Refined Simulation Calculation Method for Carbon Emissions in Regional Power Systems

Abstract

This invention discloses a method for refined simulation calculation of carbon emissions from regional power systems, belonging to the field of simulation calculation technology. The method includes the following steps: collecting regional power system data; initially calculating carbon emissions from the generation side, grid side, and user side; determining the carbon emission accounting boundary of the regional power system based on regional geographic information data; obtaining distributed storage data using distributed storage technology; extracting features from the topological data to obtain regional topological feature data; constructing a simulation calculation model by combining the distributed storage data with a neural network algorithm; outputting refined simulation data for carbon emissions from the regional power system; and correcting the refined simulation data by integrating environmental data and a random forest algorithm, finally outputting the corrected refined simulation data for carbon emissions from the regional power system. This invention achieves refined simulation calculation of carbon emissions from regional power systems throughout their entire lifecycle and across multiple dimensions.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of simulation calculation, and particularly relates to a regional power system carbon emission fine accounting simulation calculation method. BACKGROUND

[0002] Under the guidance of the double carbon target, the power industry, as a core field of carbon emission, needs precise carbon emission accounting data support for the scientific planning of its emission reduction path. Traditional regional power system carbon emission accounting relies on macro statistical data and does not fully consider various carbon emission factors, resulting in large errors in the accounting results and difficulty in meeting the fine needs of emission reduction decisions.

[0003] With large-scale renewable energy grid connection, popularization of distributed power supply and deepening of power market reform, the regional power system presents a complex form of source-grid-load-storage multi-element cooperation, and the frequent development of cross-regional power trading further increases the difficulty of defining carbon emission responsibility. The traditional accounting method cannot accurately locate the key nodes of carbon emission, which restricts the pertinence and effectiveness of emission reduction measures. At the same time, the policy level requires continuous improvement of the standardization and transparency of carbon emission accounting, and it is urgent to establish a simulation calculation system that takes into account scientificity and operability. Existing simulation tools mainly focus on power system operation simulation, and the carbon emission accounting function is weak, and there is no unified accounting standard and technical framework. SUMMARY

[0004] In view of the above deficiencies in the prior art, the purpose of the present application is to provide a regional power system carbon emission fine accounting simulation calculation method, which can realize multi-dimensional fine accounting of the whole cycle of regional power system carbon emission, and effectively improve the accounting accuracy and intelligent level.

[0005] To achieve the above purpose, the following technical solutions are used:

[0006] The regional power system carbon emission fine accounting simulation calculation method comprises the following steps:

[0007] Collect regional power system data, including power generation side data, power grid side data, user side data, regional geographic information data, environmental data and topology data, and preprocess the regional power system data;

[0008] Preliminary calculation of power generation side carbon emission, power grid side carbon emission and user side carbon emission is carried out through power generation side data, power grid side data and user side data, the carbon emission accounting boundary of the power generation side, the power grid side and the user side in the regional power system is determined based on the regional geographic information data, and distributed storage data is obtained in combination with the distributed storage technology;

[0009] The topological data is subjected to feature extraction to obtain regional topological feature data, combined with distributed storage data and a neural network algorithm, a simulation calculation model is constructed, and regional power system carbon emission fine accounting simulation data is output;

[0010] The simulation calculation correction model is constructed by combining the environmental data and the random forest algorithm, and the correction coefficient is output, and then the regional power system carbon emission fine accounting simulation data is corrected;

[0011] In the form of data report, the corrected regional power system carbon emission fine accounting simulation data is output.

[0012] Preferably, the process of collecting regional power system data includes:

[0013] Different types of collection devices are deployed to collect power generation side data, power grid side data, user side data, regional geographic information data, environmental data and topological data, and regional power system data is obtained;

[0014] The power generation side data includes unit fuel consumption, fuel unit heat value carbon content, fuel low heat value, total carbon emissions and total on-grid power of all power generation sides in the region; the power grid side data is power transmission loss; the user side data includes user terminal power consumption and user self-use new energy power generation; the regional geographic information data includes site area and boundary latitude and longitude coordinates of power plant area, path direction of power transmission line, substation area and boundary latitude and longitude coordinates, and user aggregation area and boundary latitude and longitude coordinates; the environmental data includes real-time temperature, real-time humidity, altitude, real-time precipitation pH value, real-time wind speed, real-time light intensity and real-time air density of the regional power system environment; the topological data includes regional power source topology, regional power grid topology and regional user topology.

[0015] Preferably, the process of preprocessing regional power system data includes:

[0016] The collected power generation side data, power grid side data, user side data, regional geographic information data and environmental data are subjected to data cleaning;

[0017] A unified time format is adopted to perform time series alignment processing on the power generation side data, power grid side data and user side data, forming a full-cycle carbon emission accounting data link of power generation, power transmission and power consumption;

[0018] Through a graphical tool, the structure of the regional power source topology, the regional power grid topology and the regional user topology can be visualized, and the graph theory and topology optimization algorithm can be used to remove redundancy and optimize the structure of the regional power source topology, the regional power grid topology and the regional user topology.

[0019] Preferably, the preliminary calculation process of power generation side carbon emission, power grid side carbon emission and user side carbon emission includes:

[0020] The power generation side data is used to preliminarily calculate the carbon emission of the power generation side and the average power supply carbon emission factor of the regional power system;

[0021] Based on the average power supply carbon emission factor of the regional power system, the carbon emission of the power grid side is preliminarily calculated in combination with the power grid side data;

[0022] Based on the average power supply carbon emission factor of the regional power system, the carbon emission of the user side is preliminarily calculated in combination with the user side data.

