Artificial intelligence-based carbon footprint prediction method for preparing electrolytic copper from waste copper

Abstract

The invention discloses a carbon footprint prediction method for preparing electrolytic copper from waste copper based on artificial intelligence. The method comprises the steps that S1, parameters and weather parameters of all stages of preparing the electrolytic copper from the waste copper are obtained; obtaining the carbon footprint of each stage of preparing electrolytic copper from the waste copper; the data acquisition module is used for acquiring data and pre-processing the acquired data; s2, constructing a TCN-Transformer model, and inputting the data obtained after preprocessing in the S1 into the model for training, thereby training to obtain a neural network model containing a corresponding relationship between parameters and carbon footprints; s3, taking partial parameters of each stage in the process of preparing electrolytic copper from waste copper planned in the future as input characteristics, and inputting the input characteristics into the model trained in the step S2 for prediction so as to obtain carbon footprint data of each stage in the future. And finally, performing addition calculation on the carbon footprint data of each stage to obtain the total carbon footprint in the whole life cycle of preparing the electrolytic copper from the waste copper. According to the method, the future short-term carbon emission level is accurately predicted, and the effectiveness of an emission reduction strategy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon footprint prediction, and more specifically, to an artificial intelligence-based method for predicting the carbon footprint of electrolytic copper produced from scrap copper. Background Art

[0002] AI-based carbon footprint assessment and prediction methods for electrolytic copper produced from scrap copper have garnered widespread attention amid global pressure to reduce carbon emissions and the drive for a circular economy. Due to its significant resource-saving and low-carbon advantages, electrolytic copper produced from scrap copper holds significant significance in promoting green manufacturing and achieving sustainable development. However, existing carbon footprint assessment methods typically focus on a single dimension, relying primarily on life cycle assessment (LCA) or carbon emission factor methods. These methods focus on direct carbon emissions from the production process, ignoring the complexity and carbon emission contributions of other links in the supply chain, such as scrap copper sourcing, transportation, and storage. Furthermore, traditional assessment methods often focus on energy consumption and fail to fully consider the multidimensional impacts of other key factors, such as water use, waste disposal, and other indirect carbon emission sources. More importantly, existing assessments are mostly based on static data, making it difficult to capture dynamic changes in the production process and, therefore, unable to accurately reflect the actual carbon reduction effects of process optimization or operational adjustments. AI-based carbon footprint assessment and prediction methods can combine machine learning and big data analysis technologies to construct a dynamic and intelligent carbon emissions analysis framework across multiple dimensions, encompassing the scrap copper supply chain, production process, and product lifecycle. By introducing sensor networks and IoT devices, data such as the source of scrap copper, transportation routes, and process operation status can be collected in real time, and dynamic predictions can be made using artificial intelligence machine learning algorithms, thereby achieving high-precision data collection and real-time analysis. This can provide data support for parameter adjustments in the production process, thereby effectively improving the accuracy and implementation effect of emission reduction strategies. Summary of the Invention

[0003] The present invention aims to overcome at least one defect (shortcoming) of the above-mentioned prior art and provide an artificial intelligence-based carbon footprint prediction method for electrolytic copper produced from scrap copper, which is used to solve the problem that existing assessment methods are difficult to fully and accurately reflect the actual carbon emission level of electrolytic copper produced from scrap copper, thereby affecting the effectiveness of emission reduction strategies.

[0004] The technical solution adopted by the present invention is an artificial intelligence-based method for predicting the carbon footprint of electrolytic copper produced from scrap copper, the method comprising the following steps:

[0005] S1: Obtaining scrap copper recycling process parameters, transportation route parameters, scrap copper storage parameters, scrap copper stretching parameters, scrap copper processing parameters, scrap copper treatment parameters, and weather parameters; obtaining the carbon footprint of the scrap copper recycling stage, transportation stage, storage stage, stretching stage, processing stage, and waste treatment stage; and pre-processing the above-obtained data;

[0006] S2: Build a TCN-Transformer model and input the data obtained after S1 preprocessing into the model for training, so as to train a neural network model containing the corresponding relationship between parameters and carbon footprint;

[0007] S3: Some parameters of each stage in the future planned process of preparing electrolytic copper from scrap copper are used as input features and input into the model trained in step S2 for prediction, so as to obtain the carbon footprint data of each stage in the future. Finally, the carbon footprint data of each stage are added up and calculated to obtain the total carbon footprint of the entire life cycle of electrolytic copper prepared from scrap copper.

[0008] In this application, artificial intelligence is combined to find the relationship between the parameters and carbon footprint in the process of preparing electrolytic copper from scrap copper, so that some parameters of each stage in the process of preparing electrolytic copper from scrap copper in the future and the trained model can be used to predict the carbon footprint data of each stage in the future. This can not only achieve high-precision data collection and real-time analysis, but also provide data support for parameter adjustment in the production process, realize accurate prediction of future short-term carbon emission levels, accurately reflect the actual carbon emission level of electrolytic copper from scrap copper, and improve the effectiveness of emission reduction strategies.

