An intelligent assessment management system and method for road construction carbon emission monitoring

Abstract

The present invention discloses an intelligent assessment management system and method for carbon emission monitoring in road construction, and specifically relates to the technical field of carbon emission monitoring in road construction: by real-time acquisition and analysis of carbon emission data in the construction process, combining historical data to identify abnormal carbon emissions, and predicting the impact of calculation deviations on the accuracy of carbon emission assessment based on a machine learning model, it is possible to effectively identify and classify time periods with different degrees of impact, especially conduct in-depth analysis and predictive early warning on time periods with medium impact, and timely discover potential cumulative effects of calculation deviations, thereby avoiding the risk of cumulative deviations being discovered only at the end of the project; this intelligent assessment management method improves the accuracy of carbon emission assessments, ensures that the project meets environmental protection standards and sustainable development goals, and reduces the difficulty of later corrections caused by accumulated deviations.

Description

Technical Field

[0001] The present invention relates to the technical field of road construction carbon emission monitoring, and in particular to an intelligent assessment management system and method for road construction carbon emission monitoring. Background Art

[0002] Road construction carbon emission monitoring refers to tracking, recording and analyzing the emission of greenhouse gases such as carbon dioxide generated by construction activities during the road construction process. The monitoring process usually includes measuring carbon emissions in construction machinery, material transportation, material production and other links, and evaluating their impact on the environment. Through carbon emission monitoring, the main sources of emissions can be identified and corresponding emission reduction measures can be formulated to achieve a more environmentally friendly and sustainable construction process.

[0003] During road construction, carbon emission assessment is key to ensuring that the project meets environmental standards and sustainable development goals. However, small calculation deviations may occur during the assessment process. These deviations may seem insignificant at a time, but if they are not discovered and corrected in time, they may accumulate throughout the construction cycle, causing the total carbon emission assessment to deviate significantly from the actual value. At the same time, this cumulative effect may not be discovered until the end of the project, resulting in the failure to achieve the overall environmental protection goals of the project, and since the investment of time and resources has become an established fact, it becomes almost impossible to correct these deviations. Summary of the invention

[0004] The purpose of the present invention is to provide an intelligent assessment management system and method for road construction carbon emission monitoring to overcome the shortcomings of the background technology.

[0005] In order to achieve the above object, the present invention provides the following technical solution: an intelligent assessment and management method for road construction carbon emission monitoring, comprising the following steps:

[0006] S1: Obtain carbon emission data during the construction process from several sources in real time, compare the real-time carbon emission data with historical data, and determine abnormal carbon emission data based on the comparison results;

[0007] S2: Divide the abnormal carbon emission data into K different time periods in chronological order, and extract training data of the abnormal carbon emission data in each time period, wherein the training data includes deviation time training data and high-frequency fluctuation training data;

[0008] S3: Based on the deviation time training data and high-frequency fluctuation training data, a machine learning model is trained to predict the impact of calculation deviation on the accuracy of the overall carbon emission assessment, and the impact degree of calculation deviation in each time period is determined according to the model output results;

[0009] S4: Compare the calculated deviation impact degree values ​​obtained in each time period with the gradient standard threshold, and divide the corresponding time period into a high impact degree time period, a medium impact degree time period, and a low impact degree time period according to the comparison result;

[0010] S5: For the medium impact period, further analysis will be conducted on the impact of calculation deviations on the accuracy of the overall carbon emission assessment, and early prediction and warning of abnormal cumulative effect events of calculation deviations will be conducted based on the analysis results.

[0011] Preferably, in S1, the carbon emission data acquired in real time is compared with the historical data, and the abnormal carbon emission data is determined according to the comparison result, specifically:

[0012] Real-time data of the current time period is extracted from the sensors and monitoring systems, and historical data of the corresponding time period is extracted from the database. The difference between the real-time data and the historical data is calculated. The difference calculation formula is: difference = real-time data − average value of historical data. A difference of ±5% is defined as the normal range. If the difference exceeds the set threshold range, it is judged as abnormal carbon emission data.

