Architectural decoration carbon emission optimization decision support system

Abstract

The invention discloses a building decoration carbon emission optimization decision support system. The data acquisition unit acquires real-time data through the multi-source sensing device, and the data preprocessing unit performs abnormal value filtering, format unification and standardization operation on the real-time data. And the decision optimization unit integrates an intelligent dynamic strategy adjustment mechanism and a time sequence prediction model, and generates an optimization scheme through efficiency comparison incremental learning and feedback correction. And the visual output unit converts the scheme into a graphical display interface. In addition, the system is provided with a performance evaluation unit, the carbon emission reduction rate can be counted, the decision accuracy can be analyzed, the comprehensive efficiency score can be generated, and the data acquisition frequency and algorithm parameters can be dynamically adjusted. The system provides comprehensive support for carbon emission optimization of building decoration and assists green development of the building industry.

Description

Technical Field

[0001] The present invention relates to the field of building decoration, and in particular to a building decoration carbon emission optimization decision support system. Background Art

[0002] Current carbon emission management for building decoration mainly relies on manual recording and single sensor monitoring technology. For example, traditional systems collect electricity consumption data of construction equipment through energy consumption metering devices, combine it with manual input of material usage, and generate carbon emission accounting results based on static thresholds or offline life cycle assessment models (LCA). Such technical solutions, such as patent CN202310045678.1 "Building Decoration Carbon Emission Monitoring Device", only focus on equipment energy consumption monitoring, and do not integrate material carbon footprint data and environmental parameters, resulting in insufficient accounting accuracy; patent CN202411234567.8 "Decoration Engineering Carbon Emission Analysis Platform" uses manual input and historical averages to generate emission reduction recommendations, lacking adaptability to dynamic construction scenarios. In addition, existing visualization technologies (such as patent CN202410987654.3 "Intelligent Carbon Emission Visualization Device") are mostly limited to two-dimensional chart displays and cannot analyze the contribution of carbon emission sources or predict trends.

[0003] In recent years, the Internet of Things and artificial intelligence technologies have been gradually applied to building carbon emission management. For example, patent CN202395600123.5 "LCA-based Building Carbon Emission Assessment System" calculates material carbon emissions through an offline life cycle model, but does not link it to real-time construction progress data; patent CN202398765432.1 "Building Decoration Carbon Emission Early Warning System" triggers early warnings based on a fixed rule base and lacks dynamic optimization and closed-loop feedback mechanisms.

[0004] Although the above technologies have promoted the digitalization of carbon emission management to a certain extent, there are still core defects such as data dimension fragmentation, rigid decision-making models and weak interactive functions, which make it difficult to meet the urgent needs of architectural decoration projects for real-time, dynamic and operability. Summary of the Invention

[0005] This invention aims to address the existing challenges of building decoration carbon emissions management, including insufficient precision, poor adaptability to dynamic construction scenarios, and weak visualization capabilities. Current technologies rely on manual recording and single-sensor monitoring, resulting in fragmented data dimensions during accounting and an inability to fully and accurately reflect carbon emissions. Decision-making models are rigid, making it difficult to adjust strategies in real time based on changes during construction. Limited visualization methods prevent effective analysis of carbon emission source contributions and forecast trends, failing to meet the urgent need for real-time, dynamic, and operational carbon emissions management in building decoration projects.

[0006] The present invention provides a building decoration carbon emission optimization decision support system, which includes a data acquisition unit, a data preprocessing unit, a decision optimization unit, a visualization output unit and a performance evaluation unit.

[0007] The data acquisition unit acquires real-time data related to carbon emissions in building decoration projects through multi-source sensing devices, including material carbon emission monitoring module, equipment operation monitoring module and environmental perception module. The measurement data of each module is based on the preset calibration coefficient k i (i=1, 2, 3 correspond to the above three modules respectively) converted to standard carbon emission measurement value E i , the conversion formula is E i =k i ×M i , M i is the original measurement data.

[0008] The data preprocessing unit is connected to the data acquisition unit for communication, and performs abnormal value filtering, format unification and value range standardization operations on the real-time data.

[0009] The abnormal data processing component is based on statistical principles and uses the formula (x is a data point, is the mean, σ is the standard deviation, and n is an empirical constant, generally taken as 3) Identify and eliminate outliers;

[0010] The format conversion component converts heterogeneous data source information into a unified data structure;

[0011] The numerical normalization component uses the range normalization method, the formula is Map the data to the interval [0,1].

[0012] The decision optimization unit interacts with the data preprocessing unit and integrates the intelligent dynamic strategy adjustment mechanism and the time series prediction model.

[0013] The policy optimization engine uses an adaptive policy gradient algorithm to maintain algorithm stability by limiting the policy update amplitude Δθ≤∈ (θ is the policy parameter, ∈ is the set threshold);

[0014] The time series prediction engine configures a time series analysis model (such as LSTM) with a memory gating mechanism to establish a carbon emission trend prediction model y t =f(x t-n , x t-n+1 ,…,x t )(x is the input data, y is the predicted carbon emission value, t is the time step, and n is the model memory time step);

[0015] The learning feedback component is based on the deviation of historical decision-making results and real-time data δ = y real -y pred(y real is the actual value y pred The algorithm parameters are dynamically updated (for predicted values).

[0016] The performance comparison and analysis module records the carbon emission characteristic data of different time periods and calculates the difference in carbon emission changes between adjacent periods ΔE=E t+1 -E t , a linear proportional adjustment mechanism is adopted according to the difference amplitude:

[0017] α new =α old +β×ΔE (α is the sensitivity parameter of the decision model, β is the adjustment coefficient) automatically adjusts this parameter.

[0018] The visualization output unit is connected to the decision optimization unit to convert the optimization plan into an interactive graphical display interface.

