A method for calculating the full life cycle carbon emissions of a photovoltaic module

Abstract

This invention provides a method for calculating the carbon emissions of a photovoltaic (PV) module throughout its entire lifecycle, relating to the field of carbon emission calculation technology. The method first acquires energy data for each stage of the PV module's lifecycle; then it calculates the carbon emissions per unit product during silicon mining, industrial silicon production, polysilicon production, silicon wafer production, cell production, the PV module itself, the PV module's operation during power generation, and the PV module's retirement and disposal; finally, it calculates the total carbon emissions per unit product over its entire lifecycle. This method fully considers the power loss of the PV module during use and, combined with a neural network model, utilizes an improved evolutionary algorithm to optimize the initial weights and thresholds, effectively avoiding systematic errors caused by model weights, reducing calculation errors, and improving the accuracy of carbon emission calculations.

Description

Technical Field

[0001] This invention relates to the field of carbon emission calculation technology, and in particular to a method for calculating the carbon emissions of a photovoltaic module throughout its entire life cycle. Background Technology

[0002] Key initiatives in the power sector include implementing renewable energy substitution programs, deepening power system reform, and building a new power system with new energy sources as the mainstay. Renewable energy sources, represented by wind and solar power, are clean, safe, reliable, and not limited by geographical distribution. Traditionally, solar power and other new energy sources are considered completely clean and do not produce any carbon emissions. However, the production and maintenance of solar modules involve numerous additional steps, such as raw material processing, energy use, and labor, inevitably leading to carbon emissions.

[0003] Existing methods for calculating carbon emissions mainly include the material balance method, the measured method, and the carbon emission factor method. The material balance method requires comprehensive statistical accounting of input and output within the industry, but lacks clear records of energy consumption for input and output; therefore, this method is not suitable for calculating carbon emissions throughout the entire lifecycle of photovoltaic modules. The measured method uses direct measurement to calculate carbon emissions from products, which is difficult to implement for the entire lifecycle of photovoltaic modules. The carbon emission factor method mainly calculates carbon emissions from product production based on the carbon emission coefficients of different fossil fuel combustion, making it suitable for industries primarily reliant on energy consumption, but it lacks clearly defined calculation boundaries.

[0004] Existing technologies for calculating carbon emissions throughout the entire lifecycle of photovoltaics have the following shortcomings: First, existing technologies mostly estimate carbon emissions during the production stage of photovoltaic modules, but in reality, the calculation of carbon emissions from photovoltaic modules does not fully consider the power loss during the use of photovoltaic modules, resulting in errors in the calculation of equivalent carbon emissions from photovoltaic modules. Second, existing technologies mostly use the carbon emission factor method to estimate the carbon emissions of corresponding energy sources, but the carbon emission factors used are mostly uniform carbon emission factors for the corresponding provinces. However, the carbon emission factors of different regions within a province are different due to the differences in energy consumption in time and space, which also increases the error of the carbon emission assessment results of photovoltaic modules to a certain extent.

[0005] For example, Chinese patent publication number CN111369114B discloses a method for obtaining carbon emissions from the photovoltaic power generation industry based on its entire life cycle. This method divides the total carbon emissions of the photovoltaic power generation industry into four stages: carbon emissions during the photovoltaic system production stage, carbon emissions during the photovoltaic system operation and maintenance stage, carbon emissions during the photovoltaic system power transmission loss stage, and carbon emissions during the photovoltaic system retirement stage. While this patent calculates carbon emissions for different photovoltaic production stages to obtain the effective carbon emissions of the photovoltaic power generation industry, its carbon emission estimation for photovoltaic modules does not consider the carbon emissions from power loss during the use of the photovoltaic modules. Therefore, its carbon emission estimation results contain errors. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a method for calculating the carbon emissions of photovoltaic modules throughout their entire life cycle, in order to overcome the shortcomings of the prior art.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for calculating the carbon emissions of a photovoltaic module throughout its entire life cycle, comprising:

[0008] Acquire energy data at each stage of the entire lifecycle of photovoltaic modules;

[0009] Calculate the carbon emissions per unit of product: C1 during silica mining, C2 during industrial silicon production, C3 during polysilicon production, C4 during silicon wafer production, C5 during solar cell production, C6 during photovoltaic modules, C7 during photovoltaic module power generation operation, and C8 during photovoltaic module retirement and scrapping.

[0010] Calculate the total carbon emissions C of a unit photovoltaic module over its entire life cycle. 总 =C1+C2+C3+C4+C5+C6+C7+C8.

[0011] Preferably, the energy data at each stage of the photovoltaic module's entire life cycle is collected by a monitoring unit, specifically:

[0012] Utilizing satellite remote sensing intelligent identification technology, the system identifies mining conditions within silica mining areas, including mine size, type, and transportation conditions; acquires transportation energy consumption and distance for each batch of silica over different time periods; obtains time-segmented energy data for mining enterprises via IoT transmission; acquires silica production data; uses the factory areas of industrial silicon, polysilicon, silicon wafers, solar cells, and modules as boundaries, and acquires time-segmented energy data and industrial silicon production data for these enterprises via IoT transmission; and captures unit product energy consumption and carbon emission data for scrap steel and aluminum recycling in the market, acquiring time-segmented energy data for mining enterprises via IoT transmission.

[0013] Preferably, the carbon emissions per unit product of silica mining are calculated as follows:

[0014] Step S1. Collect all electrical and oil information data during the silica mining process to form a basic database;

[0015] Step S2. Construct a carbon emission calculation model under active / reactive power in the silica mining process, and use the electricity and oil information data collected in step S1 to calculate the carbon emission under active / reactive power in the silica mining process, update the basic database to form an electricity and oil information database, and divide the training set and test set.

[0016] The carbon emission calculation model for the active / reactive power of the silica mining process is shown in the following formula:

[0017]

[0018] Among them, C L This indicates the carbon emissions during the active / reactive power generation of the silica mining process. Q represents the consumption coefficient of the i-th material; M，i This represents the annual consumption of the i-th material; CF M，i This represents the carbon emission coefficient when using the i-th material;

[0019] Step S3. Construct a BP neural network. Input the oil information data from the sample data into the input layer of the BP neural network model. Then calculate the error between the output result of the output layer of the BP neural network model and the unit carbon emission of the sample data. Using the corresponding training set and test set in S2, train and test to obtain the carbon emission factor model under active / reactive power in the silica mining process. Based on the model, give the corresponding carbon emission factor under unit active / reactive power in the silica mining process when using oil of different quality.