[0009] Preferably, in step S2, constructing the TCN-Transformer model includes:

[0010] S21: Construct a multi-layer temporal convolutional network (TCN) to extract features from the data input into the model, and perform feature concatenation and fusion on the extracted feature data;

[0011] S22: The Transformer network receives the fused features and performs position encoding on the input sequence, and then performs sequence prediction to obtain the output data matrix of the carbon footprint, thereby training a carbon footprint calculation model that includes the corresponding relationship between the parameters and the corresponding carbon footprint for each parameter of the scrap copper in step S1. The total carbon footprint outputted is:

[0012] C 总 =C 回收 +C 运输 +C 存储 +C 拉伸 +C 加工 +C 处理

[0013] Among them, C 总 Represents the total carbon footprint over the entire life cycle.

[0014] In this application, a prediction model is constructed by combining TCN and Transformer networks to achieve the prediction of the carbon footprint of electrolytic copper prepared from scrap copper. It can not only utilize the efficient local feature extraction of TCN, but also inherit the global reasoning ability of Transformer, which significantly improves the performance of the model and makes the prediction results of the trained carbon footprint calculation model more accurate.

[0015] Preferably, in step S21, the formula for feature extraction is:

[0016]

[0017] Where X is the input data matrix, To extract the feature matrix.

[0018] Preferably, in step S22, performing position encoding on the input sequence includes: constructing a position code and adding it to the input embedding to obtain an input sequence with time position information; wherein the position code is:

[0019]

[0020] PE is the positional encoding, t is the time step index, i is the position in the embedding dimension, and d model is the dimension of the input vector.

[0021] By injecting time sequence information into the model through position encoding, the model's perception of time sequence is enhanced. The sequence prediction function is directly related to the generation of the output matrix of the carbon footprint, thereby training the correspondence between parameters and carbon footprint and improving the accuracy of model prediction.

[0022] Preferably, the recycling process parameters include the mass of collected scrap copper, the energy consumption per unit mass of collected scrap copper and its first carbon emission factor, the mass of sorted scrap copper, the energy consumption per unit mass of sorted scrap copper and its second carbon emission factor;

[0023] In step S2, a corresponding relationship between the energy consumption per unit mass of scrap copper collected and its first carbon emission factor and the carbon footprint of the collection is established, as well as a corresponding relationship between the energy consumption per unit mass of scrap copper sorted and its second carbon emission factor and the carbon footprint of the sorting is established. The obtained carbon footprint calculation model includes a collection carbon emission calculation function and a sorting carbon emission calculation function.

[0024] Therefore, the collected carbon footprint of the scrap copper can be calculated based on the collected scrap copper mass and the collected carbon emission calculation function. The collected carbon emission calculation function is as follows:

[0025]

[0026] Among them, C1 represents the collection of carbon footprint, Dm M represents the energy consumption required to collect unit mass of type m scrap copper. m Indicates the mass of the mth type of scrap copper, E Dm represents the first carbon emission factor of type m of scrap copper, and h represents the number of scrap copper types;

[0027] The sorting carbon footprint of the scrap copper is calculated based on the mass of the sorted scrap copper and the sorting carbon emission calculation function. The sorting carbon emission calculation function is as follows:

[0028]

[0029] Among them, C2 represents the sorting carbon footprint, S m M represents the energy consumption required to sort the mth type of scrap copper per unit mass. m Indicates the mass of the mth type of scrap copper, E Sm represents the second carbon emission factor of type m scrap copper;

[0030] The collection carbon footprint and the sorting carbon footprint are summed up to obtain the carbon footprint of the scrap copper in the recycling stage.

[0031] Preferably, the transport route parameters include the transport route distance, the mass of the transported scrap copper and its third carbon emission factor; a corresponding relationship between the third carbon emission factor and the carbon footprint of the transport stage is established in step S2, and the obtained carbon footprint calculation model includes a transport carbon emission calculation function;

[0032] Thus, the carbon footprint of the transportation stage of the scrap copper is calculated based on the transportation route distance, the mass of the transported scrap copper, and the transportation carbon emission calculation function. The transportation carbon emission calculation function is as follows:

[0033]

[0034] Among them, C 运输 represents the carbon footprint of the transportation stage, D j represents the transport route distance of the jth transport route, F j represents the mass of scrap copper transported on the jth transport route, EY j represents the third carbon emission factor of the jth transport route, and g represents the total number of transport routes.