[0023] Preferably, the process of determining the carbon emission accounting boundary of the power generation side, the power grid side and the user side in the regional power system based on the regional geographic information data comprises:

[0024] The GIS map is initialized, the coordinate system of the GIS map is unified, the power generation side region is identified according to the site area and boundary latitude and longitude coordinates of the power plant area, and the site area and boundary latitude and longitude coordinates of the power plant area are superimposed on the GIS map to determine the carbon emission accounting boundary of the power generation side in the regional power system through the GIS technology;

[0025] The path of the power transmission line is drawn in the GIS map according to the path direction of the power transmission line, and the area and boundary latitude and longitude coordinates of the substation area are superimposed on the GIS map to determine the carbon emission accounting boundary of the power grid side in the regional power system;

[0026] The user region is drawn in the GIS map with reference to the area and boundary latitude and longitude coordinates of the user aggregation area, and the carbon emission accounting boundary of the user side in the regional power system is determined.

[0027] Preferably, the process of obtaining the distributed storage data by using the distributed storage technology comprises:

[0028] The distributed storage technology is adopted to establish three independent storage partitions based on the carbon emission accounting boundary of the power generation side, the power grid side and the user side in the regional power system, and the geographic range and data ownership of each independent storage partition are clearly defined;

[0029] According to the carbon emission accounting boundary of the power generation side, the power grid side and the user side in the corresponding regional power system and the time dimension, the preliminary calculation results of the carbon emission of the power generation side, the power grid side and the user side are respectively split into a plurality of data shards, and the data format of each data shard after the split processing is unified;

[0030] According to the size and access frequency of the data shards, the data shards are allocated to different nodes in the distributed storage cluster, a replica mechanism is adopted to store each data shard, and the establishment of the storage partition is realized;

[0031] The geographical space index and attribute index are established for each storage partition, the data storage completeness of each node is verified through data shard hash value comparison and metadata integrity checking technology, the correspondence between the data and the carbon emission accounting boundary of the power generation side, the power grid side and the user side in the regional power system is confirmed to be correct, the distributed storage of the preliminary calculation results of the carbon emissions of the power generation side, the power grid side and the user side is realized, and the distributed storage data is obtained.

[0032] Preferably, the process of feature extraction on the topology data to obtain the regional topology feature data comprises:

[0033] The regional topology feature data comprises power source type proportion, power source cluster connectivity, main grid access rate, line connection density, transformer substation level coefficient, power grid loop network rate, standby line redundancy rate, user distribution network access rate and self-use new energy coverage rate.

[0034] The topology analysis technology and the graph neural network are combined to extract the features of the regional power source topology, and the unit quantity of each type of power source, the total unit quantity, the actual connected unit pairs in the power source cluster, the power source node quantity accessing the main grid and the total power source node quantity in the power generation side of the regional power system are obtained.

[0035] The power source type proportion is obtained through the proportion of the unit quantity of each type of power source in the total unit quantity, the power source cluster connectivity is obtained by calculating the theoretical maximum connected unit pairs in the power source cluster of the power generation side of the regional power system and then obtaining the ratio of the actual connected unit pairs in the power source cluster to the theoretical maximum connected unit pairs, and the main grid access rate is obtained by calculating the ratio of the power source node quantity accessing the main grid to the total power source node quantity.

[0036] The power system topology analysis technology and the graph theory algorithm are combined to adapt to the electrical connection characteristics of the power grid side topology, the features of the regional power grid topology are extracted, and the actual existing transmission line quantity, the different voltage level transformer substation quantity, the total transformer substation quantity, the line quantity forming a loop network structure, the total transmission line quantity, the standby transmission line quantity and the operating line quantity in the power grid side of the regional power system are obtained.

[0037] The total node quantity in the power grid side of the regional power system is used to calculate the theoretical maximum possible line quantity in the power grid side of the regional power system, and the proportion of the actual existing transmission line quantity in the theoretical maximum possible line quantity is calculated to obtain the line connection density; the proportion of the different voltage level transformer substation quantity in the total transformer substation quantity is calculated to obtain the transformer substation level coefficient; the ratio of the line quantity forming a loop network structure to the total transmission line quantity is calculated to obtain the power grid loop network rate; and the ratio of the standby transmission line quantity to the operating line quantity is calculated to obtain the standby line redundancy rate.

[0038] The user topology analysis technology and attribute classification algorithm are used to extract the characteristics of the regional user topology, and the number of users connected at different voltage levels, the total number of users, and the number of users with new energy power generation on the user side in the regional power system are obtained.

[0039] The user distribution network access rate is calculated by the proportion of the number of users connected at different voltage levels on the user side in the total number of users in the regional power system. The self-use new energy coverage rate is obtained by calculating the ratio of the number of users with new energy power generation on the user side to the total number of users.

[0040] Preferably, the process of constructing a simulation calculation model by a neural network algorithm and outputting the regional power system carbon emission fine accounting simulation data includes:

[0041] The regional power system carbon emission fine accounting simulation data includes the simulation calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission;

[0042] The preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission are selected from the distributed storage data. The preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission and the regional topology characteristic data are integrated to generate a fine accounting simulation data set. The fine accounting simulation data set is divided into a first training set and a first test set;

[0043] The neural network algorithm is adopted. The first training set data is used as input data, and the regional power system carbon emission fine accounting simulation data is used as output data. The nonlinear relationship between the preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission, the regional topology characteristic data and the regional power system carbon emission fine accounting simulation data is learned to obtain a trained simulation calculation model;

[0044] The first test set data is input into the trained simulation calculation model. The stochastic gradient descent optimizer is used to adjust the parameters of the simulation calculation model, optimize the performance of the simulation calculation model, obtain the final simulation calculation model, and output the corresponding regional power system carbon emission fine accounting simulation data in combination with the current preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission and the regional topology characteristic data.