[0013] Preferably, in S2, the deviation time training data is a deviation time anomaly index, and the method for obtaining the deviation time anomaly index is:

[0014] Acquire abnormal carbon emission data in each time period in real time, describe the normal distribution and abnormal distribution of carbon emission data, calculate the prior probability that the data point is normal according to the proportion of normal data in historical data, and calculate the prior probability that the data point is abnormal, P(abnormal)=1−P(normal);

[0015] Calculate observed carbon emissions data when the data points are normal The probability of is expressed as: ; In the formula, μ is the mean of historical normal carbon emission data, is the variance of historical normal carbon emission data, is the ratio of pi. Similarly, in the case of abnormal data points, the observed carbon emission data is calculated The probability of ; Use Bayes' theorem to calculate data points is the posterior probability of anomaly , the expression is: ;in, It is observed data The marginal probability of is calculated as: ; Obtain the posterior probability through the Bayesian formula , the expression is: ; Based on the posterior probability, the deviation time anomaly index is calculated, and the calculation expression is: ; is the abnormal threshold of the posterior probability, is the deviation time anomaly index.

[0016] Preferably, in S2, the high-frequency fluctuation training data is a high-frequency fluctuation change index, and the method for obtaining the high-frequency fluctuation change index is:

[0017] The carbon emission data is sampled at uniform time intervals and marked as , where t=1,2,…,N represents the time point, determines the window function w(τ) and selects the window size L, and applies the window function w(τ) slidingly to the carbon emission data For each window position t, multiply it by the window function w(τ) and perform Fourier transform. The result of STFT is calculated for each time point t. The expression is: ; In the formula, represents the Fourier transform result of frequency f at time t, At the point in time The carbon emission data value at , f is the frequency variable, j is the imaginary unit, and the critical frequency of the high-frequency interval is determined , for frequencies above the critical frequency Calculate the square of the amplitude of all frequency components and add them up to get the energy of the high-frequency part, that is, calculate the high-frequency fluctuation change index, the expression is: ; is the amplitude of the STFT result, that is, the modulus of S(t,f), It is the high frequency fluctuation index.

[0018] Preferably, in S3, a machine learning model for predicting the impact of calculation deviation on the accuracy of overall carbon emission assessment is trained based on deviation time training data and high-frequency fluctuation training data, and the calculation deviation impact value in each time period is determined according to the model output result;

[0019] The high-frequency fluctuation change index and the deviation time anomaly index are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the calculation deviation impact value label in each time period as the prediction target, and takes minimizing the sum of prediction errors of all calculation deviation impact value labels in the time period as the training target. The machine learning model is trained until the sum of prediction errors converges and the model training is stopped. The calculation deviation impact value in each time period is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0020] Preferably, in S4, the calculated deviation influence degree value obtained in each time period is compared with the gradient standard threshold, and the corresponding time period is divided into influence degrees according to the comparison result;

[0021] The calculated deviation influence degree value in each time period obtained is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the calculated deviation influence degree value in each time period is compared with the first standard threshold and the second standard threshold respectively;

[0022] If the calculated deviation impact value within the time period is greater than the second standard threshold, it is divided into a high impact time period; if the calculated deviation impact value within the time period is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it is divided into a medium impact time period; if the calculated deviation impact value within the time period is less than the first standard threshold, it is divided into a low impact time period.

[0023] Preferably, in S5, for the medium impact period, the impact of the calculation deviation on the accuracy of the overall carbon emission assessment is further analyzed, and an early prediction and warning of the abnormal cumulative effect of the calculation deviation is carried out according to the analysis results;

[0024] When the calculated deviation impact value in a time period is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it is divided into a medium impact time period, and the divided medium impact time period is further divided into several sub-time periods. The generated calculation deviation impact value is analyzed section by section, where a=1,2,…,n represents the number of the sub-time period, and n is a positive integer greater than 0. The corresponding moving average , and form a moving average sequence, the moving average The calculation formula is: ; where k is each sub-time period The window size is , b is the number of the sub-time period, and the difference value of the moving average sequence is calculated. The expression is: ; is the difference value of the moving average series.

[0025] Preferably, if )>0, it means that at time point Moving average relative to the previous time point Increase, that is, the impact of calculation deviation on the accuracy of overall carbon emission assessment is increasing, at this time, an early warning signal is generated to remind relevant personnel to check and intervene; if <0, it means that the moving average is decreasing. There is no need to generate an early warning signal immediately and it should be monitored continuously.