[0035] Preferably, the scrap copper storage parameters include the storage time of the storage mode, the energy consumption of the equipment per unit time, and the fourth carbon emission factor thereof; in step S2, a corresponding relationship between the energy consumption of the equipment per unit time, the fourth carbon emission factor, and the carbon footprint of the storage stage is established, and the obtained carbon footprint calculation model includes a storage carbon emission calculation function;

[0036] Therefore, based on the storage time of the storage method and the storage carbon emission calculation function, the carbon footprint of the storage stage of the scrap copper is calculated. The storage carbon emission calculation function is as follows:

[0037]

[0038] Among them, C 存储 represents the carbon footprint of the storage stage, T t Indicates the storage time of the t-th storage method, P t Indicates the energy consumption per unit time of the t-th storage mode, EC t represents the fourth carbon emission factor of the t-th storage method, and f represents the number of storage methods.

[0039] Preferably, the scrap copper stretching parameters include the mass of the stretched scrap copper, the energy usage per unit mass of the scrap copper stretching, and its fifth carbon emission factor; through step S2, a corresponding relationship between the energy usage per unit mass of the scrap copper stretching, its fifth carbon emission factor, and the carbon footprint of the stretching stage is established, and the obtained carbon footprint calculation model includes a stretching carbon emission calculation function;

[0040] Therefore, based on the mass of the stretched scrap copper and the stretching carbon emission calculation function, the carbon footprint of the stretching stage of the scrap copper is calculated. The stretching carbon emission calculation function is as follows:

[0041]

[0042] Among them, C 拉伸 represents the carbon footprint of the stretching stage, M k represents the mass of the kth type of scrap copper, U k The unit mass tensile energy consumption of the kth type of scrap copper, EL k represents the fifth carbon emission factor of the kth type of scrap copper, and h represents the number of scrap copper types.

[0043] Preferably, the scrap copper processing parameters include the mass of the scrap copper processed, the energy usage per unit mass of the scrap copper processed, and its sixth carbon emission factor; in step S2, a corresponding relationship between the energy usage per unit mass of the scrap copper processed, its sixth carbon emission factor, and the carbon footprint of the processing stage is established, and the obtained carbon footprint calculation model includes a processing carbon emission calculation function;

[0044] Therefore, based on the mass of the processed scrap copper and the processing carbon emission calculation function, the carbon footprint of the processing stage of the scrap copper is calculated. The processing carbon emission calculation function is as follows:

[0045]

[0046] Among them, C 加工 represents the carbon footprint of the processing stage, M krepresents the mass of the kth type of scrap copper, Q k It represents the unit mass processing energy consumption of the kth type of scrap copper, EJ k represents the sixth carbon emission factor of the kth type of scrap copper, and h represents the number of scrap copper types.

[0047] Preferably, the copper scrap processing parameters include waste mass and the seventh carbon emission factor; a corresponding relationship between the seventh carbon emission factor and the carbon footprint of the waste processing stage is established in step S2, and the obtained carbon footprint calculation model includes a processing carbon emission calculation function;

[0048] Thus, the carbon footprint of the waste treatment stage of the scrap copper is calculated based on the waste mass and the treatment carbon emission calculation function. The treatment carbon emission calculation function is as follows:

[0049]

[0050] Among them, C 处理 represents the carbon footprint of the waste treatment stage, W v represents the waste mass of type v waste, E 处理,v represents the seventh carbon emission factor of the vth type of waste, and r represents the number of waste types.

[0051] By predicting the carbon footprint at each of the above stages, we can evaluate the carbon emissions in the process of producing electrolytic copper from scrap copper from multiple perspectives, and accurately reflect the actual carbon emission level of electrolytic copper produced from scrap copper.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] The present invention provides a method for predicting the carbon footprint of electrolytic copper prepared from scrap copper, which obtains scrap copper recovery process parameters, transportation route parameters, scrap copper storage parameters, scrap copper stretching parameters, scrap copper processing parameters, and scrap copper treatment parameters; inputs the recovery process parameters, transportation route parameters, scrap copper storage parameters, scrap copper stretching parameters, scrap copper processing parameters, and scrap copper treatment parameters into a carbon footprint calculation model to obtain the carbon footprint of the scrap copper in the recovery stage, transportation stage, storage stage, stretching stage, processing stage, and waste treatment stage output by the carbon footprint calculation model; and performs evaluation and calculation based on the carbon footprint of the recovery stage, transportation stage, storage stage, stretching stage, processing stage, and waste treatment stage to obtain the total carbon footprint of the scrap copper. Therefore, in the entire evaluation process, the scrap copper recovery process, transportation route, scrap copper storage, scrap copper stretching, scrap copper processing, and scrap copper treatment are combined to evaluate the carbon emissions in the process of producing electrolytic copper from scrap copper from multiple perspectives, thereby accurately reflecting the actual carbon emission level of electrolytic copper prepared from scrap copper, and improving the effectiveness of emission reduction strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a schematic flow chart of the artificial intelligence-based method for predicting the carbon footprint of electrolytic copper from scrap copper provided by the present invention.