[0020] Step S4. Construct an online calculation model for the carbon emissions from power generation during the silica mining process, as shown in the following formula:

[0021] C s=E×CF

[0022] Among them, C s E represents the carbon emissions from power generation during the silica mining process; E represents the total electrical energy consumed during the production phase of the silica mining process; CF represents the carbon emission coefficient of electrical energy during the production phase of the silica mining process.

[0023] Step S5. Collect power generation data of the silica mining process online, and distinguish between active and reactive power; combine the corresponding carbon emission factors obtained in S3, and calculate the corresponding carbon emissions of the silica mining process online based on the online calculation model of carbon emissions of power generation during silica mining.

[0024] The carbon emission calculation model for the silica mining process is as follows:

[0025]

[0026] C1 represents the carbon emissions from the silica mining process.

[0027] Preferably, the method for calculating carbon emissions during the industrial silicon production process is as follows:

[0028] Obtain the total operating load segment of industrial silicon production over a certain operating period; based on the total operating load segment and the preset baseline load point, determine all characteristic load points of industrial silicon production within the total operating load segment; based on the actual carbon content of the coal fed into the furnace and the actual characteristic energy consumption of each characteristic load point, determine the actual carbon emissions of each characteristic load point; based on the actual carbon emissions of all characteristic load points, determine the actual total carbon emissions of industrial silicon production over a certain operating period, as shown in the following formula:

[0029]

[0030] Among them, C p The actual carbon emissions for all characteristic load points are shown in the following formula:

[0031] C p =B′ p ×CF b

[0032] Among them, B′ p The actual characteristic energy consumption at the characteristic load point is shown in the following formula:

[0033] B′ p =B P +B P ×Ω P +B P ×σ P +B P ×R P +B P ×τP

[0034] Among them, CF b B represents the actual carbon emission coefficient at the characteristic point. P Coal consumption for power supply at characteristic load points; Ω P σ is the coefficient representing the influence of heat consumption on energy consumption. P R is the coefficient representing the impact of combustion efficiency on energy consumption. P τ is the coefficient representing the influence of coal utilization rate on energy consumption. P This is the coefficient representing the impact of plant power consumption rate on energy consumption.

[0035] Preferably, the method for calculating carbon emissions during the polysilicon production process is as follows:

[0036] First, determine the total amount of each raw material used in the polysilicon production process and its carbon dioxide equivalent emission factor per unit of production. Then, determine the carbon dioxide equivalent emission factors for each type of energy source. Next, determine the carbon emission equivalent coefficient of 1 kWh of electricity in the polysilicon production stage and the carbon emission equivalent coefficient of 1 kWh of electricity in the polysilicon transportation stage. Further calculate the carbon emissions in the polysilicon production stage and the polysilicon transportation stage. Therefore, the total carbon emissions in the polysilicon production process are the sum of the carbon emissions from the polysilicon production and polysilicon transportation stages, as shown in the following formula:

[0037] C3=∑EC m1s ×k 1s +∑ET 1s ×k 2s

[0038] Where s = 0, 1, 2, 3...N, and N represents the number of polycrystalline silicon types; EC m1s k represents the energy consumption during the polysilicon production stage. 1s The carbon emission equivalent factor for 1 kWh of electricity generated during the energy acquisition phase in the polysilicon production process; ET 1s k represents the energy consumption value during the transportation stage of polysilicon acquisition. 2s The carbon emission equivalent coefficient for 1 kWh of electricity used in the transportation of polysilicon.

[0039] Preferably, the method for calculating carbon emissions during the silicon wafer production process is as follows:

[0040] In the silicon wafer production process, the total amount of each raw material used and its unit production carbon dioxide equivalent emission factor are determined; the carbon dioxide equivalent emission factors of various energy sources are determined, and a digital twin model database is established; a carbon factor library for calculating carbon emissions from silicon wafer production is established; a life cycle model tree for silicon wafer production is generated based on the digital twin model database and the carbon factor library; a carbon emission data calculation model is established based on each stage of silicon wafer production; and carbon emissions at each stage of silicon wafer production are calculated based on the carbon emission data calculation model and the life cycle model tree for silicon wafer production, as shown in the following formula:

[0041] C4=(∑ED m2e +∑EF m2 +∑EL m2 +∑EM m2 +∑EN m2 )×k3+T m2 ×k4

[0042] C4 represents the carbon emissions during silicon wafer production, and ED... m2e The energy consumption value when performing a segment manufacturing process for the e-th type of silicon wafer raw material, where e = 1, 2, 3...u, and u represents the type and quantity of silicon wafer raw materials; EF m2 Energy consumption value during the assembly phase; EL m2 Energy consumption during the decomposition and capacity-building process; EM m2 Energy consumption value when executing grouped segments; EN m2 The energy consumption value during the integration phase; k3 is the carbon emission equivalent coefficient of 1 kWh of electricity used in silicon wafer production; T m2 k is the energy consumption value during silicon wafer transportation; k4 is the carbon emission equivalent coefficient of using 1 kWh of electricity during silicon wafer transportation.

[0043] Preferably, the method for calculating carbon emissions during the battery cell production process is as follows:

[0044] A analytic hierarchy process (AHP) is used to construct an energy and electricity carbon emission index system to measure the carbon emission level of the solar cell production process. Carbon emission models for the solar cell production process, the power transmission system, and the electricity consumption side are constructed to calculate carbon emission data during solar cell production, power transmission, and power generation. Historical carbon emission data for the solar cell production process is also calculated. The temporal characteristics of historical carbon emission data are obtained using the Empirical Mode Decomposition (EMD) method. An LSTM (Long Short-Term Memory) network is trained using this historical carbon emission data to calculate carbon emission data for power generation and transmission during solar cell production, as shown in the following formula:

[0045] C5 = f θ (P1, P2, P3, P4, X)

[0046] C5 represents the carbon emissions during the cell manufacturing process, f θ For the constructed LSTM model, θ represents the network structure parameters of the LSTM model, including the time period of memory; X represents historical carbon emission data; P1, P2, P3, and P4 represent the high-frequency, medium-frequency, low-frequency, and extremely low-frequency power characteristic components of power generation and transmission during the cell production process, respectively.

