Carbon emission monitoring method, system and equipment based on multi-source heterogeneous data

Abstract

The invention provides a carbon emission monitoring method, system and device based on multi-source heterogeneous data, and relates to the technical field of smart power grids, the method comprises the following steps: obtaining carbon emission accounting basic data of a target device in a predetermined time period, the carbon emission accounting basic data comprising energy consumption data and fuel consumption data; extracting a time sequence of energy consumption data in the carbon emission accounting basic number of the target equipment, and identifying abnormal energy consumption data from the time sequence based on the energy consumption data at the quartile position in each sliding window in the time sequence; repairing the abnormal energy consumption data to obtain repaired energy consumption data; and based on the repair energy consumption data, the normal energy consumption data in the time sequence and the fuel consumption data, determining the carbon emission in the predetermined time period. Therefore, the accuracy and credibility of carbon emission monitoring are improved.

Description

Technical Field

[0001] This application relates to the field of carbon emission monitoring technology, and more specifically, to a carbon emission monitoring method and system based on multi-source heterogeneous data. Background Technology

[0002] With increasingly severe global climate change, excessive emissions of greenhouse gases (especially carbon dioxide) have become the core of a series of ecological crises. Frequent extreme weather events, rising sea levels, and ecosystem imbalances are escalating, threatening human survival and sustainable development. Therefore, monitoring carbon emissions is crucial.

[0003] However, current mainstream carbon emission monitoring technologies still face many technical bottlenecks, among which the stability of equipment operation and the reliability of sensor data affect the accuracy of carbon emission monitoring. During equipment operation, instantaneous high energy consumption points are common. For example, a sudden drop in motor speed can cause a surge in energy consumption, while bearing wear can increase operating resistance and cause abnormal fluctuations in energy consumption. Simultaneously, sensors, as the core components for data acquisition, are susceptible to external environmental interference and their own aging, leading to problems such as data jumps and zero-value drift. These anomalies at the equipment and sensor levels directly cause distortion in the collected energy consumption data, resulting in discrepancies between the carbon emissions calculated based on energy consumption data and the actual emissions, thus lowering the accuracy and reliability of carbon emission monitoring. Summary of the Invention

[0004] This application provides a carbon emission monitoring method, system, and device based on multi-source heterogeneous data, which solves the problem of low accuracy and reliability in carbon emission monitoring.

[0005] Firstly, this application provides a carbon emission monitoring method based on multi-source heterogeneous data. The risk screening method includes the following steps: acquiring the basic carbon emission accounting data of the target equipment within a predetermined time period, the basic carbon emission accounting data including energy consumption data and fuel consumption data; extracting the time series of energy consumption data from the basic carbon emission accounting data of the target equipment, and identifying abnormal energy consumption data from the time series based on the energy consumption data at the quartile position within each sliding window in the time series; repairing the abnormal energy consumption data to obtain repaired energy consumption data; and determining the carbon emission amount within the predetermined time period based on the repaired energy consumption data, the normal energy consumption data in the time series, and the fuel consumption data.

[0006] Secondly, this application provides a carbon emission monitoring system based on multi-source heterogeneous data, including an acquisition module for acquiring basic carbon emission accounting data of a target device within a predetermined time period, the basic carbon emission accounting data including energy consumption data and fuel consumption data; an identification module for extracting the time series of energy consumption data from the basic carbon emission accounting data of the target device, and identifying abnormal energy consumption data from the time series based on the energy consumption data at the quartile position within each sliding window in the time series; a repair module for repairing the abnormal energy consumption data to obtain repaired energy consumption data; and a determination module for determining the carbon emissions within the predetermined time period based on the repaired energy consumption data, the normal energy consumption data in the time series, and the fuel consumption data.

[0007] Thirdly, this application provides an electronic device, including: a processor and a memory; wherein the memory is used to store a computer program that can run on the processor; the processor is used to execute the program stored in the memory to implement the steps of the carbon emission monitoring method based on multi-source heterogeneous data mentioned in the first aspect.

[0008] The technical solution provided by this invention employs a dynamic calculation mode using a sliding window in the time-series of energy consumption data. This mode adapts to the diurnal periodicity and shift fluctuations of energy consumption data, improving the accuracy of identifying abnormal energy consumption data. Furthermore, the anomaly identification method based on quartile positions effectively identifies extreme outliers (such as zero-value drift or instantaneous peaks caused by sensor malfunctions). Instead of directly discarding abnormal energy consumption data, the identified data is repaired, thus improving the accuracy of subsequent carbon emission calculations. Finally, when calculating carbon emissions, the system comprehensively repairs three types of multi-source data: energy consumption data, normal energy consumption data, and fuel consumption data. This covers both the accurate quantification of indirect energy consumption such as electricity and heat, and the statistics of direct fossil fuel consumption, ensuring a complete dimension for carbon emission calculation. Therefore, the carbon emissions calculated using this method are closer to actual emissions, improving the accuracy and reliability of carbon emission monitoring. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this embodiment of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart of a carbon emission monitoring method based on multi-source heterogeneous data provided in this application;

[0011] Figure 2This is a schematic diagram of a carbon emission monitoring system based on multi-source heterogeneous data provided in this application.