[0055] Figure 2 This is a flowchart of the overall process of the artificial intelligence-based method for predicting the carbon footprint of electrolytic copper from scrap copper provided by the present invention.

[0056] Figure 3 This is a schematic diagram of the structure of the TCN-Transformer model provided by the present invention. DETAILED DESCRIPTION

[0057] The accompanying drawings are for illustrative purposes only and are not to be construed as limiting the present invention. To better illustrate the following embodiments, some components in the accompanying drawings may be omitted, enlarged, or reduced in size, and do not represent actual product dimensions. Those skilled in the art will appreciate that some well-known structures and their descriptions may be omitted from the accompanying drawings.

[0058] Example 1

[0059] like Figure 1 and Figure 2 As shown, this embodiment provides an artificial intelligence-based method for predicting the carbon footprint of electrolytic copper from scrap copper, the method comprising the following steps:

[0060] Step S1, obtaining scrap copper recycling process parameters, transportation route parameters, scrap copper storage parameters, scrap copper stretching parameters, scrap copper processing parameters, scrap copper treatment parameters, and weather parameters; obtaining the carbon footprint of the scrap copper in the recycling stage, the transportation stage, the storage stage, the stretching stage, the processing stage, and the waste treatment stage; and preprocessing the above-obtained data;

[0061] Step S2: constructing a TCN-Transformer model and inputting the data obtained after preprocessing in S1 into the model for training, thereby training a neural network model containing the corresponding relationship between parameters and carbon footprints;

[0062] In step S3, some parameters of each stage of the future planned process of preparing electrolytic copper from scrap copper are used as input features and input into the model trained in step S2 for prediction, thereby obtaining the carbon footprint data of each stage in the future. Finally, the carbon footprint data of each stage are added together to obtain the total carbon footprint of the entire life cycle of preparing electrolytic copper from scrap copper.

[0063] In this application, artificial intelligence is combined to find the relationship between the parameters and carbon footprint in the process of preparing electrolytic copper from scrap copper, so that some parameters of each stage in the process of preparing electrolytic copper from scrap copper in the future and the trained model can be used to predict the carbon footprint data of each stage in the future. This can not only achieve high-precision data collection and real-time analysis, but also provide data support for parameter adjustment in the production process, realize accurate prediction of future short-term carbon emission levels, accurately reflect the actual carbon emission level of electrolytic copper from scrap copper, and improve the effectiveness of emission reduction strategies.

[0064] Preferably, in step S1, data preprocessing is performed on the carbon footprint data corresponding to the parameters of the scrap copper processing stage and the carbon footprint data output by the model according to the acquired data to obtain an input-output matrix, including:

[0065] Input data matrix of carbon footprint i:

[0066]

[0067] Similarly, the output data matrix of carbon footprint i is:

[0068]

[0069] in, is the element of the input data set of carbon footprint, n is the input feature vector dimension, is the element of the output dataset of carbon footprint i, and m is the number of input and output vector samples.

[0070] Further preferably, the weather parameters affecting the collection of scrap copper include temperature, humidity and air pressure, so:

[0071]

[0072] Among them, T k The temperature of the kth scenario, MO k represents the humidity of the kth scene, PR k Represents the air pressure of the kth scenario.

[0073] Further preferably, the copper collection methods include manual collection, mechanical collection, etc. Different collection methods correspond to different energy consumption and carbon emissions. Therefore, the recycling process parameters include the mass of scrap copper collected, the energy consumption of collecting unit mass of scrap copper and its first carbon emission factor; scrap copper needs to be sorted and preliminarily processed before entering reproduction. Sorting includes manual sorting, mechanical sorting, magnetic separation and other processes. Scrap copper processing involves steps such as compression, cutting, and impurity removal. Therefore, the recycling process parameters may also include the mass of scrap copper sorted, the energy consumption of sorting unit mass of scrap copper and its second carbon emission factor;

[0074]

[0075] in, represents the carbon footprint of the recycling of type m scrap copper, D m M represents the energy consumption per unit mass of scrap copper collected for the mth type of scrap copper. m Indicates the mass of the mth type of scrap copper, ED m represents the first carbon emission factor of the mth type of scrap copper, S m The energy consumption per unit mass of scrap copper sorting of the mth type is ES m represents the second carbon emission factor of type m scrap copper;

[0076] Preferably, the transportation of scrap copper from the collection point to the processing plant can be by land, sea or air. The energy consumption and carbon emissions of different transportation modes vary greatly. The carbon footprint of the transportation stage includes the emissions generated during the transportation of scrap copper from the collection point to the processing plant. Therefore, to obtain the carbon footprint of the transportation stage: the transportation route parameters include the transportation route distance and its third carbon emission factor, and the mass of the transported scrap copper;

[0077]

[0078] in, represents the carbon footprint of the transportation stage of the j-th transportation route, D j represents the transport route distance of the jth transport route, F j represents the mass of scrap copper transported on the jth transport route, EY j represents the third carbon emission factor of the jth transportation route.