[0047] Preferably, the carbon emissions of the photovoltaic module are calculated as follows:

[0048] The photovoltaic module is divided into three stages: building material and component production, transportation and operation, and component dismantling and reuse. The calculation boundaries for each stage are defined, and the carbon emissions of each stage are added together to obtain the total carbon emissions C6 of the module, as shown in the following formula:

[0049] C6 = EM m4 ×k6+ET m4 ×k7+T m3 ×k8

[0050] Among them, EM m4 k6 represents the energy consumption during the production of building materials and components for photovoltaic modules; k6 is the carbon emission equivalent coefficient of using 1 kWh of electricity during the production of building materials and components for photovoltaic modules; ET m4 K represents the energy consumption during the transportation and operation of photovoltaic modules; k7 represents the carbon emission equivalent coefficient of using 1 kWh of electricity during the transportation and operation of photovoltaic modules; T m3 k is the energy consumption value during the component dismantling and reuse stage of photovoltaic modules; k8 is the carbon emission equivalent coefficient of using 1 kWh of electricity during the component dismantling and reuse stage of photovoltaic modules.

[0051] Preferably, the carbon emissions from the power generation operation are calculated as follows:

[0052] First, obtain the carbon footprint assessment parameters for power generation operation in the target project, as well as the basic data on power generation operation related to carbon emissions. Then, use the Bayesian network method to calculate the data deviation between the basic power generation operation data and the actual carbon emission data. Next, construct a plan review technology distribution based on the basic power generation operation data and the upper and lower limits of the deviation value to fit the data distribution of the basic power generation operation data. Simplify the continuously changing carbon footprint assessment parameters into a triangular distribution to simulate the parameter distribution of the carbon footprint assessment parameters. Finally, combine the data distribution of the basic power generation operation data and the parameter distribution of the carbon footprint assessment parameters, and calculate the carbon footprint distribution of the power generation operation in the target project using Monte Carlo simulation. Based on the carbon footprint distribution of the power generation operation, calculate the carbon emissions of the target project's power generation operation, as shown in the following formula:

[0053]

[0054] Among them, PF u，v GWP represents the carbon emission factor of the v-th power generation operation using material u; v This represents the carbon emission loss coefficient for the v-th type of power generation operation.

[0055] Preferably, the carbon emissions from the decommissioning and scrapping of the photovoltaic modules are calculated as follows:

[0056] The process involves: acquiring historical carbon emission event data for photovoltaic (PV) modules; calculating retirement and scrapping data based on historical carbon emission time data; updating the model parameters of the constructed power grid BP neural network model until the variance E between the carbon emissions obtained from the current parameters and the expected carbon emissions is less than a set value; obtaining the determined model parameters and the trained power grid BP neural network model; the model parameters include the weight values ​​of neurons and the normalized carbon emission influence parameters for each group; acquiring the electricity consumption for retirement and scrapping at the current time point and the reference electricity consumption, inputting them into the trained power grid BP neural network model; and determining the carbon emissions of PV module retirement and scrapping based on the output of the power grid BP neural network model, as shown in the following formula:

[0057]

[0058] C8 represents the carbon emissions from decommissioning and scrapping, while AD... z Calculate the decommissioning and scrapping number for the z-th type of historical carbon emissions, EF z Calculate the carbon emission coefficient for decommissioning and scrapping for the z-th type of historical carbon emission;

[0059] The power grid BP neural network model includes:

[0060] A carbon emission neuron computational mathematical model building unit is used to construct a carbon emission neuron computational mathematical model. The parameters involved in training the neuron computational mathematical model include: the weight values ​​W from neuron p to neuron q. pq The input information x received from neuron p at time t p (t), where f(x) is the preset neuron transfer function;

[0061] The input layer acquisition unit is used to obtain the normalized carbon emission impact parameters of each group as the input layer vector x = (x1, x2, ... x) of the BP neural network based on the BP neural network structure. p , ...x P ), where P is the number of neurons in the input layer;

[0062] Hidden layer acquisition unit, used to calculate the mathematical model of carbon emission neurons, and the input layer vector x = (x1, x2, ... x) corresponding to each group of normalized carbon emission influence parameters contained in the training samples. p , ...x PTo obtain the hidden layer vector y = (y1, y2, ..., y3) of the BP neural network. q , ..., y Q )×W pq , where y q W is the q-th neuron in the Q neurons of the hidden layer. pq The vector represents the weights from the p-th neuron in the input layer to the q-th neuron in the hidden layer.

[0063] The output layer acquisition unit is used to calculate the carbon emission budget value 'a' of the output layer of the BP neural network corresponding to each group of normalized carbon emission influence parameters, based on the hidden layer vector and the preset neuron transfer function f(x), where W qk represents the weight values ​​from the q-th neuron in the hidden layer to the k-th neuron in the output layer, where k takes the value of 1;

[0064] The output error calculation unit is used to obtain the actual carbon emission value b corresponding to the normalized carbon emission influence parameter for each group based on the normalized carbon emission data contained in the training samples, and to calculate the corresponding output error for each group of normalized carbon emission influence parameters.

[0065] The parameter error acquisition and judgment unit takes the sum of the squares of the output errors err corresponding to the normalized carbon emission impact parameters of each group as the parameter error, and judges whether the parameter error is within the preset error allowable range; if yes, it outputs the carbon emission data; if no, it adjusts the weight values ​​of neurons p to q and recalculates.

[0066] The beneficial effects of adopting the above technical solution are as follows: The present invention provides a method for calculating the carbon emissions of photovoltaic modules throughout their entire life cycle. This method calculates the total carbon emissions of a photovoltaic module throughout its entire life cycle by calculating the carbon emissions per unit of product, including carbon emissions from silicon mining, industrial silicon, polycrystalline silicon, silicon wafers, solar cells, photovoltaic modules, photovoltaic module power generation and operation, and photovoltaic module retirement and scrapping. This method fully considers the power loss of photovoltaic modules during use and combines an improved evolutionary algorithm with a neural network model to optimize the initial weights and thresholds. This effectively avoids systematic errors caused by the influence of model weights, allowing for correction and compensation, reducing calculation errors, and improving the accuracy of carbon emission calculation. Attached Figure Description

[0067] Figure 1 A block diagram illustrating a method for calculating the carbon emissions of a photovoltaic module throughout its entire life cycle, as provided in an embodiment of the present invention.

[0068] Figure 2This is a flowchart of a method for calculating carbon emissions during silica mining, provided in an embodiment of the present invention.

[0069] Figure 3 A flowchart illustrating the method for calculating carbon emissions during industrial silicon production, as provided in this embodiment of the invention.

[0070] Figure 4 A flowchart illustrating the method for calculating carbon emissions during polysilicon production, as provided in this embodiment of the invention.

[0071] Figure 5 A flowchart illustrating the method for calculating carbon emissions during silicon wafer production, as provided in this embodiment of the invention.