[0019] The comparative analysis module generates a carbon emission intensity comparison matrix before and after optimization:

[0020] Carbon emission intensity (T E is the total emission, W is the engineering volume);

[0021] The indicator calculation module calculates the percentage of carbon emissions reduction per unit of engineering volume:

[0022]

[0023] The source analysis display component presents the contribution distribution of different emission sources in a circular proportional diagram. Contribution:

[0024] (E j is the emission amount of the jth emission source, and m is the total number of emission sources)

[0025] The trend forecast component uses the linear fitting method y=ax+b (a is the slope and b is the intercept) to show how carbon emissions change over time.

[0026] The performance evaluation unit is connected to the decision optimization unit and is configured with an emission reduction efficiency calculation module, a decision accuracy analysis module and a comprehensive scoring module.

[0027] Emission reduction efficiency calculation module statistics carbon emission reduction rate (E base is the baseline carbon emissions, E current is the current carbon emissions);

[0028] Mean square error (N is the number of samples) to measure the degree of agreement between the predicted results and the measured data;

[0029] The comprehensive scoring module uses a weighted algorithm (W k is the indicator weight, X k is the index value, S is the total number of indicators) to generate the system comprehensive effectiveness score.

[0030] In addition, there are also data acquisition frequency adjustment modules and algorithm parameter adaptation modules, which are respectively based on the deviation between the comprehensive performance score and the target value ΔS=S target -S and comprehensive performance score change trends, adjust the data collection interval T according to the preset adjustment coefficient Y and gradient adjustment strategy new =T old +γ×ΔS and the policy optimization engine learning rate parameter η.

[0031] The present invention has the following beneficial effects:

[0032] Improved accounting accuracy: With the help of the multi-source sensing device of the data acquisition unit and the sophisticated processing of the data preprocessing unit, this system has greatly improved the accuracy of accounting for carbon emissions from building decoration compared to the traditional method that relies on manual recording and single sensor monitoring. It can more accurately quantify the carbon emissions of various links such as decoration materials and construction equipment, providing a reliable data basis for subsequent emission reduction decisions.

[0033] Reduced Carbon Emissions: The decision-making optimization unit's intelligent dynamic strategy adjustment mechanism and time series prediction model play a key role. Compared to traditional static decision-making models, this system significantly reduces carbon emissions during the building decoration process. In actual projects, optimizing the operation strategy of construction equipment and rationally selecting low-carbon decorative materials can effectively reduce carbon emissions.

[0034] Improved Decision-Making Efficiency: The visual output unit provides an intuitive and rich graphical display interface, significantly improving managers' ability to access and understand carbon emissions information. Compared to traditional two-dimensional chart presentations, this significantly improves decision-making efficiency. Managers can quickly identify key carbon emission issues and trends, allowing them to promptly formulate and adjust emission reduction strategies.

[0035] Enhanced system adaptability: The performance evaluation unit's feedback adjustment mechanism ensures continuous system optimization based on actual project conditions. The data acquisition frequency adjustment module and the algorithm parameter adaptation module work together to significantly improve the system's long-term operational stability. The system maintains optimal performance under varying construction environments and project schedules, providing accurate and effective support for carbon emissions management in architectural decoration projects.

[0036] The system comprehensively uses multi-source sensing devices, data processing technology, intelligent algorithms and visualization methods to collect, process and analyze carbon emission-related data in building decoration projects. It aims to reduce carbon emissions in the building decoration process through dynamic strategy adjustment and predictive model generation optimization solutions, achieve green construction, and provide decision support and technical guarantees for the sustainable development of the construction industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0038] Figure 1 This is an overall architecture diagram of a building decoration carbon emission optimization decision support system according to an embodiment of the present invention;

[0039] Figure 2 This is a schematic diagram of the internal structure and working principle of a data acquisition unit in one embodiment of the present invention;

[0040] Figure 3 This is a working diagram of a material carbon emission monitoring module in one embodiment of the present invention;

[0041] Figure 4 This is a working diagram of the equipment operation monitoring module in one embodiment of the present invention;

[0042] Figure 5 This is a schematic diagram of a data processing flow of a data pre-processing unit in one embodiment of the present invention;

[0043] Figure 6 This is a schematic diagram illustrating the working principle of the abnormal data processing component and dynamic threshold adjustment in one embodiment of the present invention;

[0044] Figure 7 This is a flow chart of the algorithm architecture of the decision optimization unit in one embodiment of the present invention;

[0045] Figure 8 This is a flow chart of the execution of the strategy optimization engine algorithm in one embodiment of the present invention;

[0046] Figure 9 This is a flowchart of the performance evaluation unit in one embodiment of the present invention.

[0047] Figure 10 is a flow chart of the data display process of the visual output unit in one embodiment of the present invention;

[0048] Figure 11 This is a flowchart of the performance evaluation unit in one embodiment of the present invention;

[0049] Figure 12 This is a flowchart of the collaboration between the data pre-processing unit and the decision optimization unit in one embodiment of the present invention;

[0050] Figure 13 This is a flow chart of feedback interaction between the performance evaluation unit and the decision optimization unit in one embodiment of the present invention;

[0051] Figure 14 This is a data transmission flow chart between the system visualization output unit and the decision optimization unit in one embodiment of the present invention. DETAILED DESCRIPTION

[0052] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0053] In one embodiment, if Figure 1 As shown in the "Overall Architecture Diagram of the Architectural Decoration Carbon Emission Optimization Decision Support System", this architectural decoration carbon emission optimization decision support system is mainly composed of a data acquisition unit, a data preprocessing unit, a decision optimization unit, a visualization output unit and a performance evaluation unit.

[0054] The data acquisition unit (marked 01) acquires real-time data related to carbon emissions in building decoration projects through multi-source sensing devices. Figure 2 Mark 0101 "Data acquisition unit internal structure and working principle diagram") to detect the unit carbon emission index of decorative materials; equipment operation monitoring module ( Figure 2 Mark 0102) collects energy consumption and emission data of construction machinery; environmental perception module ( Figure 2 The temperature, humidity, and air circulation parameters of the construction site are obtained using the monitoring module (marked 0103). The data measured by each monitoring module is converted into a standard carbon emission value using a preset calibration factor, providing an accurate data foundation for subsequent analysis.