[0045] Preferably, the simulation calculation correction model is constructed by integrating environmental data and the random forest algorithm, and the correction coefficient is output to correct the regional power system carbon emission fine accounting simulation data. The process includes:

[0046] The environmental data is integrated to generate a simulation calculation correction data set, and the simulation calculation correction data set is divided into a second training set and a second test set;

[0047] The random forest algorithm is used, the second training set data is taken as an input variable, and the correction coefficient is taken as an output variable, a nonlinear relationship between environment data and the correction coefficient is learned, and a trained simulation calculation correction model is acquired.

[0048] The second test set data is input into the trained simulation calculation correction model, the simulation calculation correction model parameters are adjusted, the simulation calculation correction model performance is optimized, and finally, the simulation calculation correction model is obtained.

[0049] The current environment data is input into the simulation calculation correction model, the correction coefficient is output, the correction coefficient is referred to, and the regional power system carbon emission fine accounting simulation data is corrected.

[0050] Preferably, in the form of a data report, the process of outputting the corrected regional power system carbon emission fine accounting simulation data includes:

[0051] The dimensions and formats of the preset data report are classified according to the power generation side, the power grid side and the user side, each category contains the corresponding carbon emission simulation calculation results after correction processing;

[0052] The simulation calculation results of the corrected power generation side carbon emission, power grid side carbon emission and user side carbon emission are integrated, and are filled into the data report one by one, to generate a regional power system carbon emission fine accounting simulation calculation result output report in the preset report format.

[0053] The beneficial effects of the present application are:

[0054] On the one hand, the present application relies on GIS technology to determine the carbon emission accounting boundaries of the power generation side, the power grid side and the user side, and combines distributed storage technology to realize safe partition storage and efficient calling of data at each link, avoiding the bottleneck problem of traditional storage architecture; on the other hand, by capturing the structural characteristics of the power system (such as power supply cluster connectivity and power grid ring network rate), the simulation model is constructed by combining neural network algorithm, and the nonlinear relationship between carbon emission and multiple factors is accurately fitted, breaking the limitations of traditional methods relying on macro statistical data and ignoring topological influence, and finally realizing multi-dimensional fine accounting of the whole cycle carbon emission of regional power system generation, transmission and power consumption, and the simulation data precision is significantly better than that of the traditional average carbon emission factor method.

[0055] The application can effectively offset the interference of environmental factors on carbon emission accounting by fusing environmental data through a random forest algorithm to construct a correction model and dynamically optimizing simulation data, so that the final result is more in line with the actual operation scene; meanwhile, the corrected accounting result is visualized in the form of data report and output, providing clear carbon emission data support for users at each link, not only improving the intelligent degree of carbon emission accounting, but also accurately positioning the key nodes of carbon emission, solving the problem that the traditional method is difficult to define the responsibility of emission reduction and restrict the pertinence of emission reduction measures. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 is a flowchart of the method of the application;

[0057] Figure 2 is a comparison diagram of carbon emissions on the power generation side;

[0058] Figure 3 is a comparison diagram of carbon emissions on the power grid side;

[0059] Figure 4 is a comparison diagram of carbon emissions on the user side;

[0060] Figure 5 is an iteration diagram of the training performance of the simulation calculation model. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings.

[0062] As shown in Figure 1 The regional power system carbon emission fine accounting simulation calculation method is composed of the following steps:

[0063] Step 101, collecting regional power system data, including power generation side data, power grid side data, user side data, regional geographic information data, environmental data and topology data, and preprocessing the regional power system data;

[0064] Step 102, preliminarily calculating the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission through the power generation side data, the power grid side data and the user side data, determining the carbon emission accounting boundary of the power generation side, the power grid side and the user side in the regional power system based on the regional geographic information data, and combining the distributed storage technology to obtain distributed storage data, thereby improving the security and availability of data storage;

[0065] Step 103, extracting features from the topology data to obtain regional topology feature data, combining the distributed storage data with a neural network algorithm to construct a simulation calculation model, outputting regional power system carbon emission fine accounting simulation data, and integrating topology characteristics to improve the accuracy of regional power system carbon emission fine accounting simulation calculation;

[0066] Step 104, integrate environmental data with random forest algorithm, build simulation calculation correction model, output correction coefficient, and then correct regional power system carbon emission fine accounting simulation data, further improve the accuracy of regional power system carbon emission fine accounting simulation calculation results;

[0067] Step 105, output the corrected regional power system carbon emission fine accounting simulation data in the form of data report, and visualize the results of regional power system carbon emission fine accounting simulation calculation, which is convenient for subsequent reference and application.