[0072] Figure 6 A flowchart illustrating the method for calculating carbon emissions during the battery cell production process provided in this embodiment of the invention;

[0073] Figure 7 A flowchart illustrating the method for calculating carbon emissions from photovoltaic module power generation operation, as provided in this embodiment of the invention. Detailed Implementation

[0074] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0075] In this embodiment, a method for calculating the carbon emissions of a photovoltaic module throughout its entire life cycle is described, such as... Figure 1 As shown, firstly, energy data for each stage of the photovoltaic module's entire lifecycle is obtained; then, the carbon emissions per unit product are calculated as follows: C1 during silica mining, C2 during industrial silicon production, C3 during polysilicon production, C4 during silicon wafer production, C5 during cell production, C6 during photovoltaic module operation, C7 during photovoltaic module decommissioning and scrapping, and C8 during photovoltaic module retirement and disposal; finally, the total carbon emissions per unit product photovoltaic module throughout its entire lifecycle, C, are calculated. 总 =C1+C2+C3+C4+C5+C6+C7+C8.

[0076] I. Energy data for each stage of the photovoltaic module's entire lifecycle is collected using monitoring units, specifically:

[0077] Utilizing satellite remote sensing intelligent identification technology, the system identifies mining conditions within silica mining areas, including mine size, type, and transportation conditions. By installing IoT smart meters and GPS positioning on transportation vehicles, it acquires transportation energy consumption and distance for each batch of silica over various time periods. Meters are installed at energy inlets or storage areas of mining enterprises to obtain time-segmented energy data via IoT transmission. Meters are also installed on ore warehouses or transport vehicles to obtain silica production data. Furthermore, using the production areas of industrial silicon, polysilicon, silicon wafers, solar cells, and modules as boundaries, meters are installed at all energy inlets or storage areas throughout the plant to acquire time-segmented energy data via IoT transmission. Meters are also installed on industrial silicon product warehouses or transport vehicles to obtain industrial silicon production data. Finally, it captures unit product energy consumption and carbon emission data from the recycling and processing of scrap steel and aluminum in the market, installs meters at all energy inlets or storage areas for waste materials, and acquires time-segmented energy data from mining enterprises via IoT transmission.

[0078] The method for calculating carbon emissions (C1) per unit of silica mining process is as follows: Figure 2 As shown, specifically:

[0079] Step S1. Collect all electrical and oil information data during the silica mining process to form a basic database;

[0080] Step S2. Construct a carbon emission calculation model under active / reactive power in the silica mining process, and use the electricity and oil information data collected in step S1 to calculate the carbon emission under active / reactive power in the silica mining process, update the basic database to form an electricity and oil information database, and divide the training set and test set.

[0081] The carbon emission calculation model for the active / reactive power of the silica mining process is shown in the following formula:

[0082]

[0083] Among them, C L This indicates the carbon emissions during the active / reactive power generation of the silica mining process. Q represents the consumption coefficient of the i-th material, which includes silica rock, iron, and electrical energy, and can be obtained from the CLCD database; M，i This represents the annual consumption of the i-th material; CF M，i This represents the carbon emission coefficient when using the i-th material;

[0084] Step S3. Construct a BP neural network. Input the oil information data from the sample data into the input layer of the BP neural network model. Then calculate the error between the output result given by the output layer of the BP neural network model and the unit carbon emission of the sample data. Using the corresponding training set and test set in S2, train and test to obtain the carbon emission factor model under active / reactive power in the silica mining process. Based on the model, give the corresponding carbon emission factor under unit active / reactive power in the silica mining process when using oil of different quality.

[0085] Step S4. Construct an online calculation model for the carbon emissions from power generation during the silica mining process, as shown in the following formula:

[0086] C s =E×CF

[0087] Among them, C s E represents the carbon emissions from power generation during the silica mining process; E represents the total electrical energy consumed during the production stage of the silica mining process, which can be obtained from the National Energy Administration; CF represents the carbon emission coefficient of electrical energy during the production stage of the silica mining process, which can be obtained from the corresponding Department of Climate Change of the Ministry of Ecology and Environment of China.

[0088] Step S5. Collect power generation data of the silica mining process online, and distinguish between active and reactive power; combine the corresponding carbon emission factors obtained in S3, and calculate the corresponding carbon emissions of the silica mining process online based on the online calculation model of carbon emissions of power generation during silica mining.

[0089] The carbon emission calculation model for the silica mining process is as follows:

[0090]

[0091] This embodiment takes the ore mining process as an example. The material and energy consumption data for manufacturing a 1kWp photovoltaic module are shown in Table 1.

[0092] Table 1. Material and energy data in the silicon ore mining process.

[0093]

[0094] II. The calculation method for carbon emissions (C2) during industrial silicon production is as follows: Figure 3 As shown, specifically:

[0095] The main material consumption in industrial silicon production is raw material silica and various carbonaceous reducing agents; the main energy consumption is coal combustion and electricity; and waste in different states, including gases and solids, is generated. Therefore, carbon emission sources include: electricity (indirect carbon emissions), coal combustion (direct carbon emissions), and process emissions from the chemical reaction between silica and carbon. The process involves obtaining the total operating load segment of industrial silicon production over a certain period; determining all characteristic load points within the total operating load segment based on the total operating load segment and a preset baseline load point; determining the actual carbon emissions at each characteristic load point based on the actual carbon content of the coal fed into the furnace and the actual characteristic energy consumption at each characteristic load point; and finally, determining the actual total carbon emissions of industrial silicon production over a certain period based on the actual carbon emissions at all characteristic load points, as shown in the following formula:

[0096]

[0097] Among them, C p The actual carbon emissions for all characteristic load points are shown in the following formula:

[0098] C p =B′ p ×CF b

[0099] Among them, B′ p The actual characteristic energy consumption at the characteristic load point is shown in the following formula:

[0100] B′ p =B P +B P ×Ω P +B P ×σ P +B P ×R P +B P ×τ P

[0101] Among them, CF b B represents the actual carbon emission coefficient at the characteristic point. P Coal consumption for power supply at characteristic load points; Ω P σ is the coefficient representing the influence of heat consumption on energy consumption. P R is the coefficient representing the impact of combustion efficiency on energy consumption. P τ is the coefficient representing the influence of coal utilization rate on energy consumption. P This is the coefficient representing the impact of plant power consumption rate on energy consumption.

[0102] This embodiment takes the industrial silicon production process as an example. The material and energy consumption data for manufacturing a 1kWp photovoltaic module are shown in Table 2.