[0026] The present invention also provides an intelligent assessment and management system for road construction carbon emission monitoring, including a data acquisition module, a training data extraction module, a calculation deviation assessment module, an impact degree classification module and a prediction and early warning module;

[0027] Data acquisition module: obtains carbon emission data during the construction process from several sources in real time, compares the real-time carbon emission data with historical data, and determines abnormal carbon emission data based on the comparison results;

[0028] Training data extraction module: divide the abnormal carbon emission data into K different time periods according to the time sequence, and extract the training data of the abnormal carbon emission data in each time period, wherein the training data includes deviation time training data and high-frequency fluctuation training data;

[0029] Calculation deviation assessment module: Based on deviation time training data and high-frequency fluctuation training data, a machine learning model is trained to predict the impact of calculation deviation on the accuracy of overall carbon emission assessment, and the calculation deviation impact value in each time period is determined according to the model output results;

[0030] Impact degree division module: compares the calculated deviation impact degree values ​​obtained in each time period with the gradient standard threshold, and divides the corresponding time period into high impact degree time period, medium impact degree time period and low impact degree time period according to the comparison result;

[0031] Prediction and warning module: For the medium impact period, the impact of calculation deviation on the accuracy of the overall carbon emission assessment will be further analyzed, and based on the analysis results, early prediction and warning of abnormal cumulative effect events of calculation deviation will be carried out.

[0032] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0033] 1. The present invention accurately identifies abnormal carbon emission data by comparing real-time data acquisition with historical data, and uses deviation time training data and high-frequency fluctuation training data to build a machine learning model, so as to accurately predict the impact of calculation deviations in each time period on the overall carbon emission assessment. Not only can it effectively classify and manage time periods with different degrees of impact, but it can also timely predict and warn of potential cumulative effects of calculation deviations through in-depth analysis of time periods with medium impact and monitoring of moving average trends in sub-time periods, so as to prevent the accumulation of assessment deviations from causing inaccurate final assessment results.

[0034] 2. The present invention not only improves the accuracy and timeliness of carbon emission assessment, but also ensures that the project can respond to abnormal situations in a timely manner and avoid deviations from environmental protection goals due to cumulative deviations. This systematic and intelligent monitoring and management method can help construction projects better achieve environmental protection and sustainable development goals, while reducing the time and cost required to correct deviations in the later stage, and improving the overall management efficiency and effectiveness of the project. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0036] Figure 1 The figure is a flow chart of the method of the present invention.

[0037] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] Example 1, please refer to Figure 1 As shown, the intelligent assessment and management method for road construction carbon emission monitoring described in this embodiment includes the following steps:

[0040] S1: Obtain carbon emission data during the construction process from several sources in real time, compare the real-time carbon emission data with historical data, and determine abnormal carbon emission data based on the comparison results;

[0041] S2: Divide the abnormal carbon emission data into K different time periods in chronological order, and extract training data of the abnormal carbon emission data in each time period, wherein the training data includes deviation time training data and high-frequency fluctuation training data;

[0042] S3: Based on the deviation time training data and high-frequency fluctuation training data, a machine learning model is trained to predict the impact of calculation deviation on the accuracy of the overall carbon emission assessment, and the impact degree of calculation deviation in each time period is determined according to the model output results;

[0043] S4: Compare the calculated deviation impact degree values ​​obtained in each time period with the gradient standard threshold, and divide the corresponding time period into a high impact degree time period, a medium impact degree time period, and a low impact degree time period according to the comparison result;

[0044] S5: For the medium impact period, further analysis will be conducted on the impact of calculation deviations on the accuracy of the overall carbon emission assessment, and early prediction and warning of abnormal cumulative effect events of calculation deviations will be conducted based on the analysis results.

[0045] Among them, in S1, carbon emission data during the construction process is obtained in real time from several sources, and the real-time carbon emission data is compared with historical data, and abnormal carbon emission data is determined based on the comparison results.