[0012] Figure 3 This is a schematic diagram of the structure of an electronic device provided in this application. Detailed Implementation

[0013] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0014] To better understand the above technical solutions, a detailed description of the technical solutions will be provided below in conjunction with the accompanying drawings and specific embodiments. (Refer to...) Figure 1 As shown in the figure, this is an exemplary flowchart of a carbon emission monitoring method based on multi-source heterogeneous data according to this embodiment of the present application. The carbon emission monitoring method based on multi-source heterogeneous data includes the following steps:

[0015] In step S101, the basic data for carbon emission accounting of the target equipment within a predetermined time period are obtained.

[0016] The basic data for carbon emission accounting includes energy consumption data and fuel consumption data.

[0017] Specifically, embodiments of the present invention interface with a power monitoring system to collect real-time energy consumption data and with a fuel system to obtain fuel consumption data. Furthermore, embodiments of the present invention also extract production output and raw material consumption data from the production system and deploy edge computing nodes to collect operating parameters of target equipment, including but not limited to temperature and pressure.

[0018] Furthermore, the predetermined time period can be a complete production cycle. Target equipment includes, but is not limited to, electrical equipment and fuel system equipment. Electrical equipment includes, but is not limited to, high-power motors, heating furnaces, etc., while fuel system equipment includes, but is not limited to, coal feeding systems for coal-fired boilers and oil supply systems for fuel-fired equipment.

[0019] Furthermore, energy consumption data includes, but is not limited to, cumulative electricity consumption, total power consumption, and instantaneous energy consumption data (such as instantaneous electricity consumption). Fuel consumption data refers to the fuel consumption of the target equipment per unit time.

[0020] In step S102, the time series of energy consumption data in the carbon emission accounting basis of the target device is extracted, and abnormal energy consumption data is identified from the time series based on the energy consumption data at the quartile position in each sliding window of the time series.

[0021] Specifically, to facilitate the standardization of energy consumption data in different units and thus improve computing speed, this embodiment of the invention can convert energy consumption data in different units into ton CO2 equivalents. For example, 1 kWh = 0.785 kg CO2, thereby converting energy consumption data in different units into energy consumption data in a unified unit.

[0022] Furthermore, this embodiment of the invention extracts a time-series sequence of energy consumption data from the target device, spanning at least one complete production cycle. The time-series sequence can be a data sequence formed by sorting the energy consumption data of the target device according to time. Then, obviously invalid data, such as negative and zero values, are removed from the time-series sequence, because energy consumption cannot be negative or zero under normal conditions. Due to sensor sensitivity issues, data gaps may also exist in the time-series sequence; this embodiment of the invention records the missing data locations as NaN. The energy consumption data in the processed time-series sequence are sorted in chronological order. This embodiment of the invention uses a sliding window method to process the time-series features. The sliding window size is set to 24 data points corresponding to 15-minute data points per day, with a sliding step of 1 data point to ensure the capture of short-term fluctuations.

[0023] Furthermore, as an optional embodiment of the present invention, identifying abnormal energy consumption data from a time series based on energy consumption data at the quartile positions within each sliding window includes: arranging the energy consumption data within the sliding window in ascending order of numerical value; determining the first energy consumption data located at the upper quartile position and the second energy consumption data located at the lower quartile position within the sliding window, where the upper quartile position is the 25th percentile position and the lower quartile position is the 75th percentile position; calculating the interquartile range of the sliding window based on the second energy consumption data and the first energy consumption data; determining an upper limit and a lower limit value based on the interquartile range, the first energy consumption data, and the second energy consumption data; and marking energy consumption data within the sliding window that exceeds the upper limit and lower limit value as abnormal energy consumption data.

[0024] Specifically, this embodiment of the invention employs the Interquartile Range (IQR) algorithm to identify anomalous energy consumption data in a time series. First, this embodiment sorts the energy consumption data within each sliding window in ascending order of value. Then, it calculates the first energy consumption data Q1 located at the upper quartile position and the second energy consumption data Q3 located at the lower quartile position within the sliding window. Here, Q1 refers to the value at the 25th percentile after sorting the energy consumption data in the sliding window in ascending order, i.e., the n×25th percentile data, where n is the number of valid energy consumption data points within the sliding window. Q3 refers to the value at the 75th percentile after sorting the energy consumption data in the sliding window in ascending order, i.e., the n×75th percentile data.

[0025] Furthermore, in this embodiment of the invention, the interquartile range of the sliding window is calculated using the following formula:

[0026] IQR = Q3 - Q1

[0027] In the above formula, IQR represents the interquartile range of the sliding window. Q3 refers to the value at the 75th percentile after the energy consumption data within the sliding window is sorted in ascending order. Q1 refers to the value at the 25th percentile after the energy consumption data within the sliding window is sorted in ascending order.