[0103] Table 2. Material and Energy Data in Industrial Silicon Production Process

[0104]

[0105] III. The calculation method for carbon emissions C3 during polysilicon production is as follows: Figure 4 As shown, specifically:

[0106] First, the total amount of each raw material used in the polysilicon production process and its unit production carbon dioxide equivalent emission factor are determined by querying the CLCD database. Then, the carbon dioxide equivalent emission factors of various energy sources are determined. Subsequently, the carbon emission equivalent coefficient of 1 kWh of electricity at the polysilicon production site corresponding to the energy acquisition stage and the carbon emission equivalent coefficient of 1 kWh of electricity at the polysilicon transportation stage are determined. Since CO2 emissions (kg greenhouse gas) = ​​activity level E (TJ) * K carbon dioxide equivalent emission factor (kg greenhouse gas / TJ), the carbon dioxide equivalent emission factor is the carbon emission equivalent coefficient of 1 kWh of electricity at the polysilicon production site corresponding to the energy acquisition stage.

[0107] Further calculations are made for carbon emissions during the raw material production stage, raw material transportation stage, and polysilicon production stage. The production volume of each raw material is determined based on the unit production volume of polysilicon and the total production volume of the polysilicon enterprise. The carbon emissions from the production of each material are then summed to obtain the carbon emissions during the raw material production stage. The carbon emissions during the raw material transportation stage are calculated based on the unit transportation energy consumption and transportation distance of the transportation machinery. The polysilicon production stage is further subdivided into processes such as mixing and reaction in the reduction furnace, and the carbon emissions during the polysilicon production stage are calculated separately based on the unit shift energy consumption, number of machine shifts, and energy unit carbon emission factor for each process. Therefore, the total carbon emissions during the polysilicon production process are the sum of the carbon emissions from the raw material production and transportation stages, as shown in the following formula:

[0108] C3=∑EC m1s ×k 1s +∑ET 1s ×k 2s

[0109] Where s = 0, 1, 2, 3...N, and N represents the number of polycrystalline silicon types; EC m1s k represents the energy consumption during the polysilicon production stage. 1s The carbon emission equivalent factor for 1 kWh of electricity generated during the energy acquisition phase in the polysilicon production process; ET 1s k represents the energy consumption value during the transportation stage of polysilicon acquisition. 2s The carbon emission equivalent coefficient of 1 kWh of electricity used in the transportation of polysilicon;

[0110] This embodiment takes the polycrystalline silicon manufacturing process as an example. Table 3 shows the material and energy consumption data for manufacturing a 1kWp photovoltaic module.

[0111] Table 3. Material and Energy Data in Polysilicon Manufacturing Process

[0112]

[0113] IV. The calculation method for carbon emissions (C4) during silicon wafer production is as follows: Figure 5 As shown, specifically:

[0114] In the silicon wafer production process, the total amount of each raw material used and its unit production carbon dioxide equivalent emission factor are determined; the carbon dioxide equivalent emission factors of various energy sources are determined, and a digital twin model database is established; a carbon factor library for calculating carbon emissions from silicon wafer production is established; a life cycle model tree for silicon wafer production is generated based on the digital twin model database and the carbon factor library; a carbon emission data calculation model is established based on each stage of silicon wafer production; and carbon emissions at each stage of silicon wafer production are calculated based on the carbon emission data calculation model and the life cycle model tree for silicon wafer production, as shown in the following formula:

[0115] C4=(∑ED m2e +∑EF m2 +∑EL m2 +∑EM m2 +∑EN m2 )×k3+T m2 ×k4

[0116] C4 represents the carbon emissions during silicon wafer production, and ED... m2e The energy consumption value when performing a segment manufacturing process for the e-th type of silicon wafer raw material, where e = 1, 2, 3...u, and u represents the type and quantity of silicon wafer raw materials; EF m2 Energy consumption value during the assembly phase; EL m2 Energy consumption during the decomposition and capacity-building process; EM m2 Energy consumption value when executing grouped segments; EN m2 The energy consumption value during the integration phase; k3 is the carbon emission equivalent coefficient of 1 kWh of electricity used in silicon wafer production; T m2 k is the energy consumption value during silicon wafer transportation; k4 is the carbon emission equivalent coefficient of using 1 kWh of electricity during silicon wafer transportation.

[0117] The digital twin model database includes an engineering structure database, an engineering materials database, an energy-saving analysis database, an engineering setup database, and a silicon wafer production information model database. The silicon wafer production lifecycle model tree includes: parsing the silicon wafer production information model database based on production information during the silicon wafer production process and usage information during the silicon wafer production and use process; establishing a carbon emission data calculation model based on the parsed silicon wafer production information model database and the process of splitting the silicon wafer production lifecycle stages to form each stage of each raw material; wherein the production information includes the main products, by-products, parameter values, and transportation information of each raw material production process, and the usage information includes the production, usage process, usage amount, and waste amount of each raw material; the silicon wafer production lifecycle stages include the construction stage, usage stage, and demolition stage of silicon wafer production.

[0118] The lifecycle model tree of silicon wafer production is used to summarize and calculate data. To avoid data omissions or duplicate calculations, for example, the calculation of carbon emissions at each stage of silicon wafer production is divided into multiple energy consumption values, which is a manifestation of the model tree.

[0119] This embodiment takes the silicon wafer manufacturing process as an example. The material and energy consumption data for manufacturing a 1kWp photovoltaic module are shown in Table 4.

[0120] Table 4. Material and Energy Data in Silicon Wafer Manufacturing Process

[0121]

[0122]

[0123] V. The calculation method for carbon emissions C5 during the battery cell production process is as follows: Figure 6 As shown, specifically:

[0124] A analytic hierarchy process (AHP) is used to construct an energy and electricity carbon emission index system to measure the carbon emission level of the solar cell production process. Carbon emission models for the solar cell production process, the power transmission system, and the electricity consumption side are constructed to calculate carbon emission data during solar cell production, power transmission, and power generation. Historical carbon emission data for the solar cell production process is also calculated. The temporal characteristics of historical carbon emission data are obtained using the Empirical Mode Decomposition (EMD) method. An LSTM (Long Short-Term Memory) network is trained using this historical carbon emission data to calculate carbon emission data for power generation and transmission during solar cell production, as shown in the following formula:

[0125] C5 = f θ (P1, P2, P3, P4, X)

[0126] C5 represents the carbon emissions during the cell manufacturing process, f θFor the constructed LSTM model, θ represents the network structure parameters of the LSTM model, including the memory time period (the model selects data from the previous 11 months to calculate the carbon emissions of the current month); X represents historical carbon emission data; P1, P2, P3, and P4 represent the high-frequency, medium-frequency, low-frequency, and extremely low-frequency power characteristic components of power generation and transmission during the cell production process, respectively.

[0127] This embodiment takes the battery cell manufacturing process as an example. The material and energy consumption data for manufacturing a 1kWp photovoltaic module are shown in Table 5.