[0068] In step 101, the process of collecting regional power system data includes:

[0069] Use fuel flow meter to collect unit fuel consumption, combine with combustion heat value analyzer and combustion component analyzer to collect unit heat value carbon content, combine with automatic calorimeter and water content tester to collect low heat value, use flue gas emission monitoring system and electric energy metering device to obtain total carbon emission and total on-grid power of all power generation sides in the region; Use electric energy metering device to collect power transmission loss; Use smart electric energy meter and distributed new energy power generation special electric energy metering device to collect user terminal power consumption and user self-use new energy power generation;

[0070] Use GPS equipment to collect boundary latitude and longitude coordinates of power plant area, substation area boundary latitude and longitude coordinates and user aggregation area boundary latitude and longitude coordinates, use total station to collect path direction of power transmission line, use unmanned aerial vehicle to carry GPS equipment to collect site area of power plant area, substation area and user aggregation area;

[0071] Use temperature sensor, humidity sensor, barometer, pH sensor, anemometer, light sensor and air density sensor to collect real-time temperature, real-time humidity, altitude, real-time precipitation pH, real-time wind speed, real-time light intensity and real-time air density of regional power system environment;

[0072] Combine with power generation side smart electric energy meter, voltage transformer, unit state monitoring sensor and booster station measurement and control device to collect regional power supply topology; Combine with line fault indicator, intelligent circuit breaker, substation automation system, distribution terminal and GPS synchronous clock to collect regional power grid topology; Combine with user side smart electric energy meter, load monitoring terminal, user access switch, power information collector and Internet of Things sensor to collect regional user topology.

[0073] In step 101, the process of preprocessing regional power system data includes:

[0074] The collected power generation side data, power grid side data, user side data, regional geographic information data and environmental data are cleaned up;

[0075] The power generation side data, power grid side data and user side data are time series aligned by using a unified time format, and a full-cycle carbon emission accounting data link of power generation, power transmission and power consumption is formed;

[0076] Through a graphical tool, the structures of regional power supply topology, regional power grid topology and regional user topology are visualized, and the regional power supply topology, regional power grid topology and regional user topology are processed to remove redundancy and optimize structure by using graph theory and topology optimization algorithm.

[0077] In step 102, the preliminary calculation process of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission includes:

[0078] The power generation side carbon emission and the regional power system average power supply carbon emission factor are preliminarily calculated by using the power generation side data, and the calculation process includes:

[0079] ;

[0080] ;

[0081] Wherein, is the preliminary calculation result of the power generation side carbon emission, and the unit is ton of carbon dioxide, abbreviated as ; is the unit heat value of fuel consumption, and the unit is ton, abbreviated as t; is the carbon content of fuel unit heat value, and the unit is ton of carbon dioxide / gigajoule, abbreviated as ; is the low heat value of fuel, and the unit is gigajoule / ton, abbreviated as ; is the carbon oxidation rate, which is a industry standard value; L is the carbon conversion coefficient, which is usually 44 / 12, about equal to 3.667; and are the total carbon emission and the total on-grid power of all power generation sides in the region, and the units are ton of carbon dioxide and megawatt-hour, respectively, abbreviated as ; is the regional power system average power supply carbon emission factor, and the unit is ton of carbon dioxide / megawatt-hour, abbreviated as ;

[0082] Based on the regional power system average power supply carbon emission factor, the power grid side carbon emission is preliminarily calculated combined with the power grid side data, and the calculation process includes:

[0083] ;

[0084] wherein, is the preliminary calculation result of the carbon emission of the power grid side, the unit is ton of carbon dioxide, and the abbreviation is ; is the power transmission loss, the unit is megawatt hour, and the abbreviation is MWh;

[0085] Based on the average power supply carbon emission factor of the regional power system, combined with the user side data, the user side carbon emission is preliminarily calculated, and the preliminary calculation formula is as follows:

[0086]

[0087] wherein, is the preliminary calculation result of the carbon emission of the user side, the unit is ton of carbon dioxide, and the abbreviation is ; and are the power consumption of the user terminal and the self-use new energy power generation of the user, respectively, and the unit is megawatt hour, and the abbreviation is MWh.

[0088] In step 102, based on the regional geographic information data, the process of determining the carbon emission accounting boundary of the power generation side, the power grid side and the user side in the regional power system includes:

[0089] Through GIS technology, initialize the GIS map, unify the coordinate system of the GIS map, identify the power generation side area according to the site area and boundary latitude and longitude coordinates of the power plant area, and superimpose the site area and boundary latitude and longitude coordinates of the power plant area on the GIS map to determine the carbon emission accounting boundary of the power generation side in the regional power system;

[0090] According to the path of the power transmission line, the path of the power transmission line is drawn in the GIS map, and the area and boundary latitude and longitude coordinates of the transformer station area are superimposed on the GIS map to determine the carbon emission accounting boundary of the power grid side in the regional power system;

[0091] Referring to the area and boundary latitude and longitude coordinates of the user aggregation area, the user area is drawn in the GIS map, and then the carbon emission accounting boundary of the user side in the regional power system is determined.