[0028] Furthermore, as an optional embodiment of the present invention, determining the upper and lower limits based on the interquartile range, the first energy consumption data, and the second energy consumption data includes: calculating a first product between a predetermined multiple and the interquartile range; determining the difference between the first energy consumption data and the first product as the lower limit; and determining the sum between the second energy consumption data and the first product as the upper limit.

[0029] Specifically, the predetermined multiple can be selected based on actual circumstances. In this embodiment of the invention, the value is 1.5. The upper and lower limits are calculated using the following formula:

[0030] Lower = Q1 - 1.5 × IQR

[0031] Upper = Q3 + 1.5 × IQR

[0032] In the above formula, Lower represents the lower limit value. Upper represents the upper limit value. IQR represents the interquartile range of the sliding window. Q3 refers to the value at the 75th percentile after the energy consumption data within the sliding window is sorted in ascending order. Q1 refers to the value at the 25th percentile after the energy consumption data within the sliding window is sorted in ascending order.

[0033] Furthermore, in this embodiment of the invention, energy consumption data exceeding the range [Lower, Upper] is marked as abnormal energy consumption data. This embodiment of the invention targets the instantaneous high energy consumption caused by equipment failure, and therefore mainly focuses on abnormal energy consumption data exceeding Upper.

[0034] Furthermore, as an optional embodiment of the present invention, after marking energy consumption data exceeding the upper and lower limits within the sliding window as abnormal energy consumption data, the method further includes: if the abnormal energy consumption data is a single-point sudden increase, it is determined that the target device has a momentary fault, the single-point sudden increase indicates that only one energy consumption data exceeds the upper limit, and the preceding and following energy consumption data of the abnormal energy consumption data are both within the range of the upper and lower limits; if the abnormal energy consumption data is multiple consecutive data sudden increases, it is determined that the target device has a continuous fault, the multiple consecutive data sudden increases indicate that multiple consecutive energy consumption data exceed the upper limit.

[0035] Specifically, if the abnormal value is a single point of sudden increase, such as only one data point exceeding the upper limit, and the data before and after are normal, it is judged as a momentary equipment failure, such as a short-term overload caused by motor jamming. If three or more consecutive data points exceed the upper limit, it is necessary to check whether it is a continuous failure, such as valve leakage, in conjunction with the equipment operation log. In such cases, it is not directly repaired, but rather the equipment maintenance warning is triggered.

[0036] In step S103, the abnormal energy consumption data is repaired to obtain repaired energy consumption data.

[0037] Specifically, the method addresses the location of abnormal energy consumption data within a time series. If the abnormal energy consumption data is located between the beginning and end of the time series, this embodiment employs a weighted interpolation method based on the trend of normal data preceding and following the abnormal energy consumption data to repair it. If the abnormal energy consumption data is located at the beginning or end of the time series, it is filled using the data following the beginning or the data preceding the end.

[0038] Furthermore, as an optional embodiment of the present invention, repairing abnormal energy consumption data to obtain repaired energy consumption data includes: if the abnormal energy consumption data is located between the beginning and end of a time series, then the preceding and following energy consumption data of the abnormal energy consumption data are obtained from the time series, and the preceding and following energy consumption data of the abnormal energy consumption data are weighted to obtain the repaired energy consumption data of the abnormal energy consumption data; if the abnormal energy consumption data is located at the beginning or end of a time series, then the abnormal energy consumption data is repaired using the following energy consumption data of the beginning to obtain the repaired energy consumption data, or the abnormal energy consumption data is repaired using the preceding energy consumption data of the end to obtain the repaired energy consumption data.

[0039] Specifically, the embodiments of the present invention employ the following formula to repair abnormal energy consumption data, thereby obtaining repaired energy consumption data:

[0040] Repairing energy consumption data = Previous data point of abnormal energy consumption data × 0.6 + Next data point of abnormal energy consumption data × 0.4

[0041] In this embodiment, 0.6 and 0.4 are weights. Since energy consumption changes are usually continuous, the weights are skewed towards the preceding data in abnormal energy consumption data. Therefore, the weight of the preceding data is higher than that of the following data. Of course, the weights can be adjusted according to the actual situation, and this embodiment of the invention does not limit this.

[0042] Furthermore, if abnormal energy consumption data is located at the beginning or end of the time series, it is filled with the nearest neighbor value. For example, if the abnormal data is at the beginning, it is filled with the next data, and if the abnormal data is at the end, it is filled with the previous data.

[0043] Furthermore, after identifying anomalous energy consumption data from the time series based on the energy consumption data at the quartile positions within each sliding window in the time series, the method further includes: identifying missing data positions from each sliding window; determining the energy consumption data preceding and following the missing data position from the time series, as well as the sort number of the target position and the number of positions in the interval between the missing data positions; and determining the repair energy consumption data for the missing data position based on the energy consumption data preceding and following the missing data position, the sort number of the target position, and the number of positions in the interval.