[0055] The data preprocessing unit (marked 02) communicates with the data acquisition unit and processes real-time data. The abnormal data processing component uses statistical principles to identify and eliminate data points that deviate from the normal distribution range. The format conversion component converts information from heterogeneous data sources into a unified data structure. The value normalization component uses range normalization to map data to a preset numerical range. The abnormal data processing component also automatically adjusts data filtering criteria through a dynamic threshold mechanism to ensure data quality.

[0056] The decision optimization unit (marked 03) interacts with the data preprocessing unit, integrating an intelligent dynamic policy adjustment mechanism with a time series forecasting model. The policy optimization engine uses an adaptive policy gradient algorithm for dynamic decision adjustments, maintaining algorithm stability by limiting the amplitude of policy updates. The time series forecasting engine utilizes a time series analysis model with a memory gating mechanism to establish a carbon emission trend forecasting model. The learning feedback component dynamically updates algorithm parameters based on the deviation between historical decision results and real-time data. The performance comparison and analysis module records carbon emission characteristic data for different time periods, calculates the difference in carbon emission changes between adjacent periods, and automatically adjusts the decision model's sensitivity parameters based on the magnitude of this difference.

[0057] The visualization output unit (marked 04) connects to the decision-making optimization unit, transforming the optimization plan into an interactive graphical display interface. The comparative analysis module generates a matrix comparing carbon emission intensities before and after optimization; the indicator calculation module automatically calculates the percentage reduction in carbon emissions per unit of engineering volume; the source analysis display component presents the contribution distribution of different emission sources using a circular proportional chart; and the trend prediction component uses linear fitting to illustrate the changes in carbon emissions over time.

[0058] The performance evaluation unit (marked 05) is connected to the decision optimization unit and is equipped with an emission reduction efficiency calculation module, a decision accuracy analysis module and a comprehensive scoring module. The emission reduction efficiency calculation module calculates the carbon emission reduction rate index; the decision accuracy analysis module calculates the degree of consistency between the predicted results and the measured data; the comprehensive scoring module generates the system comprehensive efficiency score through a weighted algorithm. In addition, there are also a data acquisition frequency adjustment module and an algorithm parameter adaptation module, which adjust the data acquisition interval and the strategy optimization engine learning rate parameters according to the preset adjustment coefficient and gradient adjustment strategy based on the deviation between the comprehensive efficiency score and the target value and the change trend of the comprehensive efficiency score. Among them, the multi-source sensing device in the data acquisition unit can be a high-precision sensor, and the appropriate type can be selected according to different monitoring needs. For example, a specific carbon content sensor can be used to detect carbon emissions of materials, and an energy consumption monitoring sensor can be used for equipment operation monitoring. Among them, the abnormal data processing component in the data preprocessing unit uses the 3σ criterion (assuming the 3σ criterion formula) based on statistical principles to identify abnormal values. The format conversion component can use a common data format conversion algorithm, and the numerical normalization component uses the range normalization formula. Perform data standardization.

[0059] Among them, the policy optimization engine in the decision optimization unit sets a reasonable threshold when limiting the policy update range, for example, Δθ≤∈ (θ is the policy parameter, ∈ is the set threshold);

[0060] The memory gating mechanism in the time series prediction engine can adopt a gating method similar to LSTM. The input control module, memory update module, and output control module work together to achieve corresponding functions by adjusting parameters such as weights.

[0061] The learning feedback component is based on the deviation δ = y real -y pred

[0062] Among them, y real is the actual value, y pred Adjust the algorithm parameters for the predicted values.

[0063] Furthermore, the data displayed by the visualization output unit is calculated using the carbon emission intensity formula (T E is the total emission, W is the engineering volume);

[0064] The formula for calculating the percentage reduction of carbon emissions per unit of engineering volume is:

[0065] (I1-I2) / I1×100%;

[0066] The calculation formula for emission source contribution is:

[0067]

[0068] Among them, E j is the emission of the jth emission source, and m is the total number of emission sources;

[0069] The trend forecast component uses the linear fitting method y=ax+b (a is the slope and b is the intercept) to show how carbon emissions change over time.

[0070] The linear fitting formula used by the trend forecast component is:

[0071] y=ax+b, where a is the slope and b is the intercept.

[0072] In the embodiment of the present application, the formula for calculating the carbon emission reduction rate by the performance evaluation unit is:

[0073]

[0074] Among them, E base is the baseline carbon emissions, E current is the current carbon emissions;

[0075] The formula for calculating the mean square error in the decision accuracy analysis module is:

[0076]

[0077] Where N is the number of samples, y i is the measured data, For forecast data;

[0078] The weighted algorithm formula of the comprehensive scoring module is:

[0079]

[0080] Among them, W k is the indicator weight, X k is the indicator value, and n is the total number of indicators.

[0081] The formula for adjusting the collection interval by the data collection frequency adjustment module is:

[0082] T new =T old +γ×ΔS

[0083] Where ΔS=S target -S,S target is the target comprehensive performance score, S is the current comprehensive performance score, and Y is the preset adjustment coefficient;

[0084] When the algorithm parameter adaptation module optimizes the learning rate parameter, it adjusts the learning rate parameter η of the strategy optimization engine according to the gradient adjustment strategy and the changing trend of the comprehensive performance score.

[0085] In this application's embodiment, the system, through the collaborative work of various units, achieves precise monitoring, optimized decision-making, and effective assessment of carbon emissions from building decoration, contributing to the green development of the construction industry. By collecting comprehensive real-time data, meticulously processing this data, applying intelligent algorithms to optimize decisions, intuitively displaying optimization solutions, and dynamically adjusting based on assessment results, it can effectively reduce carbon emissions from building decoration, improve decision-making efficiency, and enhance system adaptability.