[0046] Determine the data source: including sensors installed on mechanical equipment to monitor fuel consumption, equipment operating status, exhaust gas concentration, etc. Environmental sensors placed at the construction site to collect external environmental parameters such as temperature, humidity, wind speed, air pressure, etc., which affect carbon emissions. Vehicle GPS and fuel sensors: used to track the driving routes and fuel consumption of construction vehicles. Construction logs and manual input: manually input data such as construction operation logs, equipment maintenance records, and personnel operation records.

[0047] Real-time sensor data collection: Using the Internet of Things (IoT) technology, real-time data from various sensors is transmitted to a central data platform via wireless networks. Low-power wide area networks (LPWAN) or cellular networks (such as 4G / 5G) can be used for data transmission.

[0048] Data transmission and storage: Once the data is collected, it is immediately transmitted to the cloud or local server for storage. Each piece of data is marked with a timestamp to ensure the time series integrity of the data. Data from different sources are integrated to form a unified real-time carbon emissions data set.

[0049] Historical carbon emission data includes historical data of similar construction projects, past data of the same construction phase, historical performance data of equipment operation, etc., which are stored in the database. Clean the historical data, remove outliers and invalid data, and perform standardization to ensure data consistency. Align the real-time data with the historical data based on the timestamp, and compare them with the historical data in the same time period as the benchmark. Ensure that the real-time data and historical data are compared on the same dimension, such as comparing data on the same equipment and under the same environmental conditions.

[0050] Extract real-time data for the current time period from sensors or monitoring systems, such as fuel consumption or carbon emission concentration for the current hour or day.

[0051] Historical data for the corresponding time period is extracted from the database, usually the fuel consumption or carbon emission concentration under the same historical conditions, such as data from the same day and the same time period in the past week.

[0052] Calculate the difference between real-time data and historical data. The difference calculation formula is: difference = real-time data − historical data average value. For example, if the carbon emission concentration in the current hour is 100ppm, and the average value under the same conditions in history is 90ppm, then the difference is: difference = 100ppm − 90ppm = 10ppm. The difference can directly reflect whether the current carbon emissions or fuel consumption exceeds or is lower than the historical average.

[0053] According to the fluctuation range of historical data, the threshold range is set, and the difference of ±5% is defined as the normal range. If the difference exceeds the set threshold range, it is judged as abnormal carbon emission data. For example, if the set normal threshold is ±5%, and the currently detected percentage difference is 10%, it is considered that the current carbon emissions have significant deviations, and further analysis or measures may be required.

[0054] S2: Divide the abnormal carbon emission data into K different time periods in chronological order, and extract training data of the abnormal carbon emission data in each time period, wherein the training data includes deviation time training data and high-frequency fluctuation training data.

[0055] Fixed time interval division: According to the characteristics of the construction cycle, the entire data sequence is divided into K time periods according to fixed time intervals. For example, it can be divided by time units such as hours, days, weeks, or months. If the collection cycle is one day and it is desired to be divided into 24 hour segments, each hour is a time period.

[0056] Dynamic time interval division: The time period can be adjusted dynamically according to the actual fluctuation of the data. For example, the time period can be flexibly divided according to the density of abnormal events or the rate of change of carbon emissions. Time periods with dense data or violent fluctuations can be divided into smaller ones, while relatively stable time periods can be divided into longer ones.

[0057] Select timeline: Determine the timeline you want to divide, such as by day, hour, or custom time interval.

[0058] Divide data: Divide the abnormal carbon emission data into K time periods according to the selected time axis sequence. For example, if you choose to divide by hour and want to divide the data for one day, you will get 24 time periods, corresponding to the data for each hour.

[0059] Verify the partitioning: Check whether the data points in each time period are evenly distributed to ensure that no data is missed or too concentrated in a certain time period.