[0079] After recycling, scrap copper may need to be temporarily stored, which includes simple storage or storage in a specific environment (such as a temperature-controlled warehouse). Therefore, to obtain the carbon footprint of the storage stage: scrap copper storage parameters include storage time of the storage method, energy consumption per unit time of the equipment, and the fourth carbon emission factor;

[0080]

[0081] in, represents the carbon footprint of the storage phase of the t-th storage method, T t Indicates the storage time of the t-th storage method, P t Indicates the energy consumption per unit time of the t-th storage mode, EC t Represents the fourth carbon emission factor of the tth storage method.

[0082] Drawing is the process of stretching copper rod from a thicker diameter to the desired final specifications. This process is usually carried out by mechanical equipment, involving high-energy mechanical operations such as electric stretching machines, lubrication systems, and cooling systems. Therefore, to obtain the carbon footprint of the drawing stage: scrap copper drawing parameters include the mass of scrap copper stretched, the energy used per unit mass of scrap copper stretched, and its fifth carbon emission factor;

[0083]

[0084] in, represents the carbon footprint of the k-th type of scrap copper during the stretching stage, M k represents the mass of the kth type of scrap copper, U k The energy consumption per unit mass of the k-th type of scrap copper is EL k represents the fifth carbon emission factor of the kth type of scrap copper.

[0085] Processing is the stage of further processing of copper rods after initial stretching, including surface treatment, finishing, cutting, and bending operations. The processing process usually involves multiple steps and a variety of equipment. The energy consumption of various equipment is the main source of carbon emissions. Therefore, to obtain the carbon footprint of the processing stage: scrap copper processing parameters include the processing quality of scrap copper, the energy used in processing per unit mass of scrap copper, and the sixth carbon emission factor;

[0086]

[0087] in, represents the carbon footprint of the processing stage, M k represents the mass of the kth type of scrap copper, Q k The energy consumption per unit mass of the k-th type of scrap copper is EJ k represents the sixth carbon emission factor of the kth type of scrap copper.

[0088] The carbon footprint of the waste treatment stage mainly includes the treatment of waste generated during the production process. Therefore, to obtain the carbon footprint of the waste treatment stage: the parameters of waste copper treatment include waste mass and seventh carbon emission factor;

[0089]

[0090] in, represents the carbon footprint of the waste treatment stage of type v waste, W v represents the waste mass of type v waste, EC v represents the seventh carbon emission factor of the vth type of waste, and r represents the number of waste types.

[0091] Preferably, in step S2, the processed data is input into the constructed TCN-Transformer model for training and optimization to obtain a carbon footprint calculation model.

[0092] like Figure 3 As shown, the construction of the TCN-Transformer model includes:

[0093] S21: Construct a multi-layer temporal convolutional network (TCN) to extract features from the data input into the model, and perform feature concatenation and fusion on the extracted feature data;

[0094] Preferably, in step S21, the formula for feature extraction is:

[0095]

[0096] Where X is the input data matrix, To extract the feature matrix. In this example, the TCN has 4 layers, and the dilation coefficient of each dilated convolution layer increases exponentially layer by layer. By dilated convolution, the receptive field is expanded, thereby achieving multi-scale local feature extraction of the input features;

[0097] To improve model stability and generalization, TCN also uses Batch Normalization (BN) to standardize data across batches, improving model training stability. Dropout randomly drops neurons to prevent overfitting. ReLU is the activation function, and Conv1D is a one-dimensional convolution.

[0098] S22: Since the Transformer does not have the time sequence processing capability like RNN or LSTM, after the Transformer network receives the fused features, it needs to inject the time sequence information into the model through position encoding. The purpose of position encoding is to let the model know the relative or absolute position of the elements in the sequence. Position encoding is generally generated by sine and cosine functions. The position encoding PE(t) of each time step can be expressed as:

[0099]

[0100] Where t is the time step index, i is the position in the embedding dimension, and d model is the dimension of the input vector. By injecting temporal order information into the model through position encoding, the model’s perception of temporal order is effectively enhanced, thus improving the accuracy of the model’s predictions.

[0101] The positional encoding is then added to the input embedding to obtain the input sequence with temporal position information:

[0102] Xpos =X+PE

[0103] Given an input sequence X pos , first obtain the query matrix Q, key matrix K and value matrix V through linear transformation:

[0104] Q=X pos W Q ,K=X pos W K ,V=X pos W V

[0105] Among them, W Q ,W K ,W V is the weight matrix learned during training.