[0092] In step 102, the process of obtaining distributed storage data by using distributed storage technology includes:

[0093] Using distributed storage technology, three independent storage partitions are established based on the carbon emission accounting boundary of the power generation side, the power grid side and the user side in the regional power system, and the geographic range and data ownership of each independent storage partition are clearly defined;

[0094] According to the accounting boundary and time dimension of carbon emissions of the power generation side, the grid side and the user side in the corresponding regional power system, the preliminary calculation results of the carbon emissions of the power generation side, the grid side and the user side are respectively split into several data slices, and the data format of each data slice after unified splitting processing is split, the data format includes carbon emission value unit, accounting timestamp, boundary identifier, data source and other key metadata, and each data slice is completely matched with the geographical range of the corresponding accounting boundary;

[0095] According to the size and access frequency of the data slices, the data slices are distributed to different nodes in the distributed storage cluster, and a copy mechanism is used to store each data slice, so as to realize the establishment of storage partitioning, wherein the copies are distributed in different nodes to ensure the safety and high availability of data, and to provide convenience for subsequent modeling;

[0096] Geospatial index and attribute index are established for each storage partition to support fast data query according to the carbon emission accounting boundary of the power generation side, the grid side and the user side in the regional power system. Through data slice hash value comparison and metadata integrity verification technology, the integrity of the data stored in each node is verified, and the correspondence between the data and the carbon emission accounting boundary of the power generation side, the grid side and the user side in the regional power system is confirmed to be correct, realizing the distributed storage of the preliminary calculation results of the carbon emissions of the power generation side, the grid side and the user side, and obtaining the distributed storage data.

[0097] In step 103, the process of feature extraction from the topology data to obtain regional topology feature data includes:

[0098] The regional topology feature data includes power source type proportion, power source cluster connectivity, main grid access rate, line connection density, transformer station level coefficient, power grid ring network rate, standby line redundancy rate, user distribution network access rate and self-use new energy coverage rate.

[0099] The topology analysis technology and the graph neural network are combined to extract the features of the regional power source topology, and the number of units of each type of power source, the total number of units, the number of actually connected unit pairs in the power source cluster, the number of power source nodes accessing the main grid and the total number of power source nodes in the regional power system are obtained.

[0100] The power source type proportion is obtained by the proportion of the number of units of each type of power source in the total number of units. The total number of units of the power generation side in the regional power system The theoretical maximum number of connected unit pairs in the power source cluster of the power generation side is calculated , and the formula is: , and the power source cluster connectivity is obtained by the ratio of the number of actually connected unit pairs in the power source cluster to the theoretical maximum number of connected unit pairs. The main grid access rate is calculated by the ratio of the number of power source nodes accessing the main grid to the total number of power source nodes.

[0101] In combination with the power system topology analysis technology and the graph algorithm, the electrical connection characteristics of the power grid side topology are adapted, the regional power grid topology is characterized, the number of transmission lines actually existing in the power grid side in the regional power system, the number of substations of different voltage levels, the total number of substations, the number of lines forming a loop network structure, the total number of transmission lines, the number of standby transmission lines and the number of operating lines are obtained;

[0102] The total number of nodes of the power grid side in the regional power system is utilized The theoretical maximum possible number of lines of the power grid side in the regional power system is calculated The calculation formula is as follows: The proportion of the number of transmission lines actually existing in the theoretical maximum possible number of lines is calculated, and the line connection density is obtained; the proportion of the number of substations of different voltage levels to the total number of substations is calculated, and the substation level coefficient is obtained; the ratio of the number of lines forming a loop network structure to the total number of transmission lines is calculated, and the grid loop network rate is obtained; the ratio of the number of standby transmission lines to the number of operating lines is calculated, and the standby line redundancy rate is obtained;

[0103] Through the user topology analysis technology and the attribute classification algorithm, the characteristics of the regional user topology are extracted, and the number of users of different voltage levels connected to the user side in the regional power system, the total number of users and the number of users with new energy power generation on the user side are obtained;

[0104] The proportion of the number of users of different voltage levels connected to the user side in the total number of users is calculated, and the user distribution network access rate is obtained; the ratio of the number of users with new energy power generation on the user side to the total number of users is calculated, and the self-use new energy coverage rate is obtained;

[0105] The power source type proportion is the proportion of the number of units of each type of power source to the total number of units on the power generation side in the regional power system; the power source cluster connectivity is the ratio of the number of actually connected unit pairs in the power source cluster on the power generation side to the theoretical maximum number of connected unit pairs; the main grid access rate is the ratio of the number of power source nodes connected to the main grid on the power generation side in the regional power system to the total number of power source nodes;

[0106] The line connection density is the ratio of the number of transmission lines actually existing in the power grid side in the regional power system to the theoretical maximum possible number of lines; the substation level coefficient is the proportion of the number of substations of different voltage levels in the power grid side in the regional power system to the total number of substations; the grid loop network rate is the ratio of the number of lines forming a loop network structure in the power grid side in the regional power system to the total number of transmission lines; the standby line redundancy rate is the ratio of the number of standby transmission lines in the power grid side in the regional power system to the number of operating lines;

[0107] The user network access rate is the proportion of the number of users of different voltage levels accessing the user side in the total number of users in the regional power system; the self-use new energy coverage rate is the proportion of the number of users with new energy power generation on the user side in the total number of users. These explained regional topological feature data are dimensionless data.

[0108] In step 103, a simulation calculation model is constructed by a neural network algorithm, and the process of outputting the regional power system carbon emission fine accounting simulation data includes:

[0109] The regional power system carbon emission fine accounting simulation data includes the simulation calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission;

[0110] The preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission are screened out from the distributed storage data, the preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission and the regional topological feature data are integrated, a fine accounting simulation data set is generated, and the fine accounting simulation data set is divided into a first training set and a first test set according to a 7:3 ratio;

[0111] The neural network algorithm is adopted, the first training set data is taken as the input data, the regional power system carbon emission fine accounting simulation data is taken as the output data, the nonlinear relationship between the preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission, the regional topological feature data and the simulation calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission is learned, and a trained simulation calculation model is obtained;

[0112] The first test set data is input into the trained simulation calculation model, a stochastic gradient descent optimizer is adopted to adjust the parameters of the simulation calculation model, the performance of the simulation calculation model is optimized, a final simulation calculation model is obtained, and corresponding regional power system carbon emission fine accounting simulation data is output in combination with the current preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission and the regional topological feature data.