[0044] Specifically, in this embodiment of the invention, for NaN values, a linear interpolation method is used. If the energy consumption data before and after the missing position are a and b, with an interval of k points, then the energy consumption data for repairing the missing data position is a + (ba) × (i / k), where i is the position of the missing point in the interval.

[0045] Furthermore, embodiments of the present invention can also calculate the standard deviation of the corrected time series composed of the repaired energy consumption data and compare it with the standard deviation of the historical normal cycle to determine the validity of the repaired energy consumption data. If the deviation between the standard deviation of the corrected time series and the standard deviation of the historical normal cycle is ≤5%, it proves that the repaired energy consumption data is valid, ensuring that the repaired data conforms to the energy consumption fluctuation law of the equipment.

[0046] For example, in order to provide a detailed description of the process for repairing abnormal energy consumption data, the following specific embodiments are provided in this invention:

[0047] First, the original data sequence (sorted chronologically) in this embodiment of the invention is: [5.2, 5.4, 5.3, 28.7, 5.5, NaN, 5.1]. Among them, 28.7 is relatively prominent in the entire data sequence. Therefore, the suspected abnormal energy consumption value is 28.7, including the missing value: NaN.

[0048] Furthermore, invalid data is removed: no negative values ​​or zero values ​​that are not in a shutdown state, valid data is retained and the missing positions are marked, i.e., NaN.

[0049] Furthermore, the original data sequence is the energy consumption data of a certain device at the 15-minute level, with a total of 7 points covering approximately 1.75 hours. Due to the small amount of data, the entire sequence is directly used as one analysis window (in actual scenarios, it is recommended that the window have ≥24 points; this is a simplified demonstration).

[0050] Furthermore, after removing missing NaN values, the valid data is: [5.2, 5.4, 5.3, 28.7, 5.5, 5.1]. The data sequence of the above valid data is sorted in ascending order according to the numerical value to calculate the quartiles: [5.1, 5.2, 5.3, 5.4, 5.5, 28.7]. The number of valid data is n = 6.

[0051] Furthermore, Q1 selects the 25th percentile position in the data sequence [5.1,5.2,5.3,5.4,5.5,28.7].

[0052] Furthermore, the upper quartile position = 6 × 25% = 1.5, that is, take the mean of the first and second data (5.1 + 5.2) ÷ 2 = 5.15.

[0053] Furthermore, Q3 selects the 75th position in the data sequence [5.1,5.2,5.3,5.4,5.5,28.7].

[0054] Furthermore, the lower quartile position = 6 × 75% = 4.5, that is, take the mean of the 4th and 5th data (5.4 + 5.5) ÷ 2 = 5.45.

[0055] Then IQR = Q3 - Q1 = 5.45 - 5.15 = 0.3.

[0056] Furthermore, determine the lower and upper limits:

[0057] The lower limit is calculated as follows: Lower = Q1 - 1.5 × IQR = 5.15 - 1.5 × 0.3 = 4.7.

[0058] Upper value = Q3 + 1.5 × IQR = 5.45 + 1.5 × 0.3 = 5.9.

[0059] Further, outlier determination: 28.7 > 5.9 → determined to be an outlier with instantaneous high energy consumption.

[0060] Furthermore, the position of the outlier in the time series was located. The original sequence was [5.2, 5.4, 5.3, 28.7, 5.5, NaN, 5.1]. The data before 28.7 was 5.3, 15 minutes earlier in time, and the data after 28.7 was 5.5, 15 minutes later in time.

[0061] Furthermore, a weighted interpolation method is used for repair: repair value = previous data × 0.6 + next data × 0.4, that is, repair value = 5.3 × 0.6 + 5.5 × 0.4 = 3.18 + 2.2 = 5.38.

[0062] Furthermore, the position of the missing value in the time series is located: between 5.5 (before) and 5.1 (after), with an interval of 1 data point (k=1), the missing point is the first position in the interval (i=1).

[0063] Furthermore, linear interpolation is used for repair: repair value = 5.5 + (5.1 - 5.5) × (1 / 1) = 5.5 - 0.4 = 5.1. After repair, the complete sequence is arranged in the original time order: [5.2, 5.4, 5.3, 5.38, 5.5, 5.1, 5.1].

[0064] Therefore, through the IQR algorithm, the instantaneous high energy consumption anomaly of 28.7 in the original sequence was corrected to 5.38, and the missing value NaN was corrected to 5.1. The corrected data can be used for carbon emission accounting. For example, based on the conversion of 1 kWh = 0.785 kg CO2, the accounting error can be reduced by approximately 28.7 × 0.785 - 5.38 × 0.785 ≈ 18.4 kg CO2, effectively solving the problem of inaccurate carbon emission calculation caused by instantaneous anomalies.

[0065] In step S104, carbon emissions within a predetermined time period are determined based on the repaired energy consumption data, normal energy consumption data in the time series, and fuel consumption data.