[0106] The self-attention mechanism of Transfromer can be expressed as:

[0107]

[0108] The multi-head attention mechanism can be expressed as:

[0109] head i =Attention(QW i Q ,KW i K ,VW i V )

[0110] MultiHead(Q,K,V)=Concat(head1,head2,…,head h )W0

[0111] Among them, head i is the output of the i-th attention head, and W0 is the linear transformation matrix.

[0112] The final output of Transformer is:

[0113]

[0114] Among them, LayerNorm is residual connection and normalization, and FFN is feedforward neural network.

[0115] The total carbon footprint output by the model is:

[0116] C 总 =C 回收 +C 运输 +C 存储 +C 拉伸+C 加工 +C 处理

[0117] Among them, C 总 Represents the total carbon footprint over the entire life cycle.

[0118] Through the training of the above model architecture, a carbon footprint calculation model is obtained for each parameter of the scrap copper in step S1, which includes the corresponding relationship between the parameters and the corresponding carbon footprint. This realizes the prediction of the carbon footprint of electrolytic copper prepared from scrap copper throughout its entire life cycle. It can not only utilize the efficient local feature extraction of TCN, but also inherit the global reasoning ability of Transformer, significantly improving the performance of the model and making the prediction results of the trained carbon footprint calculation model more accurate.

[0119] Preferably, in step S3, based on the trained model, some parameters of each stage in the process of preparing electrolytic copper from scrap copper in the future are predicted to obtain the total carbon footprint of the scrap copper in the future.

[0120] Specifically, obtaining the carbon footprint of the recycling stage includes collecting the carbon footprint and sorting the carbon footprint.

[0121] The collection methods of scrap copper include manual collection and mechanical collection. Different collection methods correspond to different energy consumption and carbon emissions. Therefore, the recycling process parameters include the mass of scrap copper collected, the energy consumption per unit mass of scrap copper collected, and its first carbon emission factor. Step S2 establishes a correspondence between the energy consumption per unit mass of scrap copper collected, its first carbon emission factor, and the collection carbon footprint. The resulting carbon footprint calculation model includes a collection carbon emission calculation function. Therefore, the mass of scrap copper collected is input into the carbon footprint calculation model to obtain the carbon footprint of the scrap copper recycling stage output by the carbon footprint calculation model, including:

[0122] The collected carbon footprint of the scrap copper can be calculated based on the collected scrap copper mass and the collected carbon emission calculation function. The collected carbon emission calculation function is as follows:

[0123]

[0124] Where C1 represents the carbon footprint of the collection (kg CO2-eq), D m Indicates the energy consumption required to collect unit mass of type m scrap copper (MJ / kg), M m represents the mass of the mth type of scrap copper (kg), E Dm represents the first carbon emission factor of the mth type of scrap copper ((kgCO2-eq) / MJ), h represents the number of scrap copper types; further, D m *M mIt can be understood as collection energy consumption, which represents the energy used in the scrap copper collection process (such as fuel, electricity, etc.), and is related to the type of energy used and the quality of the scrap copper.

[0125] Scrap copper needs to be sorted and preliminarily processed before entering reproduction. Sorting includes manual sorting, mechanical sorting, magnetic separation and other processes. Scrap copper processing involves steps such as compression, cutting, and impurity removal. Therefore, the recycling process parameters can also include the mass of sorted scrap copper, the energy consumption of sorting unit mass of scrap copper and its second carbon emission factor. Through step S2, the corresponding relationship between the energy consumption of sorting unit mass of scrap copper and its second carbon emission factor and the sorting carbon footprint is established. At this time, the carbon footprint calculation model includes a sorting carbon emission calculation function. Therefore, based on the sorted scrap copper mass and the sorting carbon emission calculation function, the sorting carbon footprint of the scrap copper is calculated. The sorting carbon emission calculation function is as follows:

[0126]

[0127] Among them, C2 represents the sorting carbon footprint, S m Indicates the energy consumption required to sort the mth type of scrap copper per unit mass (MJ / kg), M m represents the mass of the mth type of scrap copper (kg), E Sm represents the second carbon emission factor of the mth type of scrap copper ((kg CO2-eq) / MJ); further, S m *M m It can be understood as sorting energy consumption.

[0128] Furthermore, the carbon footprint of the collection and the carbon footprint of the sorting are summed to obtain the carbon footprint of the recycling stage of the scrap copper. 回收 =C1+C2.

[0129] Preferably, the transportation of scrap copper from the collection point to the processing plant can be by land, sea, or air. Different transportation methods have significantly different energy consumption and carbon emissions. The transportation phase carbon footprint includes emissions generated during the transportation of scrap copper from the collection point to the processing plant. Therefore, to obtain the transportation phase carbon footprint: the transportation route parameters include the transportation route distance, the mass of the transported scrap copper, and its third carbon emission factor; through step S2, a corresponding relationship is established between the third carbon emission factor and the transportation phase carbon footprint. The resulting carbon footprint calculation model includes a transportation carbon emission calculation function.