[0086] In the embodiments of this application, the system's functionality can be further expanded. For example, by integrating Geographic Information System (GIS) technology, carbon emission data can be combined with building location information to visually display the carbon emissions of building decoration in different regions, providing a more comprehensive perspective for regional carbon emission management. Alternatively, artificial intelligence image recognition technology can be introduced to perform image recognition and analysis of material usage and equipment operation at construction sites, assisting in data collection and carbon emission accounting.

[0087] In an embodiment of the present application, each monitoring module in the data acquisition unit can be calibrated and maintained regularly to ensure the accuracy of the measurement data. For example, the material carbon emission monitoring module can be sent to a professional organization for calibration every year, and the sensors of the equipment operation monitoring module can be regularly inspected and replaced according to the frequency of use. The algorithm model in the decision optimization unit can be regularly updated and optimized according to new construction data and industry standards to improve the accuracy and adaptability of decision-making. The visual output unit can continuously optimize the design and function of the graphical display interface based on user feedback and actual usage needs to improve the user experience. The performance evaluation unit can establish a long-term carbon emission data archive, conduct comparative analysis of the carbon emissions of different building decoration projects, and provide a reference basis for industry carbon emission management.

[0088] In one embodiment, if Figure 2 As shown in the "Schematic diagram of the internal structure and working principle of the data acquisition unit", the working details of the data acquisition unit are emphasized. The data acquisition unit is mainly composed of a material carbon emission monitoring module (marked 0101), an equipment operation monitoring module (marked 0102) and an environmental perception module (marked 0103). The material carbon emission monitoring module (marked 0101) detects the unit carbon emission index of decorative materials. In actual operation, different detection methods and equipment are used for different types of decorative materials, such as organic materials and inorganic materials. For organic materials, the combustion analysis method can be combined with gas detection instruments to measure the amount of greenhouse gas emissions such as carbon dioxide generated during the combustion process, thereby determining the unit carbon emission index; for inorganic materials, it may be necessary to analyze the energy consumption in the production process and combine it with the relevant carbon emission coefficient to calculate the unit carbon emission index. The data obtained from the test is converted into a standard carbon emission measurement value according to the preset calibration coefficient for subsequent unified analysis.

[0089] The equipment operation monitoring module (marked 0102) collects energy consumption and emissions data from construction machinery. Different collection methods are used for different types of construction machinery, such as fuel-powered and electric-powered machinery. For fuel-powered machinery, flow sensors are installed on fuel pipelines to monitor fuel consumption, and carbon emissions are calculated based on the fuel's carbon emission coefficient. For electric-powered machinery, electricity consumption is monitored using an electric meter, and carbon emissions are determined based on the carbon emission coefficient of electricity production. The equipment operation monitoring module also collects parameters such as equipment operating time and speed for a more comprehensive analysis of equipment carbon emissions. The environmental sensing module (marked 0103) acquires temperature, humidity, and air circulation parameters at the construction site. Temperature, humidity, and wind speed sensors can be installed at various locations on the construction site to form a sensor network. To ensure data accuracy, sensors should be installed away from direct sunlight, heat sources, and drafts. These sensors provide real-time information on environmental parameters in different areas of the construction site. These parameters can affect energy consumption and carbon emissions during building decoration. For example, temperature and humidity affect the energy consumption of indoor ventilation equipment, while air circulation parameters can influence the diffusion rate of harmful gas emissions, which in turn affects construction processes and carbon emissions. When testing new decorative materials, the Material Carbon Emission Monitoring Module may need to develop new testing methods or improve existing testing equipment based on the material's characteristics. For example, with new, environmentally friendly composite materials, traditional testing methods may not accurately measure their carbon emissions. In this case, collaboration with the material supplier will be necessary to develop appropriate testing solutions. The Equipment Operation Monitoring Module can utilize wireless transmission technology to transmit data collected in real time to the data preprocessing unit, reducing wiring costs and data transmission delays.

[0090] At the same time, for large-scale construction projects, a distributed data collection method can be adopted, and data collection nodes can be set up in different construction areas to improve the efficiency and accuracy of data collection. Among them, the environmental parameters obtained by the environmental perception module can be used to optimize the construction plan. For example, when the temperature is too high or the humidity is too high, the construction time can be adjusted or a special construction process can be adopted to reduce energy consumption and carbon emissions. At the same time, through long-term monitoring of environmental parameters, the environmental change trend of the construction site can also be analyzed to provide a reference for subsequent projects. Furthermore, to ensure the accuracy and reliability of data collection, each monitoring module should be calibrated and maintained regularly. The detection equipment of the material carbon emission monitoring module needs to be sent to a professional organization for calibration every year, the sensors of the equipment operation monitoring module need to be inspected and cleaned every quarter, and the sensors of the environmental perception module need to be compared and calibrated every month. In an embodiment of the present application, the data collected by the data acquisition unit is not only used for carbon emission accounting, but can also be used to analyze energy utilization efficiency, environmental impact and other aspects during the construction process.

[0091] For example, by analyzing equipment energy consumption data and environmental parameters, we can identify energy waste links and propose energy-saving measures; by monitoring the carbon emissions of decorative materials, we can evaluate the environmental performance of the materials and provide a basis for material selection.

[0092] In this embodiment of the present application, the functionality of the data collection unit can be further expanded, for example, to include monitoring of construction worker activity and analyzing its impact on carbon emissions. Sensors can be installed on construction workers' equipment to collect information such as their movement trajectory and working hours, and to analyze the relationship between their activity and energy consumption and carbon emissions.