[0060] Deviation time refers to the time point when abnormal carbon emission data deviates significantly from historical data. The purpose of extracting deviation time training data is to identify the timing and duration of abnormal events. The deviation time training data is the deviation time anomaly index. The method for obtaining the deviation time anomaly index is:

[0061] Acquire abnormal carbon emission data in each time period in real time, describe the normal distribution and abnormal distribution of carbon emission data, calculate the prior probability that the data point is normal according to the proportion of normal data in historical data, for example, if 95% of the data in historical data is normal, then P(normal)=0.95; calculate the prior probability that the data point is abnormal, P(abnormal)=1−P(normal);

[0062] Calculate observed carbon emissions data when the data points are normal The probability of is expressed as: ; In the formula, μ is the mean of historical normal carbon emission data, is the variance of historical normal carbon emission data, is the ratio of pi. Similarly, in the case of abnormal data points, the observed carbon emission data is calculated The probability of ; Use Bayes' theorem to calculate data points is the posterior probability of anomaly , the expression is: ;in, It is observed data The marginal probability of is calculated as: ; Obtain the posterior probability through the Bayesian formula , the expression is: ; Based on the posterior probability, the deviation time anomaly index is calculated, and the calculation expression is: ; is the abnormal threshold of the posterior probability, is the deviation time anomaly index.

[0063] The larger the deviation time anomaly index, the more significant the impact of the calculation deviation on the accuracy of the overall carbon emissions assessment. In other words, the more serious the deviation of the data point from the historical normal value, or the longer the anomaly lasts, the more likely this deviation will lead to significant errors in the overall carbon emissions assessment results, thereby affecting the reliability of the assessment and the effectiveness of the decision. Therefore, a large deviation time anomaly index is a warning signal, indicating that measures need to be taken to correct or further analyze the source of the deviation.

[0064] High-frequency fluctuations refer to frequent changes in abnormal carbon emission data in various time periods, which may indicate equipment instability, improper operation, or external environmental interference. High-frequency fluctuation training data is used to identify system instability.

[0065] Wherein, the high-frequency fluctuation training data is a high-frequency fluctuation change index, and the method for obtaining the high-frequency fluctuation change index is:

[0066] The carbon emission data is sampled at uniform time intervals and marked as , where t=1,2,…,N represents the time point, determines the window function w(τ) and selects the window size L, and applies the window function w(τ) slidingly to the carbon emission data For each window position t, multiply it by the window function w(τ) and perform Fourier transform. The result of STFT is calculated for each time point t. The expression is: ; In the formula, represents the Fourier transform result of frequency f at time t, At the point in time The carbon emission data value at , f is the frequency variable, j is the imaginary unit, and the critical frequency of the high-frequency interval is determined , depending on the data characteristics and analysis requirements. Usually, the frequency is higher than The part above the critical frequency is regarded as high frequency component. Calculate the square of the amplitude (ie, energy) of all frequency components and add them up to get the energy of the high-frequency part, that is, calculate the high-frequency fluctuation change index, the expression is: ; is the amplitude of the STFT result, that is, the modulus of S(t, f), It is the high frequency fluctuation index.

[0067] The larger the high-frequency fluctuation change index, the greater the impact of calculation deviation on the accuracy of the overall carbon emissions assessment. This means that frequent and drastic fluctuations in the data may lead to instability and unreliability in the assessment results, increasing the difficulty and error of the overall carbon emissions forecast. High-frequency fluctuations usually reflect potential problems in the system, such as equipment instability or external interference. Therefore, a larger high-frequency fluctuation change index is an important risk indicator in the assessment process, indicating that measures need to be taken to reduce the impact of these fluctuations on the assessment results.

[0068] S3: Based on the deviation time training data and high-frequency fluctuation training data, a machine learning model is trained to predict the impact of calculation deviation on the accuracy of the overall carbon emission assessment, and the impact degree of the calculation deviation in each time period is determined according to the model output results.

[0069] The high-frequency fluctuation change index and the deviation time anomaly index are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the calculation deviation impact value label in each time period as the prediction target, and takes minimizing the sum of prediction errors of all calculation deviation impact value labels in the time period as the training target. The machine learning model is trained until the sum of prediction errors converges and the model training is stopped. The calculation deviation impact value in each time period is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0070] The method for obtaining the calculation deviation impact value in each time period is as follows: from the comprehensive feature vector training data of the trained machine learning model, obtain the corresponding function expression: ; Where G is the output function of the model, is the high-frequency fluctuation index, is the deviation time anomaly index, The calculated deviation impact value for each time period.

[0071] S4: Compare the calculated deviation impact degree values ​​obtained in each time period with the gradient standard threshold, and divide the corresponding time period into high impact degree time period, medium impact degree time period and low impact degree time period according to the comparison result.