[0113] In step 104, the simulation calculation correction model is constructed by integrating the environmental data and the random forest algorithm, the correction coefficient is output, and the process of correcting the regional power system carbon emission fine accounting simulation data includes:

[0114] The environmental data is integrated to generate a simulation calculation correction data set, and the simulation calculation correction data set is divided into a second training set and a second test set according to an 8:2 ratio;

[0115] The random forest algorithm is used to learn the nonlinear relationship between the real-time temperature, real-time humidity, altitude, real-time precipitation pH, real-time wind speed, real-time light intensity, and real-time air density of the regional power system environment and the correction coefficient by taking the second training set data as input variables and the correction coefficient as output variables, and a trained simulation calculation correction model is obtained.

[0116] The second test set data is input into the trained simulation calculation correction model, the simulation calculation correction model parameters are adjusted, the simulation calculation correction model performance is optimized, and a final simulation calculation correction model is obtained.

[0117] The current environmental data is input into the simulation calculation correction model, and the correction coefficient is output. The correction coefficient is used to correct the regional power system carbon emission fine accounting simulation data.

[0118] The specific correction process includes: the correction coefficient is between 0 and 1. When the correction coefficient is less than 0.1, the regional power system carbon emission fine accounting simulation data is not corrected.

[0119] When the correction coefficient is greater than or equal to 0.1 and less than 0.4, the simulation calculation result of the carbon emission amount of the power generation side is corrected. The specific correction process is as follows: when the correction coefficient is less than or equal to 0.3, the simulation calculation result of the carbon emission amount of the power generation side is corrected according to the formula: when the correction coefficient is greater than 0.3, the simulation calculation result of the carbon emission amount of the power generation side is not corrected. The simulation calculation result of the carbon emission amount of the power generation side after correction is The simulation calculation result of the carbon emission amount of the power generation side is E is the correction coefficient.

[0120] When the correction coefficient is greater than or equal to 0.4 and less than 0.7, the simulation calculation result of the carbon emission amount of the power grid side is corrected. The specific correction process is as follows: when the correction coefficient is less than or equal to 0.5, the simulation calculation result of the carbon emission amount of the power grid side is corrected according to the formula: when the correction coefficient is greater than 0.5, the simulation calculation result of the carbon emission amount of the power grid side is not corrected. The simulation calculation result of the carbon emission amount of the power grid side after correction is The simulation calculation result of the carbon emission amount of the power grid side is

[0121] When the correction coefficient is greater than or equal to 0.7 and less than or equal to 1, the simulation calculation result of the carbon emission amount of the user side is corrected. The specific correction process is as follows: when the correction coefficient is less than or equal to 0.9, the simulation calculation result of the carbon emission amount of the user side is corrected according to the formula: when the correction coefficient is greater than 0.9, the simulation calculation result of the carbon emission amount of the user side is not corrected. ​The formula is used to correct the simulation calculation result of the user side carbon emission. When the correction coefficient is greater than 0.9, the simulation calculation result of the user side carbon emission is not corrected. The simulation calculation result of the user side carbon emission after correction is The simulation calculation result of the user side carbon emission is

[0122] In step 105, the process of outputting the corrected regional power system carbon emission fine accounting simulation data in the form of data report includes:

[0123] The dimensions and formats of the preset data report are classified according to the power generation side, the power grid side and the user side. Each category contains the simulation calculation result of the corrected carbon emission;

[0124] The simulation calculation results of the corrected carbon emission of the power generation side, the power grid side and the user side are integrated, and they are filled into the data report one by one to generate the regional power system carbon emission fine accounting simulation calculation result output report in the preset report format.

[0125] As shown in Figures 2-4 The regional power system carbon emission accounting full cycle comparison chart is a static comparison chart, Figures 2-4 which respectively shows the changes of carbon emissions of the power generation side, the power grid side and the user side, and clearly presents the time sequence differences of the carbon emission preliminary calculation result, the carbon emission accounting simulation calculation result and the corrected carbon emission accounting simulation calculation result in the power generation side, the power grid side and the user side;

[0126] Figures 2-4 The horizontal axis is a 24-hour cycle, and the vertical axis is the carbon emission. The visualization comparison of the carbon emission preliminary calculation result, the carbon emission accounting simulation calculation result and the corrected carbon emission accounting simulation calculation result in the power generation side, the power grid side and the user side is realized by different colors and different line types of lines;

[0127] Figures 2-4 It intuitively reflects the whole process effect from preliminary calculation to neural network simulation calculation and then to random forest correction in the method of the embodiment, and the corrected data curve is more stable and accurate, which verifies the promotion effect of topological features and environmental factors on the accounting accuracy, and meets the goal of full cycle, multi-dimensional fine accounting.

[0128] Figure 5 The simulation calculation model training performance iteration chart is mainly for the visualization output of the training performance in the content of constructing the simulation calculation model by the neural network algorithm in the embodiment;

[0129] The horizontal axis is the training round, and there are 8 rounds. The vertical axis contains performance indicators and error values, which shows the training process of the neural network simulation calculation model.Figure 5 The training, validation, test and best four curves of the clear annotation, the key information shows that the fourth round of training reaches the best performance (the best point circled by the green circle, that is, the intersection of the two best lines), and the error value approaches 10 as the round progresses -5 The order of magnitude verifies the training effectiveness and convergence of the simulation model, and proves that the model can stably learn the preliminary calculation results of the carbon emissions of the power generation side, the power grid side and the user side, the topological feature data and the nonlinear relationship between the regional power system carbon emission fine accounting simulation data.