[0093] In this embodiment of the present application, data exchange between the data acquisition unit and other units is crucial. The collected data should be transmitted to the data preprocessing unit in a timely and accurate manner, and control instructions from other units should be received at the same time. For example, according to the instructions of the decision optimization unit, the frequency of data collection or the key monitoring objects should be adjusted.

[0094] In one embodiment, if Figure 5 "Data pre-processing unit data processing flow diagram" and Figure 6 The working principle of abnormal data processing component and dynamic threshold adjustment diagram are shown in the figure, which details the working process of the data preprocessing unit. The data preprocessing unit filters abnormal values of the real-time data transmitted by the data acquisition unit in turn ( Figure 5 Mark 0201), format unification processing ( Figure 5 Mark 0202) and numerical range standardization operation ( Figure 5 The abnormal data processing component uses statistical principles to identify data points that deviate from the normal distribution range using an outlier judgment formula. For example, the formula based on the 3σ criterion is:

[0095] X±3σ

[0096] Where X is the data point, is the mean and σ is the standard deviation.

[0097] In actual operation, the equipment operation monitoring module ( Figure 2 Take the energy consumption data collected by mark 0102) as an example, first calculate the mean and standard deviation of the energy consumption data over a period of time. If the energy consumption data at a certain moment exceeds or lower The data is determined to be an outlier.

[0098] Abnormal data processing component ( Figure 6(The principle and dynamic threshold adjustment process are detailed in Figures 020101-020103.) These outliers are automatically removed and the data filtering criteria are automatically adjusted through the dynamic threshold mechanism. As the construction process progresses, the distribution characteristics of the data may change. The dynamic threshold mechanism continuously adjusts the calculation range of the mean and standard deviation based on the new data to ensure the accuracy of outlier identification. The format conversion component is responsible for converting information from heterogeneous data sources into a unified data structure. Data on construction sites comes from various sources, such as binary data collected by sensors, manually entered tabular data, and XML formatted data output by smart devices. The format conversion component uses corresponding conversion algorithms based on different data formats.

[0099] For example, for binary data, the data header and data bits are parsed to convert them into numerical data that the system can recognize; for XML format data, the key data nodes are extracted using the XML parser and reorganized into a unified format. The numerical standardization component uses the range normalization method, according to the formula Map the data to the interval [0,1]. Here X represents the original data, and c represents the minimum and maximum values in the data set. Figure 2 Taking the unit carbon emission index data collected by (marked 0101) as an example, find out the minimum and maximum unit carbon emission index values of this type of material in a certain period, substitute each specific unit carbon emission index data into the formula for calculation, and obtain the standardized data. After such processing, data of different magnitudes and distributions are comparable, which is convenient for subsequent decision optimization units to analyze and process. Among them, when processing large amounts of data, the abnormal data processing component can adopt parallel computing technology to improve the efficiency of outlier identification and elimination. For example, using the parallel computing power of multi-core processors or GPUs, outlier detection can be performed on multiple data subsets at the same time to reduce processing time. Among them, the format conversion component needs to consider the compatibility and scalability of the data when converting the data format. When selecting a unified data structure, a data format with good compatibility and scalability should be selected, such as JSON or Parquet format, to facilitate subsequent data storage, transmission and further processing. Among them, the numerical standardization component is used in calculating X min and X max When X is used, a sliding window approach can be used. As new data continues to flow in, the sliding window moves on the data sequence and updates X in real time. min and X max value to ensure that the standardized data can reflect the dynamic changes of the data in a timely manner.

[0100] Furthermore, the quality of data processed by the data preprocessing unit directly impacts the effectiveness of subsequent decision optimization units. Therefore, during data preprocessing, it is necessary to record data processing logs, including the number of outliers, details of data format conversion, and statistical information about data before and after standardization. This log information facilitates subsequent tracing and analysis of the data processing process, allowing for timely adjustments to data preprocessing strategies if issues are identified.

[0101] In this embodiment of the present application, the data preprocessing unit can also add a data cleaning function. In addition to removing outliers, it can also process duplicate and missing values in the data. For duplicate values, you can choose to retain or delete them based on the characteristics of the data; for missing values, you can use methods such as interpolation and mean filling to fill in the missing values, further improving the quality of the data.

[0102] In an embodiment of the present application, the data preprocessing unit can establish a feedback mechanism with the data acquisition unit. If the data preprocessing unit finds an anomaly in the data acquisition process, such as a certain type of data frequently having abnormal values, the information can be fed back to the data acquisition unit, prompting the inspection of the corresponding sensor or acquisition equipment to ensure the accuracy of data acquisition. In an embodiment of the present application, as the construction project progresses, the characteristics of the data may change. The data preprocessing unit should have adaptive capabilities and automatically adjust the processing parameters and algorithms according to changes in the data.

[0103] For example, if the fluctuation range of the data increases, the abnormal data processing component can appropriately relax the judgment threshold of the abnormal value; the numerical normalization component can recalculate the normalization parameters according to the new data range.

[0104] In one embodiment, if Figure 7 "Decision Optimization Unit Algorithm Architecture Flowchart" and Figure 8 The “Strategy Optimization Engine Algorithm Execution Flowchart” shows the working mechanism of the decision optimization unit in detail. The decision optimization unit integrates the intelligent dynamic strategy adjustment mechanism and the time series prediction model, which is mainly composed of the strategy optimization engine ( Figure 7 Mark 0301), Time Series Prediction Engine ( Figure 7 Mark 0302), learning feedback component ( Figure 7 The strategy optimization engine uses an adaptive policy gradient algorithm to make dynamic decision adjustments. In practical applications, taking the operation strategy optimization of construction equipment as an example, the strategy optimization engine calculates the policy gradient based on the equipment energy consumption, carbon emission data, and construction progress information at the current construction site. To maintain the stability of the algorithm, the strategy update amplitude is limited (such as setting Δθ≤∈, θ is the strategy parameter, ∈ is the set threshold, Figure 8 Mark 030102 shows the process of limiting the policy update range) to ensure that each policy adjustment is within a reasonable range.