[0066] Specifically, this invention calculates carbon emissions in two parts: direct emissions calculation and indirect emissions calculation. Direct emissions calculation refers to carbon dioxide generated from fuel consumption data, while indirect emissions calculation refers to carbon dioxide generated from electricity emissions. As an optional embodiment of this invention, determining carbon emissions within a predetermined time period based on repaired energy consumption data, normal energy consumption data in a time series, and fuel consumption data includes: determining a first emission factor of the regional power grid where the target device is located and a second emission factor of the fuel type corresponding to the fuel consumption data; determining a first carbon emission amount based on the repaired energy consumption data, normal energy consumption data in a time series, and the first emission factor; determining a second carbon emission amount based on the fuel consumption data and the second emission factor; and determining the sum of the first and second carbon emissions as the total carbon emissions.

[0067] Specifically, the first and second emission factors can be referenced in the "National Greenhouse Gas Emission Factor Database (2024 Edition)" and adjusted annually according to database updates. For example, the natural gas emission factor is 2.16 kg CO2 / m³. 3 .

[0068] Furthermore, in this embodiment of the invention, the following formula is used to calculate direct emissions:

[0069] The second carbon emission factor calculated from direct emissions: CO2 produced by burning fossil fuels = fuel consumption data × second emission factor.

[0070] The first carbon emission calculated from indirect emissions: Purchased electricity emissions = (Repair energy consumption data + Normal energy consumption data in the time series) × Regional power grid emission factor (first emission factor).

[0071] The sum of the first and second carbon emissions is taken as the total carbon emissions for the predetermined period.

[0072] Furthermore, as an optional embodiment of the present invention, after determining the carbon emissions within a predetermined period based on the repair energy consumption data, normal energy consumption data in the time series, and fuel consumption data, the method further includes: determining weekly, monthly, and quarterly average values ​​based on the carbon emissions within the predetermined period; triggering a yellow alert to optimize production scheduling when the daily carbon emissions exceed a first threshold, the first threshold being determined based on the weekly average; triggering an orange alert and generating an emission reduction recommendation report when the carbon emissions for the first consecutive day exceed a second threshold, the emission reduction recommendation report including adjustments to the operating parameters of the target equipment, the second threshold being determined based on the monthly average; and triggering a red alert and a production shutdown recommendation when the carbon emissions for the second consecutive day exceed a third threshold, the third threshold being determined based on the quarterly average, where the number of the second day is greater than the number of the first day.

[0073] Specifically, in this embodiment of the invention, the first threshold can be a weekly average of 10%. The second threshold can be a monthly average of 20%. The third threshold can be a quarterly average of 30%. Of course, the first to third thresholds can also be determined according to the actual situation, and this embodiment of the invention does not limit them here.

[0074] Furthermore, embodiments of the present invention can determine weekly, monthly, and quarterly averages based on carbon emissions within a predetermined time period. Specifically, the weekly, monthly, and quarterly averages can be determined based on the daily carbon emissions within the predetermined time period; that is, the daily carbon emissions multiplied by seven days yield the weekly average, the daily carbon emissions multiplied by thirty days yield the monthly average, and the daily carbon emissions multiplied by 90 days yield the quarterly average. In addition, the daily carbon emissions within the task period can be predicted based on production output, raw material consumption, and task time data in the production task plan, combined with equipment energy consumption, and the weekly, monthly, and quarterly averages can be determined based on the predicted total carbon emissions.

[0075] Furthermore, during subsequent monitoring, if daily carbon emissions exceed the weekly average by 10%, a yellow alert is triggered, and an SMS message is sent to the company's energy management department, prompting them to optimize production scheduling. If carbon emissions exceed the monthly average by 20% for three consecutive days, an orange alert is triggered, automatically generating an emission reduction recommendation report, which is synchronized to the government regulatory platform. The emission reduction recommendation report includes, but is not limited to, adjusting equipment operating parameters. If carbon emissions exceed the quarterly quota by 30% for five consecutive days, a red alert is triggered, at which point a production shutdown recommendation is issued.

[0076] The technical solution provided by this invention employs a dynamic calculation mode using a sliding window in the time-series of energy consumption data. This mode adapts to the diurnal periodicity and shift fluctuations of energy consumption data, improving the accuracy of identifying abnormal energy consumption data. Furthermore, the anomaly identification method based on quartile positions effectively identifies extreme outliers (such as zero-value drift or instantaneous peaks caused by sensor malfunctions). Instead of directly discarding abnormal energy consumption data, the identified data is repaired, thus improving the accuracy of subsequent carbon emission calculations. Finally, when calculating carbon emissions, the system comprehensively repairs three types of multi-source data: energy consumption data, normal energy consumption data, and fuel consumption data. This covers both the accurate quantification of indirect energy consumption such as electricity and heat, and the statistics of direct fossil fuel consumption, ensuring a complete dimension for carbon emission calculation. Therefore, the carbon emissions calculated using this method are closer to actual emissions, improving the accuracy and reliability of carbon emission monitoring.