[0072] The calculated deviation influence degree value in each time period obtained is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the calculated deviation influence degree value in each time period is compared with the first standard threshold and the second standard threshold respectively;

[0073] If the calculation deviation impact value within a time period is greater than the second standard threshold, it is considered that the calculation deviation within the time period has a significant impact on the accuracy of the overall carbon emission assessment and is classified as a high impact time period;

[0074] If the calculation deviation impact value within a time period is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it is considered that the calculation deviation within the time period has a certain impact on the accuracy of the overall carbon emission assessment, but the impact is medium, and it is divided into a medium impact time period;

[0075] If the impact value of the calculation deviation within a time period is less than the first standard threshold, it is considered that the calculation deviation within the time period has little impact on the overall carbon emission assessment and is classified as a low impact time period.

[0076] S5: For the medium impact period, further analysis will be conducted on the impact of calculation deviations on the accuracy of the overall carbon emission assessment, and early prediction and warning of abnormal cumulative effect events of calculation deviations will be conducted based on the analysis results.

[0077] When the calculated deviation impact value in a time period is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it is divided into a medium impact time period, and the divided medium impact time period is further divided into several sub-time periods. For example, if the medium impact time period is 1 hour, it can be divided into a sub-time period every 5 minutes. The generated calculation deviation impact value is analyzed section by section, where a=1, 2, ..., n represents the number of the sub-time period, and n is a positive integer greater than 0. The calculation of each sub-time period The corresponding moving average , and form a moving average sequence, the moving average The calculation formula is: ; where k is each sub-time period The window size is usually a smaller value, such as 3 or 5. b is the number of the sub-time period. The difference value of the moving average sequence is calculated. The expression is: ; is the difference value of the moving average series;

[0078] if )>0, it means that at time point Moving average relative to the previous time point Increase, that is, the impact of calculation deviation on the accuracy of overall carbon emission assessment is increasing, at this time, an early warning signal is generated to remind relevant personnel to check and intervene; if <0, it means that the moving average is decreasing. There is no need to generate an early warning signal immediately and it should be monitored continuously.

[0079] In this application, by continuously monitoring the calculation deviation impact value and its moving average difference in each sub-time period, it is ensured that the upward trend can be captured in time and an early warning signal can be generated. When an early warning signal is generated, the system can automatically trigger the corresponding response mechanism, such as notifying management personnel or adjusting construction strategies, to prevent the accumulation of calculation deviations from causing inaccurate evaluation results.

[0080] In this embodiment, carbon emission data in the construction process is obtained in real time from several sources, and the real-time obtained carbon emission data is compared with historical data, and abnormal carbon emission data is determined according to the comparison result; the abnormal carbon emission data is divided in chronological order, divided into K different time periods, and training data of abnormal carbon emission data in each time period is extracted respectively, and the training data includes deviation time training data and high-frequency fluctuation training data; a machine learning model for predicting the influence of calculation deviation on the accuracy of overall carbon emission assessment is trained based on deviation time training data and high-frequency fluctuation training data, and the influence degree value of calculation deviation in each time period is determined according to the output result of the model; the influence degree value of calculation deviation in each time period obtained is compared with the gradient standard threshold, and the corresponding time period is divided into high influence degree time period, medium influence degree time period and low influence degree time period according to the comparison result; for the medium influence degree time period, the influence degree of calculation deviation on the accuracy of overall carbon emission assessment is further analyzed, and the calculation deviation abnormal cumulative effect event is predicted and warned in advance according to the analysis result. It not only improves the accuracy of carbon emission monitoring, but also enhances the responsiveness of the system, which helps to better support environmental protection decision-making and construction management.