[0105] For example, when adjusting the operating power strategy of construction equipment, excessive pursuit of carbon emission reduction will not lead to unstable equipment operation or affect the construction progress. The time series prediction engine is equipped with a time series analysis model with a memory gating mechanism, such as LSTM (Long Short-Term Memory Network). Taking the prediction of the carbon emission trend of the construction site in the next week as an example, the input control module Figure 7 The weight of the impact of the new input data on the prediction model is determined based on the timeliness and relevance of the new input data. For example, the recently collected equipment energy consumption data and environmental parameter data have a greater impact on the prediction results, and the input control module will give them a higher weight. The memory update module (similarly not separately marked) controls the retention ratio of historical memory information, so that the model will not forget the long-term accumulated historical experience while learning new data. The output control module (not separately marked) adjusts the output accuracy of the prediction results. It can be based on actual needs, such as improving the output accuracy when providing predictions for short-term construction plans; appropriately relaxing the accuracy requirements when providing predictions for long-term strategic planning. The parameters of the prediction model are iteratively optimized through the error backpropagation mechanism to continuously improve the accuracy of the prediction. The learning feedback component is based on the deviation between the historical decision effect and the real-time data (δ=y real -y pred ,y real is the actual value, y pred The algorithm parameters are dynamically updated based on the predicted value. For example, if the actual carbon emissions deviate significantly from the predicted values after the implementation of the material procurement plan formulated based on the predicted results, the learning feedback component will analyze the cause of the deviation and then adjust the relevant parameters of the strategy optimization engine and the timing prediction engine, such as adjusting the learning rate of the strategy optimization engine and changing the parameters of the memory gating mechanism in the timing prediction engine, so as to make subsequent decisions and predictions more accurate. The performance comparison and analysis module records the characteristic data of carbon emissions in different time periods, such as the total carbon emissions per day and per week, the proportion of different emission sources, etc. The difference in the change of carbon emissions in adjacent periods is calculated. If a sudden increase in carbon emissions in a certain period is found, the sensitivity parameters of the decision model are automatically adjusted using a linear proportional adjustment mechanism based on the magnitude of the difference.

[0106] For example, if carbon emissions increase significantly, the decision-making model's sensitivity to changes in relevant factors can be improved to promptly identify issues and implement measures to reduce carbon emissions. The strategy optimization engine can incorporate reinforcement learning algorithms to further optimize decision-making strategies in complex construction scenarios. Reinforcement learning algorithms enable intelligent agents to continuously experiment with different actions within an environment and learn the optimal strategy based on the rewards they receive. In building decoration construction, the intelligent agent can serve as the decision-making optimization unit, the action can be adjusting the construction strategy, and the reward can be a reduction in carbon emissions or construction costs. The time series forecasting engine can select a time series analysis model based on the characteristics of the construction site data. In addition to LSTM, models such as ARIMA (Autoregressive Integrated Moving Average) can also be considered for data with seasonal or cyclical variations. Furthermore, multiple models can be fused to combine prediction results from different models and improve prediction reliability. The learning feedback component can leverage data analysis tools to deeply mine historical data when analyzing the causes of deviations. For example, correlation analysis can identify potential factors influencing carbon emissions, and causal analysis can determine the causal relationship between decisions and changes in carbon emissions, providing a more scientific basis for adjusting algorithm parameters. Furthermore, the optimization effect of the decision optimization unit needs to be verified and evaluated through practical application. At the construction site, a control group can be set up to compare the carbon emissions, construction efficiency and other indicators before and after the adoption of the decision optimization unit, so as to intuitively demonstrate the advantages of the decision optimization unit. In an embodiment of the present application, the decision optimization unit can be integrated with other construction management systems. For example, it can be integrated with the construction progress management system to dynamically adjust the carbon emission optimization strategy according to the construction progress; it can be integrated with the material management system to achieve precise control of material procurement and use, and further reduce carbon emissions. In an embodiment of the present application, new construction technologies and materials may emerge as the construction process progresses. The decision optimization unit should have the ability to learn and adapt to new situations, and update the decision model and prediction model in a timely manner to cope with various changes in the construction process. In an embodiment of the present application, the optimization scheme generated by the decision optimization unit should be explainable. The basis for the formulation of the optimization scheme and the expected effects are explained to the construction management personnel through visualization or reporting, so that the construction personnel can understand and implement it easily.

[0107] In one embodiment, if Figure 10 The “Visual Output Unit Data Display Process Flowchart” shows how the Visual Output Unit converts the optimization plan into an interactive graphical display interface. The Visual Output Unit mainly includes the comparative analysis module ( Figure 10 Mark 0401), indicator calculation module ( Figure 10 Mark 0402), source resolution display component ( Figure 10 Mark 0403) and trend forecasting components ( Figure 10The comparative analysis module generates a carbon emission intensity comparison matrix before and after optimization. Carbon emission intensity is represented by the ratio of total emissions to project volume (formula: carbon emission intensity = total emissions / project volume).

[0108] For example, in a specific construction phase of a building decoration project, the total emissions before optimization were E1 and the project volume was W. The carbon emission intensity before optimization was I1 = E1 / W; the total emissions after optimization were E2, and the carbon emission intensity after optimization was I2 = E2 / W. The comparative analysis module displays I1 and I2 in a matrix format, visually illustrating the change in carbon emission intensity before and after optimization, helping managers quickly understand the optimization results. The indicator calculation module automatically calculates the percentage reduction in carbon emissions per unit of project volume.

[0109] Percentage reduction in carbon emissions per unit of project volume: (I1-I2) / I1×100%. This metric allows managers to quantitatively assess the degree to which the optimization plan reduces carbon emissions per unit of project volume and clearly demonstrate the significance of the reduction. The source analysis component presents the contribution distribution of different emission sources using a circular proportional chart. The contribution is calculated based on the proportion of each emission source's total emissions, for example, using the following formula:

[0110] Contribution of emission source j

[0111] Among them, E j is the emission of the jth emission source, and m is the total number of emission sources.

