,β2,……βq} is the regression coefficient of each indicator of the selected regression equation.

The stepwise regression algorithm is a regression algorithm that is widely used today. This algorithm works by introducing explanatory variables one by one, each time the explanatory variable introduced has the most significant effect on the explanatory variable Y. Each time a new explanatory variable is introduced, the old explanatory variables previously introduced into the regression equation are tested one by one, and the explanatory variables that are not significant in the current equation are eliminated one by one, starting from those that have the least influence on the explanatory variable Y, until no new explanatory variables can be introduced. Eventually the explanatory variables that were kept in the regression equation were all significant in affecting the explained variable Y, while the explanatory variables that were not in the regression equation were all insignificant in their effect on Y.

This study categorizes the mechanism influencing factors of green transformation in the construction industry into six aspects: mechanism research and development capability, mechanism standard improvement degree, transformation and application of mechanism results, green building design level, green building construction level, and green building operation level. 1)

Mechanism R&D capability. Relative to the traditional building mechanism, green building mechanism process is more complex, more difficult to operate, and relatively high cost, so its development and application is not overnight, and mechanism research and development is undoubtedly the basic link in the development process of green building technology. If we can continuously increase the investment in scientific research, take various measures to develop more green building mechanisms, and promote the enhancement of green building mechanism research and development capabilities, it will lay a good foundation for the progress of green building technology.

In this study, we designed a questionnaire and adopted the method of distributing the questionnaire to green building related professionals to obtain the relevant data. The distribution of questionnaires primarily focuses on government authorities, universities and research institutions, managers and technicians of related enterprises. The designed questionnaire uses a 1~5 semantic difference scale to divide the impact of the factors of the green building mechanism on sustainable development into five levels, which are “very large”, “large”, “average”, “small” and “very small”. The scale asks respondents to score each factor based on the extent to which it affects sustainable development, based on the current state of the green building mechanism, the sustainable development path, and their personal expertise. For example, if a factor has a great influence on sustainable development, the factor is given a value of 5 points; if a factor has a small influence on sustainable development, the factor is given a value of 1 point.

This study through the adoption of e-mail, telephone interviews and on-site research and other ways to issue a total of 150 questionnaires, returned 126 questionnaires, the questionnaire recovery rate of 84%, excluding some of the answers are unclear, there are problems with the questionnaire, and finally after the collation of the questionnaires, to get the effective questionnaire 109, the effective questionnaire rate of 72.7%, in line with the minimum questionnaire validity rate of not less than 50% of the requirements of the statistics to meet the needs of this study, the questionnaire validity rate of not less than 50%. The validity rate reached 72.7%, which meets the requirement of a minimum validity rate of not less than 50% and meets the needs of this study.

Based on the questionnaire on the impact factors of green building mechanisms on sustainable development, the number of people in the valid questionnaire who chose five levels of impact degree as the rating of the importance degree of this impact factor, and standardize the data, so as to get the standardized matrix of the impact factors of green building mechanisms on sustainable development, as shown in Table 1.

In this section, we will continue to regression analysis of the green building mechanism on the path of sustainable development and its various influencing factors, the research data selection of the collection and organization of the sustainable development of construction as the dependent variable data, selecting the questionnaire in the various indicators of the impact of the data as the independent variable data.

According to the multivariate linear regression model of green building mechanism development and its influencing factors on sustainable development constructed earlier, the data of each variable after logarithmic processing are imported into SPSS20.0 statistical software, and the results of the model are obtained through regression analysis, Table 2 is the result of ANOVA, and Table 3 is the table of regression coefficients.

From the ANOVA table in Table 2, the F-statistic used to test the significance of the regression effect of the model is 110.674, and its p-value is 0<0.05, which indicates that the regression model passed the test of significance at the 5% confidence level, and the

The regression effect is significant.

According to the partial regression coefficients of each independent variable in the regression model and its corresponding standard deviation and t-test results, it can be seen that through the t-test (Sig < 0.05) the variables are mechanism research and development ability, mechanism standard improvement, mechanism results transformation and application, green building design level, green building construction level and green building operation level six independent variables, the partial regression coefficients are respectively 0.294, 0.231, 0.143, 0.328, 0.354 and 0.184, the explanatory variables all have more significant linear relationship with the dependent variables, and the constant term regression coefficient is -3.544. The correlation of the degree of perfection of its mechanism standard is skewed to 0.078<0.01, then it is considered that there is no more significant relationship between its independent variables and the dependent variable, that is to say, there is no more significant relationship between the degree of perfection of the mechanism standard and the path of sustainable development. The last two columns of covariance statistics data can be seen. The VIF values are below 5.1, which indicates that the covariance problem between independent variables entering the regression equation is not obvious. In addition, the standardized regression coefficients of mechanism research and development ability, transformation and application of mechanism results, green building design level, green building construction level, and green building operation level are 0.108, 0.158, 0.214, 0.263, and 0.134, respectively (the standardized regression coefficient of the degree of perfection of the mechanism standard is 0.063<0.1, therefore, it is not involved in the discussion) This coefficient is used to remove the influence of the data scale on the equation. This coefficient is used to remove the influence of the data outline on the equation, and its value reflects the influence of the independent variable on y. It can be seen that the importance of the influence of the five significant variables on the number of green building projects is in the order of the level of green building construction > level of green building design > level of transformation and application of the mechanism’s achievements > level of green building operation > mechanism research and development ability.

Generally can be taken to draw the predicted value of the dependent variable and the residuals of the scatterplot to carry out the test of chi-square, residuals histogram and cumulative probability plot can be derived directly from the results of the regression analysis in the SPSS software, at the same time it is the test of the regression model residuals whether to obey the normal distribution of the most intuitive way to check whether the results are as shown in Fig. 1 and Fig. 2.

Figure 1.

Regression normalized residual histogram

Figure 2.

Normalized residual P-P plot

From the histogram of regression standardized residuals in Figure 1, we can see that the standardized residuals basically obey the normal distribution, and from the standardized residuals P-P chart in Figure 2, we can see that most of the scatter points are distributed on the diagonal straight line, and we can assume that the residuals basically obey the normal distribution, and the model passes the test.

From the empirical results, it can be seen that the impact of five factors on sustainable development is more significant, namely, the system research and development ability, the mechanism of the results of the transformation and application, the level of green building design, the level of green building construction and the level of green building operation. Specifically: 1)

The effect of green building construction level on sustainable development is the most significant, the contribution to y among the significant variables is relatively large (standardized regression coefficient is 0.263), indicating that the level of green building construction has the greatest impact on sustainable development development, in the case of other variables remain unchanged, every 1% increase in the level of green building construction, the sustainable development of the construction of the construction of a subsequent increase of 0.354%.