[0130] Thus, the carbon footprint of the transportation stage of the scrap copper is calculated based on the transportation route distance, the mass of the transported scrap copper, and the transportation carbon emission calculation function. The transportation carbon emission calculation function is as follows:

[0131]

[0132] Among them, C 运输 represents the carbon footprint of the transportation stage, D j represents the transport route distance (km) of the jth transport route, F j represents the mass of scrap copper transported along the jth transport route (kg), EYj j represents the third carbon emission factor of the j-th transport route ((kgCO2-eq) / (kg*km)), and g represents the total number of transport routes.

[0133] Preferably, the scrap copper may need to be temporarily stored after recycling, which includes simple storage or storage in a specific environment (such as a temperature-controlled warehouse). Therefore, to obtain the carbon footprint of the storage stage: the scrap copper storage parameters include the storage time of the storage method, the energy consumption of the equipment per unit time, and its fourth carbon emission factor; through step S2, a corresponding relationship between the energy consumption of the equipment per unit time and its fourth carbon emission factor and the carbon footprint of the storage stage is established, and the obtained carbon footprint calculation model includes a storage carbon emission calculation function;

[0134] Therefore, based on the storage time of the storage method and the storage carbon emission calculation function, the carbon footprint of the storage stage of the scrap copper is calculated. The storage carbon emission calculation function is as follows:

[0135]

[0136] Among them, C 存储 represents the carbon footprint of the storage stage, T t Indicates the storage time (s) of the t-th storage method, P t represents the energy consumption per unit time of the t-th storage mode (MW), EC t represents the fourth carbon emission factor ((kg CO2-eq) / MJ) of the t-th storage method, and f represents the number of storage methods.

[0137] Preferably, stretching is the process of stretching a copper rod from a relatively large diameter to the desired final specifications. This process is typically performed mechanically and involves high-energy mechanical operations, such as electric stretching machines, lubrication systems, and cooling systems. Therefore, to obtain the carbon footprint of the stretching stage: the scrap copper stretching parameters include the mass of the stretched scrap copper, the energy usage per unit mass of the scrap copper stretching, and its fifth carbon emission factor; through step S2, a corresponding relationship is established between the energy usage per unit mass of the scrap copper stretching, its fifth carbon emission factor, and the carbon footprint of the stretching stage. The resulting carbon footprint calculation model includes a stretching carbon emission calculation function.

[0138] Therefore, based on the mass of the stretched scrap copper and the stretching carbon emission calculation function, the carbon footprint of the stretching stage of the scrap copper is calculated. The stretching carbon emission calculation function is as follows:

[0139]

[0140] Among them, C 拉伸 represents the carbon footprint of the stretching stage, M k represents the mass of the kth type of scrap copper (kg), U k represents the unit mass tensile energy consumption of the kth type of scrap copper (kWh / kg), EL k represents the fifth carbon emission factor of the kth type of scrap copper ((kg CO2-eq) / kWh), and h represents the number of scrap copper types.

[0141] Preferably, processing is a stage of further processing of the copper rod after preliminary stretching, including surface treatment, finishing, cutting, and bending operations. The processing process usually involves multiple steps and multiple types of equipment. The energy consumption of various equipment is the main source of carbon emissions. Therefore, to obtain the carbon footprint of the processing stage: the scrap copper processing parameters include the mass of the processed scrap copper, the energy usage per unit mass of scrap copper processing, and its sixth carbon emission factor; through step S2, a corresponding relationship is established between the energy usage per unit mass of scrap copper processing and its sixth carbon emission factor and the carbon footprint of the processing stage. The obtained carbon footprint calculation model includes a processing carbon emission calculation function;

[0142] Therefore, based on the mass of the processed scrap copper and the processing carbon emission calculation function, the carbon footprint of the processing stage of the scrap copper is calculated. The processing carbon emission calculation function is as follows:

[0143]

[0144] Among them, C 加工 represents the carbon footprint of the processing stage, M k represents the mass of the kth type of scrap copper (kg), Q k represents the unit mass processing energy consumption of the kth type of scrap copper (kWh / kg), EJ k represents the sixth carbon emission factor ((kg CO2-eq) / kWh) of the kth type of scrap copper, and h represents the number of scrap copper types.