[0081] Example 2, please refer to Figure 2 As shown, the intelligent assessment and management system for road construction carbon emission monitoring described in this embodiment includes a data acquisition module, a training data extraction module, a calculation deviation assessment module, an impact degree classification module and a prediction and warning module;

[0082] Data acquisition module: obtains carbon emission data during the construction process from several sources in real time, compares the real-time carbon emission data with historical data, and determines abnormal carbon emission data based on the comparison results;

[0083] Training data extraction module: divide the abnormal carbon emission data into K different time periods according to the time sequence, and extract the training data of the abnormal carbon emission data in each time period, wherein the training data includes deviation time training data and high-frequency fluctuation training data;

[0084] Calculation deviation assessment module: Based on deviation time training data and high-frequency fluctuation training data, a machine learning model is trained to predict the impact of calculation deviation on the accuracy of overall carbon emission assessment, and the calculation deviation impact value in each time period is determined according to the model output results;

[0085] Impact degree division module: compares the calculated deviation impact degree values ​​obtained in each time period with the gradient standard threshold, and divides the corresponding time period into high impact degree time period, medium impact degree time period and low impact degree time period according to the comparison result;

[0086] Prediction and warning module: For the medium impact period, the impact of calculation deviation on the accuracy of the overall carbon emission assessment will be further analyzed, and based on the analysis results, early prediction and warning of abnormal cumulative effect events of calculation deviation will be carried out.

[0087] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0088] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0089] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0090] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0091] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0092] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. An intelligent assessment and management method for carbon emission monitoring in road construction, characterized by: The steps include: S1: Obtain carbon emission data during the construction process from several sources in real time, compare the real-time carbon emission data with historical data, and determine abnormal carbon emission data based on the comparison results; S2: Divide the abnormal carbon emission data into K different time periods in chronological order, and extract training data of the abnormal carbon emission data in each time period, wherein the training data includes deviation time training data and high-frequency fluctuation training data; The step S2 specifically includes: the deviation time training data is a deviation time anomaly index, and the method for obtaining the deviation time anomaly index is: Acquire abnormal carbon emission data in each time period in real time, describe the normal distribution and abnormal distribution of carbon emission data, calculate the prior probability that the data point is normal according to the proportion of normal data in historical data, and calculate the prior probability that the data point is abnormal, P(abnormal)=1−P(normal); Calculate observed carbon emissions data when the data points are normal The probability of is expressed as: ; In the formula, μ is the mean of historical normal carbon emission data, is the variance of historical normal carbon emission data, is the ratio of pi. Similarly, in the case of abnormal data points, the observed carbon emission data is calculated The probability of ; Use Bayes' theorem to calculate data points is the posterior probability of anomaly , the expression is: ;in, It is observed data The marginal probability of is calculated as: ; Obtain the posterior probability through the Bayesian formula , the expression is: ; Based on the posterior probability, the deviation time anomaly index is calculated, and the calculation expression is: ; is the abnormal threshold of the posterior probability, is the deviation time anomaly index; The high-frequency fluctuation training data is a high-frequency fluctuation change index, and the method for obtaining the high-frequency fluctuation change index is: The carbon emission data is sampled at uniform time intervals and marked as , where t=1,2,…,N represents the time point, determines the window function w(τ) and selects the window size L, and applies the window function w(τ) slidingly to the carbon emission data For each window position t, multiply it by the window function w(τ) and perform Fourier transform. The result of STFT is calculated for each time point t. The expression is: ; In the formula, represents the Fourier transform result of frequency f at time t, At the point in time The carbon emission data value at , f is the frequency variable, j is the imaginary unit, and the critical frequency of the high-frequency interval is determined , for frequencies above the critical frequency Calculate the square of the amplitude of all frequency components and add them up to get the energy of the high-frequency part, that is, calculate the high-frequency fluctuation change index, the expression is: ; is the amplitude of the STFT result, that is, the modulus of S(t,f), It is the high-frequency fluctuation index; S3: Based on the deviation time training data and high-frequency fluctuation training data, a machine learning model is trained to predict the impact of calculation deviation on the accuracy of the overall carbon emission assessment, and the impact degree of calculation deviation in each time period is determined according to the model output results; S4: Compare the calculated deviation impact degree values ​​obtained in each time period with the gradient standard threshold, and divide the corresponding time period into a high impact degree time period, a medium impact degree time period, and a low impact degree time period according to the comparison result; S5: For the medium impact period, further analyze the impact of calculation deviation on the accuracy of overall carbon emission assessment, and make early predictions and warnings for abnormal cumulative effect events of calculation deviation based on the analysis results; The step S5 specifically includes: when the calculated deviation impact value in a time period is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, dividing it into a medium impact time period, and further dividing the divided medium impact time period into several sub-time periods, The generated calculation deviation impact value is analyzed section by section, where a=1,2,…,n represents the number of the sub-time period, and n is a positive integer greater than 0. The corresponding moving average , and form a moving average sequence, the moving average The calculation formula is: ; where k is each sub-time period The window size is , b is the number of the sub-time period, and the difference value of the moving average sequence is calculated. The expression is: ; is the difference value of the moving average series; if )>0, it means that at time point Moving average relative to the previous time point Increase, that is, the impact of calculation deviation on the accuracy of overall carbon emission assessment is increasing, at this time, an early warning signal is generated to remind relevant personnel to check and intervene; if <0, it means that the moving average is decreasing. There is no need to generate an early warning signal immediately and it should be monitored continuously.