[0077] This application provides a carbon emission monitoring system based on multi-source heterogeneous data, referencing... Figure 2 As shown, Figure 2This is a schematic diagram of a carbon emission monitoring system based on multi-source heterogeneous data provided in an embodiment of the present invention. The carbon emission monitoring system based on multi-source heterogeneous data includes: an acquisition module 201, used to acquire basic carbon emission accounting data of a target device within a predetermined time period, the basic carbon emission accounting data including energy consumption data and fuel consumption data; an identification module 202, used to extract the time series of energy consumption data from the basic carbon emission accounting data of the target device, and identify abnormal energy consumption data from the time series based on the energy consumption data at the quartile position in each sliding window in the time series; a repair module 203, used to repair the abnormal energy consumption data to obtain repaired energy consumption data; and a determination module 204, used to determine the carbon emission amount within the predetermined time period based on the repaired energy consumption data, the normal energy consumption data in the time series, and the fuel consumption data.

[0078] Furthermore, the identification module 202 includes: a determining unit, configured to sort the energy consumption data within the sliding window in ascending order of numerical value, and determine a first energy consumption data located at the upper quartile position and a second energy consumption data located at the lower quartile position within the sliding window, wherein the upper quartile position is the 25th percentile position and the lower quartile position is the 75th percentile position within the sliding window; a calculation unit, configured to calculate the interquartile range of the sliding window based on the second energy consumption data and the first energy consumption data; the determining unit, further configured to determine an upper limit value and a lower limit value based on the interquartile range, the first energy consumption data, and the second energy consumption data; and a marking unit, configured to mark energy consumption data within the sliding window that exceeds the upper limit value and the lower limit value as abnormal energy consumption data.

[0079] Furthermore, the determining unit is also configured to determine the upper limit and lower limit values ​​based on the interquartile range, the first energy consumption data, and the second energy consumption data, including: calculating a first product between a predetermined multiple and the interquartile range; determining the difference between the first energy consumption data and the first product as the lower limit value; and determining the sum between the second energy consumption data and the first product as the upper limit value.

[0080] Furthermore, it also includes: a determination module, used to determine that the target device has a momentary fault if the abnormal energy consumption data is a single-point sudden increase, wherein the single-point sudden increase indicates that only one energy consumption data exceeds the upper limit value, and the preceding and following energy consumption data of the abnormal energy consumption data are both within the range of the upper limit value and the lower limit value; if the abnormal energy consumption data is a continuous multiple data sudden increase, it is determined that the target device has a continuous fault, wherein the continuous multiple data sudden increase indicates that multiple consecutive energy consumption data exceed the upper limit value.

[0081] Furthermore, the repair module 203 is also configured to, if the abnormal energy consumption data is located between the beginning and the end of the time series, obtain the preceding and following energy consumption data of the abnormal energy consumption data from the time series, and weight the preceding and following energy consumption data of the abnormal energy consumption data to obtain the repaired energy consumption data of the abnormal energy consumption data; if the abnormal energy consumption data is located at the beginning or the end of the time series, repair the abnormal energy consumption data using the following energy consumption data of the beginning to obtain the repaired energy consumption data, or repair the abnormal energy consumption data using the preceding energy consumption data of the end to obtain the repaired energy consumption data.

[0082] Furthermore, the identification module 202 is also used to identify the missing data position from each sliding window; the determination module 204 is also used to determine the energy consumption data of the position preceding and following the missing data position from the time sequence, as well as the sort number of the target position of the missing data position in the interval between the energy consumption data of the preceding and following positions and the number of positions in the interval; and to determine the repair energy consumption data of the missing data position based on the energy consumption data of the position preceding and following the missing data position, the sort number of the target position and the number of positions in the interval.

[0083] Furthermore, the determining module 204 is also used to determine a first emission factor of the regional power grid where the target device is located and a second emission factor of the fuel type corresponding to the fuel consumption data; determine the first carbon emission based on the repair energy consumption data, the normal energy consumption data in the time series and the first emission factor; determine the second carbon emission based on the fuel consumption data and the second emission factor; and determine the sum of the first carbon emission and the second carbon emission as the carbon emission.

[0084] Furthermore, the determining module 204 is also used to determine the weekly average, monthly average, and quarterly average carbon emissions based on the carbon emissions within the predetermined time period; if the daily carbon emissions exceed the first threshold, a yellow alert is triggered to optimize production scheduling, the first threshold being determined based on the weekly average; if the carbon emissions for a consecutive number of days exceed the second threshold, an orange alert is triggered, and an emission reduction recommendation report is generated, the emission reduction recommendation report including adjustments to the operating parameters of the target equipment, the second threshold being determined based on the monthly average; if the carbon emissions for a consecutive number of days exceed the third threshold, a red alert is triggered, and a production shutdown recommendation is triggered, the third threshold being determined based on the quarterly average, the number of the second days being greater than the number of the first days.