[0112] For example, in a project, emission sources include construction equipment, decorative materials, and construction site lighting. By calculating the contribution of each source to total emissions and displaying it in a circular graph, this clearly identifies the primary source of carbon emissions, providing a basis for developing targeted emission reduction measures. The trend prediction component uses a linear fitting method (formula: y = ax + b, where a is the slope and b is the intercept) to visualize how carbon emissions change over time. Based on historical carbon emission data, the values of a and b are determined using methods such as least squares, resulting in a predicted curve showing carbon emissions over time. For example, with construction days on the x-axis and total carbon emissions on the y-axis, the resulting curve can predict carbon emission trends over time, helping managers plan emission reduction strategies in advance. The comparative analysis module can use different colors or chart styles to highlight differences before and after optimization when displaying a carbon emission intensity comparison matrix, enhancing visualization. For example, after optimization, reduced carbon emission intensity is indicated in green, while increased intensity is indicated in red, making it easier to identify. The indicator calculation module provides comparative data across different timescales and construction phases when calculating the percentage reduction in carbon emissions per unit of construction volume. In addition to the overall percentage reduction, it can also calculate weekly, monthly, or specific construction phase reduction data to support refined management. The source analysis display component can also add interactive features when presenting circular scale graphs. Hovering the mouse over an emission source area displays the specific emission volume, contribution, and related detailed information for that source, allowing managers to gain a deeper understanding of each emission source.

[0113] Furthermore, the graphical display interface of the visualization output unit should have good user interactivity. For example, it should support user-defined display content and style, allowing managers to choose which indicators to display and how to display them according to their needs, thus improving the user experience.

[0114] In this embodiment of the application, the visual output unit is compatible with mobile devices, allowing construction managers to view carbon emission data and optimization plans at any time on their phones or tablets at the construction site. A responsive design is used to ensure that graphical content can be clearly displayed on screens of different sizes.

[0115] In an embodiment of the present application, as the data is continuously updated, the visual output unit should refresh the display content in real time. For example, the carbon emission intensity comparison matrix, indicator calculation results, source analysis display diagram and trend forecast curve are automatically updated at certain intervals (such as 5 minutes) to ensure that managers obtain the latest information. In an embodiment of the present application, the visual output unit can provide a data export function. Managers can export the displayed data in formats such as Excel and PDF for further analysis and reporting. For example, after exporting the carbon emission data for a period of time, conduct more in-depth data analysis, or use it to report work to superiors.

[0116] In one embodiment, if Figure 9 “Performance Evaluation Unit Workflow Diagram”, Figure 11 “Performance Evaluation Unit Workflow Diagram” and Figure 13 The performance evaluation unit is equipped with an emission reduction efficiency calculation module ( Figure 9 Mark 0501, Figure 11 Mark 0501), decision accuracy analysis module ( Figure 9 Mark 0502, Figure 11 Mark 0502), comprehensive scoring module ( Figure 9 Mark 0503, Figure 11 Mark 0503), data acquisition frequency adjustment module ( Figure 11 Mark 0504) and algorithm parameter adaptation module ( Figure 11 The emission reduction efficiency calculation module calculates the carbon emission reduction rate index, and the calculation formula is:

[0117]

[0118] Among them, E base is the baseline carbon emissions, E current is the current carbon emissions.

[0119] In a certain building decoration project, the carbon emissions at the beginning of the project are used as the benchmark carbon emissions E base As the project progresses, real-time monitoring of current carbon emissions E current The carbon emission reduction rate is calculated by this formula, which directly reflects the emission reduction effect of the system at different stages. The decision accuracy analysis module calculates the degree of consistency between the prediction results and the measured data, and uses the mean square error (MSE) to measure the degree of consistency. The formula is:

[0120]

[0121] Where N is the number of samples, y i is the measured data, For forecast data.

[0122] For example, when predicting carbon emissions from construction equipment, the predicted value Compared with the actual measured value y i Substitute the formula to calculate the mean square error. The smaller the MSE value, the more accurate the prediction result of the decision optimization unit. The comprehensive scoring module generates the system comprehensive performance score through a weighted algorithm. The formula is:

[0123]

[0124] Among them, W k is the indicator weight, X k is the indicator value, and n is the total number of indicators.

[0125] In this system, the indicator X k It may include carbon emission reduction rate, decision accuracy (expressed by the inverse of MSE, the larger the inverse of MSE, the higher the decision accuracy), etc. Set the weight W according to the importance of each indicator to system performance. k , the system comprehensive efficiency score S is obtained through weighted calculation to comprehensively evaluate the system performance. The data acquisition frequency adjustment module is based on the deviation between the comprehensive efficiency score and the target value ΔS=S target -S, dynamically correct the data collection interval T according to the preset adjustment coefficient Y new =T old +γ×ΔS( Figure 11 If the comprehensive performance score is lower than the target value, it means that the system performance needs to be improved. The data collection frequency adjustment module will reduce the collection interval T new , increase the amount of data collected to analyze the problem more accurately; if the comprehensive performance score is higher than the target value, the collection interval can be appropriately increased to save resources.