2. The intelligent assessment and management method for road construction carbon emission monitoring according to claim 1 is characterized in that: In S1, the real-time carbon emission data is compared with the historical data, and the abnormal carbon emission data is determined according to the comparison result, specifically: Extract the real-time data of the current time period from the sensor and monitoring system, extract the historical data of the corresponding time period from the database, and calculate the difference between the real-time data and the historical data. The difference calculation formula is: ; A difference of ±5% is defined as the normal range. If the difference exceeds the set threshold range, it is judged as abnormal carbon emission data.

3. The intelligent assessment and management method for road construction carbon emission monitoring according to claim 1 is characterized in that: In S3, a machine learning model is trained based on the deviation time training data and the high-frequency fluctuation training data to predict the impact of the calculation deviation on the accuracy of the overall carbon emission assessment, and the calculation deviation impact value in each time period is determined according to the model output results; The high-frequency fluctuation change index and the deviation time anomaly index are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the calculation deviation impact value label in each time period as the prediction target, and takes minimizing the sum of prediction errors of all calculation deviation impact value labels in the time period as the training target. The machine learning model is trained until the sum of prediction errors converges and the model training is stopped. The calculation deviation impact value in each time period is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

4. The intelligent assessment and management method for road construction carbon emission monitoring according to claim 3 is characterized by: In S4, the calculated deviation influence degree value obtained in each time period is compared with the gradient standard threshold, and the corresponding time period is divided into influence degrees according to the comparison result; The calculated deviation influence degree value in each time period obtained is compared with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and the calculated deviation influence degree value in each time period is compared with the first standard threshold and the second standard threshold respectively; If the calculated deviation impact value within the time period is greater than the second standard threshold, it is divided into a high impact time period; if the calculated deviation impact value within the time period is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it is divided into a medium impact time period; if the calculated deviation impact value within the time period is less than the first standard threshold, it is divided into a low impact time period.

5. An intelligent assessment and management system for road construction carbon emission monitoring, used to implement the intelligent assessment and management method for road construction carbon emission monitoring according to any one of claims 1 to 4, characterized in that: It includes data acquisition module, training data extraction module, calculation deviation assessment module, impact degree classification module and prediction and early warning module; Data acquisition module: obtains carbon emission data during the construction process from several sources in real time, compares the real-time carbon emission data with historical data, and determines abnormal carbon emission data based on the comparison results; Training data extraction module: divide the abnormal carbon emission data into K different time periods according to the time sequence, and extract the training data of the abnormal carbon emission data in each time period, wherein the training data includes deviation time training data and high-frequency fluctuation training data; Calculation deviation assessment module: Based on deviation time training data and high-frequency fluctuation training data, a machine learning model is trained to predict the impact of calculation deviation on the accuracy of overall carbon emission assessment, and the calculation deviation impact value in each time period is determined according to the model output results; Impact degree division module: compares the calculated deviation impact degree values ​​obtained in each time period with the gradient standard threshold, and divides the corresponding time period into high impact degree time period, medium impact degree time period and low impact degree time period according to the comparison result; Prediction and warning module: For the medium impact period, the impact of calculation deviation on the accuracy of the overall carbon emission assessment will be further analyzed, and based on the analysis results, early prediction and warning of abnormal cumulative effect events of calculation deviation will be carried out.

Images