[0085] The technical solution provided by this invention employs a dynamic calculation mode using a sliding window in the time-series of energy consumption data. This mode adapts to the diurnal periodicity and shift fluctuations of energy consumption data, improving the accuracy of identifying abnormal energy consumption data. Furthermore, the anomaly identification method based on quartile positions effectively identifies extreme outliers (such as zero-value drift or instantaneous peaks caused by sensor malfunctions). Instead of directly discarding abnormal energy consumption data, the identified data is repaired, thus improving the accuracy of subsequent carbon emission calculations. Finally, when calculating carbon emissions, the system comprehensively repairs three types of multi-source data: energy consumption data, normal energy consumption data, and fuel consumption data. This covers both the accurate quantification of indirect energy consumption such as electricity and heat, and the statistics of direct fossil fuel consumption, ensuring a complete dimension for carbon emission calculation. Therefore, the carbon emissions calculated using this method are closer to actual emissions, improving the accuracy and reliability of carbon emission monitoring.

[0086] Corresponding to the carbon emission monitoring method based on multi-source heterogeneous data provided in the above embodiments, and based on the same technical concept, this invention also provides a carbon emission monitoring system based on multi-source heterogeneous data. This system is used to execute the aforementioned carbon emission monitoring method based on multi-source heterogeneous data. Figure 3 A schematic diagram of a carbon emission monitoring system based on multi-source heterogeneous data is provided as an embodiment of the present invention, as shown below. Figure 3 As shown. Carbon emission monitoring systems based on multi-source heterogeneous data can vary significantly due to differences in configuration or performance. They may include one or more processors 301 and memory 302. Memory 302 stores computer programs that can run on processor 301. Processor 301 executes the programs stored in memory 302 to achieve the above... Figure 1 The various steps in the method embodiment are described. The memory 302 can be temporary or persistent storage. The application stored in the memory 302 may include one or more modules (not shown in the figures), each module may include a series of computer-executable instructions for the carbon emission monitoring system based on multi-source heterogeneous data.

[0087] Furthermore, the processor 301 can be configured to communicate with the memory 302 and execute a series of computer-executable instructions stored in the memory 302 on the carbon emission monitoring system based on multi-source heterogeneous data. The carbon emission monitoring system based on multi-source heterogeneous data may also include one or more power supplies 303, one or more wired or wireless network interfaces 304, one or more input / output interfaces 305, and one or more keyboards 306.

[0088] Specifically, in this embodiment, the carbon emission monitoring system based on multi-source heterogeneous data includes a processor, a communication interface, a memory, and a communication bus; wherein, the processor, communication interface, and memory communicate with each other via the bus; the memory is used to store computer programs; and the processor is used to execute the programs stored in the memory to achieve the above... Figure 1 The various steps in the method embodiments are the same as those in the above method embodiments, and have the same beneficial effects. To avoid repetition, the embodiments of the present invention will not be described again here.

[0089] It should be noted that the carbon emission monitoring system based on multi-source heterogeneous data provided in this embodiment of the invention and the carbon emission monitoring method based on multi-source heterogeneous data provided in this embodiment of the invention are based on the same application concept. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned carbon emission monitoring method based on multi-source heterogeneous data, and has the same or similar beneficial effects. Repeated parts will not be described again.

[0090] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0091] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compactdisc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0092] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A carbon emission monitoring method based on multi-source heterogeneous data, characterized in that, The carbon emission monitoring method based on multi-source heterogeneous data comprises: obtaining carbon emission accounting basic data of a target device in a predetermined period, wherein the carbon emission accounting basic data comprises energy consumption data and fuel consumption data; extracting a time sequence of the energy consumption data in the carbon emission accounting basic data of the target device, and identifying abnormal energy consumption data from the time sequence based on energy consumption data at quartile positions in each sliding window in the time sequence; repairing the abnormal energy consumption data to obtain repaired energy consumption data; determining the carbon emission amount in the predetermined period based on the repaired energy consumption data, normal energy consumption data in the time sequence, and the fuel consumption data.

2. The carbon emission monitoring method based on multi-source heterogeneous data according to claim 1, wherein, The method for identifying abnormal energy consumption data from the time sequence based on energy consumption data at quartile positions in each sliding window in the time sequence comprises: arranging the energy consumption data in the sliding window in ascending order according to numerical values, determining first energy consumption data at an upper quartile position in the sliding window and second energy consumption data at a lower quartile position in the sliding window, wherein the upper quartile position is the 25th position in the sliding window, and the lower quartile position is the 75th position in the sliding window; calculating the interquartile range of the sliding window based on the second energy consumption data and the first energy consumption data; determining an upper limit value and a lower limit value according to the interquartile range, the first energy consumption data, and the second energy consumption data; marking energy consumption data in the sliding window that is outside the range of the upper limit value and the lower limit value as the abnormal energy consumption data.

3. The method for carbon emission monitoring based on multi-source heterogeneous data according to claim 2, characterized in that, The method for determining an upper limit value and a lower limit value according to the interquartile range, the first energy consumption data, and the second energy consumption data comprises: calculating a first product between a predetermined multiple value and the interquartile range; determining the difference between the first energy consumption data and the first product as the lower limit value, and determining the sum between the second energy consumption data and the first product as the upper limit value.