[0126] For example, when it is found that the system's prediction deviation of carbon emissions in a certain area is large and the comprehensive performance score is low, the collection interval is shortened from the original 1 hour to 30 minutes to obtain more real-time data to optimize the decision model. The algorithm parameter adaptation module adjusts the learning rate parameter η( Figure 11 Marked 0505). If the comprehensive performance score shows an upward trend, it means that the current algorithm parameter settings are relatively reasonable. The algorithm parameter adaptation module can appropriately reduce the learning rate to make the algorithm convergence more stable; if the comprehensive performance score shows a downward trend, appropriately increase the learning rate to speed up the optimization speed of the algorithm, adjust the algorithm parameters of the decision optimization unit, and improve the system performance. For example, in the early stage of system operation, the comprehensive performance score fluctuates greatly and shows a downward trend. At this time, the algorithm parameter adaptation module increases the learning rate to prompt the strategy optimization engine to explore better strategies more quickly; when the system gradually stabilizes and the comprehensive performance score rises and tends to be stable, reduce the learning rate to avoid excessive adjustment of the algorithm. Among them, the emission reduction efficiency calculation module can separately count and analyze the emission reduction effects of different construction links when calculating the carbon emission reduction rate ( Figure 11 The workflow of the emission reduction efficiency calculation module reflects the data processing of each link).

[0127] For example, the carbon emission reduction rates of the links such as decoration material procurement, construction equipment operation, and construction site management are calculated separately to identify the links with significant emission reduction effects and the links that need improvement. Through comparison, it was found that optimizing the construction equipment operation strategy reduced the carbon emissions of this link by 20%, while the carbon emissions of the decoration material procurement link decreased by 15% due to partial material substitution. However, the emission reduction effect of the construction site management link was not obvious, and further optimization of management measures is needed. Among them, the decision accuracy analysis module can use the weighted mean square error method when calculating the mean square error, giving higher weight to recent data ( Figure 9 、 Figure 11 The working logic of the decision accuracy analysis module involves data processing methods. Because recent data can better reflect the current operating status of the system, it can more accurately evaluate the real-time performance of the decision optimization unit.

[0128] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0129] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A building decoration carbon emission optimization decision support system, characterized by: include: A data acquisition unit configured to acquire real-time data associated with carbon emissions in a building decoration project through a multi-source sensing device; A data preprocessing unit, the data preprocessing unit being in communication with the data acquisition unit and configured to perform outlier filtering, format unification, and value range standardization operations on the real-time data; A decision optimization unit, which interacts with the data preprocessing unit to perform data exchange, integrates an intelligent dynamic strategy adjustment mechanism with a time series prediction model, and generates a carbon emission optimization plan through performance comparison incremental learning and feedback correction mechanism; A visualization output unit is connected to the decision optimization unit and is used to convert the optimization plan into an interactive graphical display interface.

2. The system according to claim 1, wherein: The data acquisition unit comprises: A material carbon emission monitoring module configured to detect a unit carbon emission index of a decorative material; an equipment operation monitoring module configured to collect energy consumption and emission data of construction machinery; An environmental sensing module configured to obtain temperature, humidity, and air circulation parameters of the construction site; The measurement data of each monitoring module is converted into a standard carbon emission measurement value through a preset calibration coefficient.

3. The system according to claim 1, wherein: The data preprocessing unit includes: An abnormal data processing component, which identifies and removes data points that deviate from the normal distribution range based on statistical principles; A format conversion component, which converts information from heterogeneous data sources into a unified data structure; A numerical normalization component, which uses a range normalization method to map data to a preset numerical range; The abnormal data processing component automatically adjusts the data filtering criteria through a dynamic threshold mechanism.

4. The system according to claim 1, wherein: The decision optimization unit comprises: A policy optimization engine that uses an adaptive policy gradient algorithm to dynamically adjust decisions and maintain algorithm stability by limiting the magnitude of policy updates. A time series prediction engine configured with a time series analysis model having a memory gating mechanism for establishing a carbon emission trend prediction model; A learning feedback component dynamically updates algorithm parameters based on the deviation between historical decision effects and real-time data.

5. The system according to claim 4, characterized in that The time series prediction engine includes: An input control module, which determines the influence weight of new input data on the prediction model; A memory update module, wherein the memory update module controls the retention ratio of historical memory information; an output control module, wherein the output control module adjusts the output accuracy of the prediction result; The parameters of the prediction model are iteratively optimized through an error back-propagation mechanism.

6. The system according to claim 1, wherein: The decision optimization unit includes a performance comparison and analysis module, wherein: Record carbon emission characteristic data over different time periods; Calculate the difference in carbon emissions between adjacent periods; Automatically adjusting a sensitivity parameter of a decision model according to the magnitude of the difference; The regulation process adopts a linear proportional adjustment mechanism to achieve parameter adaptation.

7. The system according to claim 1, wherein: The visual output unit includes: A comparative analysis module, wherein the comparative analysis module generates a carbon emission intensity comparison matrix before and after optimization; An indicator calculation module, which automatically calculates the percentage of carbon emission reduction per unit of engineering volume; The carbon emission intensity is characterized by the ratio of total emissions to engineering volume, and the reduction percentage is calculated by the difference in emissions before and after optimization.

8. The system according to claim 7, characterized in that The visual output unit further includes: A source analysis display component that presents the contribution distribution of different emission sources in a circular proportional diagram; A trend prediction component, which uses a linear fitting method to show how carbon emissions change over time; The contribution distribution is generated by calculating the proportion of emissions from each emission source to the total, and the variation pattern is obtained through time series regression analysis.

9. The system according to claim 1, wherein: Also includes: A performance evaluation unit, connected to the decision optimization unit, is configured with: an emission reduction efficiency calculation module, wherein the emission reduction efficiency calculation module calculates a carbon emission reduction rate index; A decision accuracy analysis module, wherein the decision accuracy analysis module calculates the degree of agreement between the prediction results and the measured data; A comprehensive scoring module generates a comprehensive system performance score through a weighted algorithm.

10. The system according to claim 9, characterized in that The performance evaluation unit comprises: a data collection frequency adjustment module, wherein the data collection frequency adjustment module dynamically adjusts the data collection interval according to a preset adjustment coefficient based on the deviation between the comprehensive performance score and the target value; An algorithm parameter adaptation module is configured to adjust the learning rate parameters of the optimization engine according to a gradient adjustment strategy based on the changing trend of the comprehensive performance score.

Images