[0128] Table 5 Data from the cell manufacturing process

[0129]

[0130] VI. The calculation method for the carbon emissions (C6) of photovoltaic modules is as follows:

[0131] The photovoltaic module is divided into three stages: building material and component production, transportation and operation, and component dismantling and reuse. The calculation boundaries for each stage are defined, and the carbon emissions of each stage are added together to obtain the total carbon emissions C6 of the module, as shown in the following formula:

[0132] C6 = EM m4 ×k6+ET m4 ×k7+T m3 ×k8

[0133] Among them, EM m4 k6 represents the energy consumption during the production of building materials and components for photovoltaic modules; k6 is the carbon emission equivalent coefficient of using 1 kWh of electricity during the production of building materials and components for photovoltaic modules; ET m4 K represents the energy consumption during the transportation and operation of photovoltaic modules; k7 represents the carbon emission equivalent coefficient of using 1 kWh of electricity during the transportation and operation of photovoltaic modules; T m3 k is the energy consumption value during the component dismantling and reuse stage of photovoltaic modules; k8 is the carbon emission equivalent coefficient of using 1 kWh of electricity during the component dismantling and reuse stage of photovoltaic modules.

[0134] This embodiment takes the photovoltaic module manufacturing process as an example. The material and energy consumption data for manufacturing a 1kWp photovoltaic module are shown in Table 6.

[0135] Table 6. Material and Energy Data in Photovoltaic Module Manufacturing Process

[0136]

[0137] VII. The calculation method for carbon emissions (C7) of photovoltaic module power generation operation is as follows: Figure 7 As shown, specifically:

[0138] First, obtain the carbon footprint assessment parameters for power generation operation in the target project, as well as the basic data on power generation operation related to carbon emissions. Then, use the Bayesian network method to calculate the data deviation between the basic power generation operation data and the actual carbon emission data. Next, construct a plan review technology distribution based on the basic power generation operation data and the upper and lower limits of the deviation value to fit the data distribution of the basic power generation operation data. Simplify the continuously changing carbon footprint assessment parameters into a triangular distribution to simulate the parameter distribution of the carbon footprint assessment parameters. Finally, combine the data distribution of the basic power generation operation data and the parameter distribution of the carbon footprint assessment parameters, and calculate the carbon footprint distribution of the power generation operation in the target project using Monte Carlo simulation. Based on the carbon footprint distribution of the power generation operation, calculate the carbon emissions of the target project's power generation operation, as shown in the following formula:

[0139]

[0140] Among them, PF u，v The carbon emission coefficient (GWP) for the v-th power generation operation using material u is available from the CLCD database. v The carbon emission loss coefficient for the v-th type of power generation operation can be obtained from the CLCD database.

[0141] This embodiment takes the photovoltaic module power generation operation process as an example. The material and energy consumption data for manufacturing a 1kWp photovoltaic module are shown in Table 7.

[0142] Table 7. Material and Energy Data During Power Generation Operation

[0143]

[0144] 8. The calculation method for carbon emissions C8 from the retirement and scrapping of photovoltaic modules is as follows:

[0145] Historical carbon emission calculation event data for photovoltaic modules is obtained; based on this data, retirement and scrapping data are calculated, and the model parameters of a pre-built grid BP neural network model are updated until the variance E between the carbon emissions obtained from the current parameters and the expected carbon emissions is less than a set value, thus obtaining the determined model parameters and the trained grid BP neural network model. The model parameters include the weight values ​​of neurons and the normalized carbon emission influence parameters for each group. The electricity consumption for retirement and scrapping at the current time point and the reference electricity consumption are obtained and input into the trained grid BP neural network model. Based on the output of the grid BP neural network model, the carbon emissions from the retirement and scrapping of photovoltaic modules are determined, as shown in the following formula:

[0146]

[0147] C8 represents the carbon emissions from decommissioning and scrapping, while AD... zCalculate the decommissioning and scrapping number for the z-th type of historical carbon emissions, EF z Calculate the carbon emission coefficient for decommissioning and scrapping for the z-th type of historical carbon emission.

[0148] The power grid BP neural network model includes:

[0149] A carbon emission neuron computational mathematical model building unit is used to construct a carbon emission neuron computational mathematical model. The parameters involved in training the neuron computational mathematical model include: the weight values ​​W from neuron p to neuron q. pq The input information x received from neuron p at time t p (t), where f(x) is the preset neuron transfer function;

[0150] The input layer acquisition unit is used to obtain the normalized carbon emission impact parameters of each group as the input layer vector x = (x1, x2, ... x) of the BP neural network based on the BP neural network structure. p , ...x P ), where P is the number of neurons in the input layer;

[0151] Hidden layer acquisition unit, used to calculate the mathematical model of carbon emission neurons, and the input layer vector x = (x1, x2, ... x) corresponding to each group of normalized carbon emission influence parameters contained in the training samples. p , ...x P To obtain the hidden layer vector y = (y1, y2, ..., y3) of the BP neural network. q , ..., y Q )×W pq , where y q W is the q-th neuron in the Q neurons of the hidden layer. pq The vector represents the weights from the p-th neuron in the input layer to the q-th neuron in the hidden layer.

[0152] The output layer acquisition unit is used to calculate the carbon emission budget value 'a' of the output layer of the BP neural network corresponding to each group of normalized carbon emission influence parameters, based on the hidden layer vector and the preset neuron transfer function f(x), where W qk represents the weight values ​​from the q-th neuron in the hidden layer to the k-th neuron in the output layer, where k takes the value of 1;

[0153] The output error calculation unit is used to obtain the actual carbon emission value b corresponding to the normalized carbon emission influence parameter for each group based on the normalized carbon emission data contained in the training samples, and to calculate the corresponding output error for each group of normalized carbon emission influence parameters.

[0154] The parameter error acquisition and judgment unit takes the sum of the squares of the output errors err corresponding to the normalized carbon emission impact parameters of each group as the parameter error, and judges whether the parameter error is within the preset error allowable range; if yes, it outputs the carbon emission data; if no, it adjusts the weight values ​​of neurons p to q and recalculates.

[0155] This embodiment takes the retirement and scrapping process of photovoltaic modules as an example. The material and energy consumption data for manufacturing a 1kWp photovoltaic module are shown in Table 8.

[0156] Table 8 shows the data during the decommissioning and scrapping process of the volt-watt modules.

[0157]

[0158] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the claims of the present invention.