4. The method for carbon emission monitoring based on multi-source heterogeneous data according to claim 2, characterized in that, After the energy consumption data in the sliding window that is outside the range of the upper limit value and the lower limit value is marked as the abnormal energy consumption data, the method further comprises: if the abnormal energy consumption data is a single-point sudden increase case, determining that the target device has a transient fault, wherein the single-point sudden increase case indicates that only one energy consumption data is outside the upper limit value, and the previous energy consumption data and the subsequent energy consumption data of the abnormal energy consumption data are within the range of the upper limit value and the lower limit value; if the abnormal energy consumption data is a continuous multiple data sudden increase case, determining that the target device has a continuous fault, wherein the continuous multiple data sudden increase case indicates that a plurality of continuous energy consumption data are outside the upper limit value.

5. The method for carbon emission monitoring based on multi-source heterogeneous data according to claim 1, characterized in that, The method for repairing the abnormal energy consumption data to obtain repaired energy consumption data comprises: if the abnormal energy consumption data is between the head and the tail of the time sequence, obtaining the previous energy consumption data and the subsequent energy consumption data of the abnormal energy consumption data from the time sequence, and weighting the previous energy consumption data and the subsequent energy consumption data of the abnormal energy consumption data to obtain the repaired energy consumption data of the abnormal energy consumption data. If the abnormal energy consumption data is at the head or tail of the time sequence, the abnormal energy consumption data is repaired by using the energy consumption data of the next bit at the head or the energy consumption data of the previous bit at the tail to obtain the repaired energy consumption data.

6. The method for carbon emission monitoring based on multi-source heterogeneous data according to claim 1, characterized in that, After identifying the abnormal energy consumption data from the time sequence based on the energy consumption data at the quartile position in each sliding window in the time sequence, the method further comprises: identifying a data missing position from each sliding window; determining the energy consumption data at a previous position and the energy consumption data at a next position of the data missing position from the time sequence, and the number of positions and the target position in the interval between the energy consumption data at the previous position and the energy consumption data at the next position of the missing position; determining the repaired energy consumption data of the data missing position based on the energy consumption data at the previous position and the energy consumption data at the next position of the data missing position, the number of positions and the target position in the interval.

7. The method for carbon emission monitoring based on multi-source heterogeneous data according to claim 1, characterized in that, The determination of the carbon emission amount in the predetermined period based on the repaired energy consumption data, the normal energy consumption data in the time sequence and the fuel consumption data comprises: determining a first emission factor of a regional power grid where the target device is located and a second emission factor of a fuel type corresponding to the fuel consumption data; determining the first carbon emission amount based on the repaired energy consumption data, the normal energy consumption data in the time sequence and the first emission factor; determining a second carbon emission amount based on the fuel consumption data and the second emission factor; determining the sum of the first carbon emission amount and the second carbon emission amount as the carbon emission amount.

8. The method for carbon emission monitoring based on multi-source heterogeneous data according to claim 1, characterized in that, After determining the carbon emission amount in the predetermined period based on the repaired energy consumption data, the normal energy consumption data in the time sequence and the fuel consumption data, the method further comprises: determining the weekly average, the monthly average and the quarterly average based on the carbon emission amount in the predetermined period; in the case that the carbon emission amount of a single day exceeds the first threshold value, triggering a yellow warning to optimize production scheduling, the first threshold value being determined based on the weekly average; in the case that the carbon emission amount of a continuous first number of days exceeds a second threshold value, triggering an orange warning and generating an emission reduction suggestion report, the emission reduction suggestion report including adjusting the operating parameters of the target device, the second threshold value being determined based on the monthly average; in the case that the carbon emission amount of a continuous second number of days exceeds a third threshold value, triggering a red warning and triggering a production suspension suggestion, the third threshold value being determined based on the quarterly average, the second number of days being greater than the first number of days.

9. A carbon emission monitoring system based on multi-source heterogeneous data, characterized in that, comprises: an acquisition module configured to acquire carbon emission accounting basic data of a target device in a predetermined period, the carbon emission accounting basic data comprising energy consumption data and fuel consumption data; an identification module configured to extract a time sequence of the energy consumption data in the carbon emission accounting basic data of the target device, and identify abnormal energy consumption data from the time sequence based on the energy consumption data at the quartile position in each sliding window in the time sequence; a repairing module configured to repair the abnormal energy consumption data to obtain repaired energy consumption data; a determining module configured to determine carbon emission in the predetermined period based on the repaired energy consumption data, normal energy consumption data in the time sequence, and the fuel consumption data.

10. An electronic device, comprising: comprising: a processor and a memory; wherein the memory is configured to store a computer program executable on the processor; the processor is configured to execute the program stored on the memory to implement the steps of the carbon emission monitoring method based on multi-source heterogeneous data according to any one of claims 1-8.

Images