Claims

1. A method for simulating and calculating fine accounting of carbon emissions of a regional power system, characterized in that, The method comprises the following steps: Collecting regional power system data, including power generation side data, power grid side data, user side data, regional geographic information data, environmental data and topology data, the topology data including regional power source topology, regional power grid topology and regional user topology, and preprocessing the regional power system data; Based on the power generation side data, the power grid side data and the user side data, the carbon emissions of the power generation side, the power grid side and the user side are preliminarily calculated, the carbon emission accounting boundaries of the power generation side, the power grid side and the user side in the regional power system are determined based on the regional geographic information data, and distributed storage data is obtained in combination with the distributed storage technology; Feature extraction is performed on the topology data to obtain regional topology feature data, and the process includes: The regional topology feature data includes power source type proportion, power source cluster connectivity, main grid access rate, line connection density, transformer substation level coefficient, power grid ring network rate, standby line redundancy rate, user distribution network access rate and self-use new energy coverage rate; The topology analysis technology and the graph neural network are combined to extract features from the regional power source topology, and the number of units of each type of power source in the regional power system, the total number of units, the number of actually connected unit pairs in the power source cluster, the number of power source nodes connected to the main grid and the total number of power source nodes are obtained; The power source type proportion is obtained by the proportion of the number of units of each type of power source in the total number of units, the theoretical maximum number of connected unit pairs in the power source cluster is calculated based on the total number of units in the regional power system, and the power source cluster connectivity is obtained by the ratio of the number of actually connected unit pairs in the power source cluster to the theoretical maximum number of connected unit pairs; The topology analysis technology and the graph theory algorithm are combined to adapt to the electrical connection characteristics of the power grid side topology, features are extracted from the regional power grid topology, and the number of actual existing transmission lines in the regional power system, the number of substations of different voltage levels, the total number of substations, the number of lines forming a ring network structure, the total number of transmission lines, the number of standby transmission lines and the number of operating lines are obtained; The theoretical maximum possible number of lines in the regional power system is calculated based on the total number of nodes in the power grid side, the proportion of the number of actual existing transmission lines in the theoretical maximum possible number of lines is calculated, and the line connection density is obtained; the proportion of the number of substations of different voltage levels in the total number of substations is calculated to obtain the transformer substation level coefficient; the power grid ring network rate is calculated by the ratio of the number of lines forming a ring network structure to the total number of transmission lines; the standby line redundancy rate is obtained by the ratio of the number of standby transmission lines to the number of operating lines; The user topology analysis technology and the attribute classification algorithm are combined to extract features from the regional user topology, and the number of users of different voltage levels connected to the user side in the regional power system, the total number of users and the number of users with new energy generation on the user side are obtained; The user distribution network access rate is calculated by the proportion of the number of users of different voltage levels connected to the user side in the total number of users; the self-use new energy coverage rate is obtained by the ratio of the number of users with new energy generation on the user side to the total number of users. In combination with the distributed storage data and the neural network algorithm, a simulation calculation model is constructed to output the simulation data of the fine accounting of the carbon emission of the regional power system; In combination with the comprehensive environmental data and the random forest algorithm, a simulation calculation correction model is constructed to output the correction coefficient, and then the simulation data of the fine accounting of the carbon emission of the regional power system is corrected; In the form of data reports, the simulation data of the fine accounting of the carbon emission of the regional power system after correction is output.

2. The regional power system carbon emission fine accounting simulation calculation method according to claim 1, characterized in that, The collection process of the regional power system data includes: Different types of collection devices are deployed to collect the power generation side data, the power grid side data, the user side data, the regional geographic information data, the environmental data and the topology data, so as to obtain the regional power system data; The power generation side data includes the unit fuel consumption, the carbon content of the unit heat value of the fuel, the low heat value of the fuel, the total carbon emission and the total power grid power of all the power generation sides in the region; the power grid side data is the power transmission loss; the user side data includes the user terminal power consumption and the user self-use new energy power generation; the regional geographic information data includes the site area and the boundary latitude and longitude coordinates of the power plant area, the path direction of the power transmission line, the site area and the boundary latitude and longitude coordinates of the transformer substation area, and the site area and the boundary latitude and longitude coordinates of the user aggregation area; the environmental data includes the real-time temperature, the real-time humidity, the altitude, the real-time precipitation pH value, the real-time wind speed, the real-time light intensity and the real-time air density of the regional power system environment.

3. The regional power system carbon emission fine accounting simulation calculation method according to claim 1, characterized in that, The pre-processing process of the regional power system data includes: The collected power generation side data, power grid side data, user side data, regional geographic information data and environmental data are subjected to data cleaning; The power generation side data, power grid side data and user side data are subjected to time sequence alignment processing in a unified time format, so as to form a full-cycle carbon emission accounting data link of power generation, power transmission and power consumption; Through a graphical tool, the structures of the regional power source topology, the regional power grid topology and the regional user topology are visualized, and the regional power source topology, the regional power grid topology and the regional user topology are subjected to redundant removal and structure optimization processing by using the graph theory and the topology optimization algorithm.