Claims

1. A method for calculating the carbon emissions of a photovoltaic module throughout its entire life cycle, characterized in that: Acquire energy data at each stage of the entire lifecycle of photovoltaic modules; Calculate the carbon emissions per unit of silica mining process Carbon emissions during industrial silicon production Carbon emissions during polysilicon production Carbon emissions during silicon wafer production Carbon emissions during the battery cell production process Carbon emissions from photovoltaic modules The carbon emissions from photovoltaic module power generation operation, considering the energy loss during use. Carbon emissions from the retirement and scrapping of photovoltaic modules ; Calculating total life cycle carbon emissions for a unit product photovoltaic module ​​ The method for calculating the carbon emissions per unit of silica mining process is as follows: Step S1. Collect all electrical and oil information data during the silica mining process to form a basic database; Step S2. Construct a carbon emission calculation model under active / reactive power in the silica mining process, and use the electricity and oil information data collected in step S1 to calculate the carbon emission under active / reactive power in the silica mining process, update the basic database to form an electricity and oil information database, and divide the training set and test set. The carbon emission calculation model for the active / reactive power of the silica mining process is shown in the following formula: ； in, This indicates the carbon emissions during the active / reactive power generation of the silica mining process. This represents the consumption coefficient of the i-th material; This represents the annual consumption of the i-th material. This represents the carbon emission coefficient when using the i-th material; Step S3. Construct a BP neural network. Input the oil information data from the sample data into the input layer of the BP neural network model. Then calculate the error between the output result of the output layer of the BP neural network model and the unit carbon emission of the sample data. Using the corresponding training set and test set in S2, train and test to obtain the carbon emission factor model under active / reactive power in the silica mining process. Based on the model, give the corresponding carbon emission factor under unit active / reactive power in the silica mining process when using oil of different quality. Step S4. Construct an online calculation model for the carbon emissions from power generation during the silica mining process, as shown in the following formula: = ； in, This indicates the carbon emissions from power generation during the silica mining process. This represents the total electrical energy consumed during the production phase of the silica mining process. This represents the carbon emission coefficient of electricity generated during the production stage of silica mining. Step S5. Collect power generation data of the silica mining process online, and distinguish between active and reactive power; combine the corresponding carbon emission factors obtained in S3, and calculate the corresponding carbon emissions of the silica mining process online based on the online calculation model of carbon emissions of power generation during silica mining. The carbon emission calculation model for the silica mining process is as follows: ； wherein, carbon emissions for the silica mining process.

2. The method for calculating the carbon emissions of a photovoltaic module throughout its entire life cycle according to claim 1, characterized in that: The energy data for each stage of the photovoltaic module's entire lifecycle is collected by a monitoring unit, specifically: Utilizing satellite remote sensing intelligent identification technology, the system identifies mining conditions within silica mining areas, including mine size, type, and transportation conditions; acquires transportation energy consumption and distance for each batch of silica over different time periods; obtains time-segmented energy data for mining enterprises via IoT transmission; acquires silica production data; uses the factory areas of industrial silicon, polysilicon, silicon wafers, solar cells, and modules as boundaries, and acquires time-segmented energy data and industrial silicon production data for these enterprises via IoT transmission; and captures unit product energy consumption and carbon emission data for scrap steel and aluminum recycling in the market, acquiring time-segmented energy data for mining enterprises via IoT transmission.

3. The method for calculating the carbon emissions of a photovoltaic module throughout its entire life cycle according to claim 2, characterized in that: The method for calculating carbon emissions during the industrial silicon production process is as follows: Obtain the total operating load segment of industrial silicon production over a certain period of time; based on the total operating load segment of industrial silicon production and the preset basic load point, determine all characteristic load points of industrial silicon production within the total operating load segment; Based on the actual carbon content of the coal fed into the furnace for industrial silicon production and the actual characteristic energy consumption at each characteristic load point, the actual carbon emissions at each characteristic load point are determined. Based on the actual carbon emissions at all characteristic load points, the actual total carbon emissions of industrial silicon production over a certain operating period are determined, as shown in the following formula: ； wherein, is the actual carbon emission for all feature load points as shown in the following equation: ； wherein, is the actual feature energy consumption at the feature load point, as shown in the following equation: ； in, The actual carbon emission coefficient at the characteristic point; Coal consumption for power supply at characteristic load points; The coefficient representing the influence of heat consumption on energy consumption; The coefficient representing the impact of combustion efficiency on energy consumption; The coefficient representing the impact of coal utilization rate on energy consumption; This is the coefficient representing the impact of plant power consumption rate on energy consumption.

4. A method of calculating the full lifecycle carbon emissions of a photovoltaic module according to claim 3, wherein: The method for calculating carbon emissions during the polysilicon production process is as follows: First, determine the total amount of each raw material used in the polysilicon production process and its carbon dioxide equivalent emission factor per unit of production. Then, determine the carbon dioxide equivalent emission factors for each type of energy source. Next, determine the carbon emission equivalent coefficient of 1 kWh of electricity in the polysilicon production stage and the carbon emission equivalent coefficient of 1 kWh of electricity in the polysilicon transportation stage. Further calculate the carbon emissions in the polysilicon production stage and the polysilicon transportation stage. Therefore, the total carbon emissions in the polysilicon production process are the sum of the carbon emissions from the polysilicon production and polysilicon transportation stages, as shown in the following formula: ； Where s = 0, 1, 2, 3...N, and N represents the number of polycrystalline silicon types; This represents the energy consumption during the polysilicon production stage. The carbon emission equivalent coefficient of 1 kWh of electricity in the polysilicon production process corresponding to the energy acquisition stage; This represents the energy consumption during the transportation phase of polysilicon acquisition. The carbon emission equivalent coefficient for 1 kWh of electricity in the polysilicon transportation process.

5. The method for calculating the carbon emissions of a photovoltaic module throughout its entire life cycle according to claim 4, characterized in that: The method for calculating carbon emissions during the silicon wafer production process is as follows: In the silicon wafer production process, the total amount of each raw material used and its unit production carbon dioxide equivalent emission factor are determined; the carbon dioxide equivalent emission factors of various energy sources are determined, and a digital twin model database is established; a carbon factor library for calculating carbon emissions from silicon wafer production is established; a life cycle model tree for silicon wafer production is generated based on the digital twin model database and the carbon factor library; a carbon emission data calculation model is established based on each stage of silicon wafer production; and carbon emissions at each stage of silicon wafer production are calculated based on the carbon emission data calculation model and the life cycle model tree for silicon wafer production, as shown in the following formula: ； in, This refers to carbon emissions during the silicon wafer manufacturing process. The energy consumption value when performing a manufacturing process for the e-th type of silicon wafer raw material, e = 1, 2, 3...u, where u represents the type and quantity of silicon wafer raw materials; This represents the energy consumption value during the assembly phase. This represents the energy consumption value during the fractionation and decomposition process. This represents the energy consumption value when executing grouped segments; Energy consumption value during the execution of the integration phase; The carbon emission equivalent coefficient for using 1 kWh of electricity in silicon wafer production; This represents the energy consumption during the transportation of silicon wafers. The carbon emission equivalent of using 1 kWh of electricity in the silicon wafer transportation process.