4. The regional power system carbon emission fine accounting simulation calculation method according to claim 1, characterized in that, The preliminary calculation process of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission includes: The power generation side carbon emission and the average power supply carbon emission factor of the regional power system are preliminarily calculated by using the power generation side data; The power grid side carbon emission is preliminarily calculated based on the average power supply carbon emission factor of the regional power system and in combination with the power grid side data; The user side carbon emission is preliminarily calculated based on the average power supply carbon emission factor of the regional power system and in combination with the user side data.

5. The regional power system carbon emission fine accounting simulation calculation method according to claim 2, characterized in that, The process of determining the carbon emission accounting boundary of the power generation side, the power grid side and the user side in the regional power system based on the regional geographic information data includes: Through the GIS technology, a GIS map is initialized, the coordinate system of the GIS map is unified, the power generation side region is identified according to the site area and the boundary latitude and longitude coordinates of the power plant area, and the site area and the boundary latitude and longitude coordinates of the power plant area are superimposed on the GIS map to determine the carbon emission accounting boundary of the power generation side in the regional power system; According to the path of the power transmission line, the path of the power transmission line is drawn in the GIS map, and the area and boundary latitude and longitude coordinates of the substation area are superimposed on the GIS map to determine the carbon emission accounting boundary of the power grid side in the regional power system; Referring to the area and boundary latitude and longitude coordinates of the user aggregation area, the user area is drawn in the GIS map, and then the carbon emission accounting boundary of the user side in the regional power system is determined.

6. The regional power system carbon emissions fine accounting simulation calculation method according to claim 1, characterized in that, The process of obtaining distributed storage data by using distributed storage technology includes: Using distributed storage technology, three independent storage partitions are established based on the carbon emission accounting boundaries of the power generation side, the power grid side and the user side in the regional power system, and the geographical range and data ownership of each independent storage partition are clearly defined; According to the carbon emission accounting boundaries of the power generation side, the power grid side and the user side in the corresponding regional power system and the time dimension, the preliminary calculation results of the carbon emissions of the power generation side, the power grid side and the user side are respectively split into several data shards, and the data format of each data shard after unified splitting processing is unified; According to the size and access frequency of the data shards, the data shards are distributed to different nodes in the distributed storage cluster, and a replica mechanism is used to store each data shard, thereby realizing the establishment of the storage partition; Geospatial index and attribute index are established for each storage partition, the integrity of the data stored by each node is verified through data shard hash value comparison and metadata integrity checking technology, the correspondence between the data and the carbon emission accounting boundaries of the power generation side, the power grid side and the user side in the regional power system is confirmed to be correct, and the distributed storage of the preliminary calculation results of the carbon emissions of the power generation side, the power grid side and the user side is realized, thereby obtaining the distributed storage data.

7. The regional power system carbon emissions fine accounting simulation calculation method according to claim 1, characterized in that, The process of constructing a simulation calculation model by using a neural network algorithm to output regional power system carbon emission fine accounting simulation data includes: The regional power system carbon emission fine accounting simulation data includes simulation calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission; The preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission are selected from the distributed storage data, the preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission, and the regional topology feature data are integrated to generate a fine accounting simulation data set, and the fine accounting simulation data set is divided into a first training set and a first test set; Using a neural network algorithm, the first training set data is used as input data, and the regional power system carbon emission fine accounting simulation data is used as output data, the nonlinear relationship between the preliminary calculation results of the power generation side carbon emission, the power grid side carbon emission and the user side carbon emission, the regional topology feature data and the regional power system carbon emission fine accounting simulation data is learned, and a trained simulation calculation model is obtained. The first test set data is input into the trained simulation calculation model, the random gradient descent optimizer is used to adjust the parameters of the simulation calculation model, the performance of the simulation calculation model is optimized, the final simulation calculation model is obtained, and the preliminary calculation results of the current power generation side carbon emission, the power grid side carbon emission and the user side carbon emission and the regional topology feature data are combined to output the corresponding regional power system carbon emission fine accounting simulation data.

8. The regional power system carbon emissions fine accounting simulation calculation method according to claim 1, characterized in that, The process of correcting the regional power system carbon emission fine accounting simulation data by combining the environmental data and the random forest algorithm to build a simulation calculation correction model and output a correction coefficient includes: Integrate the environmental data to generate a simulation calculation correction data set, and divide the simulation calculation correction data set into a second training set and a second test set; Using the random forest algorithm, the second training set data is used as the input variable, the correction coefficient is used as the output variable, the nonlinear relationship between the environmental data and the correction coefficient is learned, and a trained simulation calculation correction model is obtained; The second test set data is input into the trained simulation calculation correction model, the parameters of the simulation calculation correction model are adjusted, the performance of the simulation calculation correction model is optimized, and the final simulation calculation correction model is obtained; The current environmental data is input into the simulation calculation correction model, the correction coefficient is output, the correction coefficient is referred to, and the regional power system carbon emission fine accounting simulation data is corrected.

9. The regional power system carbon emission fine accounting simulation calculation method according to claim 7, characterized in that, In the form of data report, the process of outputting the corrected regional power system carbon emission fine accounting simulation data includes: The dimensions and formats of the data report are preset, classified according to the power generation side, the power grid side and the user side, and each category contains the corresponding corrected carbon emission simulation calculation result; The simulation calculation results of the corrected power generation side carbon emission, the power grid side carbon emission and the user side carbon emission are integrated, and they are filled into the data report one by one to generate the regional power system carbon emission fine accounting simulation calculation result output report in the preset report format.

Images