6. The method for calculating the carbon emissions of a photovoltaic module throughout its entire life cycle according to claim 5, characterized in that: The method for calculating carbon emissions during the battery cell production process is as follows: A analytic hierarchy process (AHP) is used to construct an energy and electricity carbon emission index system to measure the carbon emission level of the solar cell production process. Carbon emission models for the solar cell production process, the power transmission system, and the electricity consumption side are constructed to calculate carbon emission data during solar cell production, power transmission, and power generation. Historical carbon emission data for the solar cell production process is also calculated. The temporal characteristics of historical carbon emission data are obtained using the Empirical Mode Decomposition (EMD) method. An LSTM (Long Short-Term Memory) network is trained using this historical carbon emission data to calculate carbon emission data for power generation and transmission during solar cell production, as shown in the following formula: ； in, This refers to the carbon emissions during the production of solar cells. For the constructed LSTM model, θ represents the network structure parameters of the LSTM model, including the time period to be remembered; X represents historical carbon emission data. These are the high-frequency, medium-frequency, low-frequency, and ultra-low-frequency electrical characteristic components of power generation and transmission during the battery cell production process.

7. A method of calculating the full lifecycle carbon emissions of a photovoltaic module according to claim 6, wherein: The carbon emissions of the photovoltaic module are calculated as follows: The photovoltaic module is divided into three stages: building material and component production, transportation and operation, and component dismantling and reuse. The calculation boundaries for each stage are defined, and the carbon emissions of each stage are added together to obtain the total carbon emissions of the module. As shown in the formula below: ； in, This refers to the energy consumption during the production of building materials and components for photovoltaic modules. The carbon emission equivalent coefficient for using 1 kWh of electricity in the building materials and component production stage of photovoltaic modules; This represents the energy consumption during the transportation and operation of photovoltaic modules. The carbon emission equivalent factor for using 1 kWh of electricity during the transportation and operation of photovoltaic modules; This refers to the energy consumption during the component disassembly and reuse phase of photovoltaic modules. The carbon emission equivalent factor for using 1 kWh of electricity during the component dismantling and reuse phase of photovoltaic modules.

8. The method for calculating the carbon emissions of a photovoltaic module throughout its entire life cycle according to claim 7, characterized in that: The method for calculating the carbon emissions from the power generation operation is as follows: First, obtain the carbon footprint assessment parameters for power generation operation in the target project, as well as the basic data on power generation operation related to carbon emissions. Then, use the Bayesian network method to calculate the data deviation between the basic power generation operation data and the actual carbon emission data. Next, construct a plan review technology distribution based on the basic power generation operation data and the upper and lower limits of the deviation value to fit the data distribution of the basic power generation operation data. Simplify the continuously changing carbon footprint assessment parameters into a triangular distribution to simulate the parameter distribution of the carbon footprint assessment parameters. Finally, combine the data distribution of the basic power generation operation data and the parameter distribution of the carbon footprint assessment parameters, and calculate the carbon footprint distribution of the power generation operation in the target project using Monte Carlo simulation. Based on the carbon footprint distribution of the power generation operation, calculate the carbon emissions of the target project's power generation operation, as shown in the following formula: ； in, This represents the carbon emission coefficient of the v-th power generation operation using the u-th material; This represents the carbon emission loss coefficient for the v-th type of power generation operation.

9. A method of calculating the full lifecycle carbon emissions of a photovoltaic module according to claim 8, wherein: The method for calculating the carbon emissions from the decommissioning and scrapping of the photovoltaic modules is as follows: The process involves: acquiring historical carbon emission event data for photovoltaic (PV) modules; calculating retirement and scrapping data based on historical carbon emission time data; updating the model parameters of the constructed power grid BP neural network model until the variance E between the carbon emissions obtained from the current parameters and the expected carbon emissions is less than a set value; obtaining the determined model parameters and the trained power grid BP neural network model; the model parameters include the weight values ​​of neurons and the normalized carbon emission influence parameters for each group; acquiring the electricity consumption for retirement and scrapping at the current time point and the reference electricity consumption, inputting them into the trained power grid BP neural network model; and determining the carbon emissions of PV module retirement and scrapping based on the output of the power grid BP neural network model, as shown in the following formula: ； wherein, is the carbon emissions for decommissioning, is the number of decommissioning for the zth historical carbon emission, is the carbon emission factor for decommissioning for the zth historical carbon emission; The power grid BP neural network model includes: The carbon emission neuron computational mathematical model building unit is used to construct a carbon emission neuron computational mathematical model. The parameters involved in training the neuron computational mathematical model include the weight values ​​from neuron p to neuron q. The input information received from neuron p at time t Preset neuron transfer function ; The input layer acquisition unit is used to obtain the normalized carbon emission impact parameters of each group as the input layer vector of the BP neural network based on the BP neural network structure. ,in, This represents the number of neurons in the input layer. Hidden layer acquisition unit, used to calculate the mathematical model of carbon emission neurons and the input layer vector corresponding to each group of normalized carbon emission influence parameters contained in the training samples. Obtain the hidden layer vectors of the BP neural network. ,in, It is the q-th neuron in the Q neurons of the hidden layer. The vector represents the weights from the p-th neuron in the input layer to the q-th neuron in the hidden layer. Output layer acquisition unit, used to obtain information based on hidden layer vectors and preset neuron transfer functions. Calculate the carbon emission budget value output by the output layer of the BP neural network corresponding to the normalized carbon emission influence parameters for each group. ,in, represents the weight values ​​from the q-th neuron in the hidden layer to the k-th neuron in the output layer, where k takes the value of 1; The output error calculation unit is used to obtain the actual carbon emission value b corresponding to the normalized carbon emission influence parameter for each group based on the normalized carbon emission data contained in the training samples, and to calculate the corresponding output error for each group of normalized carbon emission influence parameters. ; The parameter error acquisition and judgment unit outputs the error corresponding to the normalized carbon emission influence parameters for each group. The sum of squares is used as the parameter error, and it is determined whether the parameter error is within the preset error allowable range; if yes, the carbon emission data is output; if no, the weight values ​​of neurons p to q are adjusted and recalculated.

Images