[0145] Preferably, the carbon footprint of the waste treatment stage mainly includes the treatment of waste generated during the production process. Therefore, to obtain the carbon footprint of the waste treatment stage: the scrap copper treatment parameters include waste mass and the seventh carbon emission factor; through step S2, a corresponding relationship between the seventh carbon emission factor and the carbon footprint of the waste treatment stage is established, and the obtained carbon footprint calculation model includes a treatment carbon emission calculation function;

[0146] Thus, the carbon footprint of the waste treatment stage of the scrap copper is calculated based on the waste mass and the treatment carbon emission calculation function. The treatment carbon emission calculation function is as follows:

[0147]

[0148] Among them, C 处理 represents the carbon footprint of the waste treatment stage, W v represents the waste mass of type v waste, EC v represents the seventh carbon emission factor ((kg CO2-eq) / kg) of the vth type of waste, and r represents the number of waste types.

[0149] By predicting the carbon footprint at each of the above stages, we can evaluate the carbon emissions in the process of producing electrolytic copper from scrap copper from multiple perspectives, and accurately reflect the actual carbon emission level of electrolytic copper produced from scrap copper.

[0150] Further preferably, the carbon footprint data of each stage are summed up to obtain the total carbon footprint of the entire life cycle of preparing electrolytic copper from scrap copper, including:

[0151] The total carbon footprint of copper scrap is calculated by summing up the carbon footprint of the recycling stage, the carbon footprint of the transportation stage, the carbon footprint of the storage stage, the carbon footprint of the stretching stage, the carbon footprint of the processing stage, and the carbon footprint of the waste treatment stage. Therefore, the calculation formula of the total carbon footprint is as follows:

[0152] C 总 =C 回收 +C 运输 +C 存储 +C 拉伸 +C 加工 +C 处理

[0153] Among them, C 总 Represents the total carbon footprint over the entire life cycle.

[0154] Preferably, if Figure 2 As shown, Figure 2 The overall flow chart of the method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence provided by the present invention. The embodiment of the present invention can also be dynamically adjusted and continuously optimized. The total carbon footprint calculated above is a static carbon footprint. In the actual production process, various dynamic factors (such as equipment wear, operator efficiency, and fluctuations in raw material quality) will affect carbon emissions. In order to better capture these factors, the model introduces a dynamic adjustment coefficient. Through real-time monitoring and data updates, such as process improvements and energy structure adjustments, the accounting model is regularly optimized to ensure the accuracy and timeliness of carbon footprint assessments. The dynamically adjusted carbon footprint after considering the dynamic adjustment coefficient can be calculated by the following formula:

[0155]

[0156] Among them, C 总 Represents the static carbon footprint, which is the total carbon footprint calculated according to the above formula.

[0157] ΔOperating efficiency, ΔRaw material quality, and ΔEquipment wear are relative changes observed in the actual production process and are used to dynamically adjust carbon emission values.

[0158] In one specific example, Company A processes 1,000 tons of copper scrap annually and recycles it into copper rod. The scrap is primarily collected from nearby suppliers, mechanically collected without sorting, and transported using diesel trucks. During the production process, electric stretching machines are used for processing, and waste is disposed of in landfills.

[0159] 1. Data Collection

[0160] Recycling Data: Mechanical collection was used, with a collection energy consumption of 0.9 MJ / kg and a carbon emission factor of 0.07 kg CO2-eq / MJ. Transportation Data: 60 km, transport method: diesel truck, transportation carbon emission factor: 0.1 kg CO2-eq / ton-km. Storage Data: Standard warehouse storage, storage equipment energy consumption of 25 kW, storage for 2 days, storage carbon emission factor: 0.05 kg CO2-eq / MJ. Stretching Processing Data: Electric stretching machine was used, with energy consumption of 90 kWh / ton during the stretching phase and a carbon emission factor of 0.45 kg CO2-eq / kWh. Energy consumption during the processing phase was 80 kWh / ton and a carbon emission factor of 0.5 kg CO2-eq / kWh. Waste Disposal Data: 60 tons of waste were generated during the production process, all of which was landfilled, with a carbon emission factor of 0.5 kg CO2-eq / kg.

[0161] 2. Data processing

[0162]

[0163] 3. Carbon footprint prediction and calculation

[0164] Assume that in the prediction algorithm, the learning rate is 0.001, the number of attention mechanism heads is 8, the number of encoder-decoder stacks is 2, and the expansion coefficient is [1, 2, 4, 8].

[0165] The model's prediction performance was continuously tested on a dataset covering the entire year of 2019. In this example, the average time for a single model training run was 10 seconds, and the standard root mean square error of the predicted data was 4.92%, making this model of certain significance for practical engineering applications.

[0166] This embodiment obtains scrap copper recycling process parameters, transportation route parameters, scrap copper storage parameters, scrap copper stretching parameters, scrap copper processing parameters, and scrap copper treatment parameters; obtains the carbon footprint of the scrap copper during the recycling phase, transportation phase, storage phase, stretching phase, processing phase, and waste treatment phase. The recycling process parameters, transportation route parameters, scrap copper storage parameters, scrap copper stretching parameters, scrap copper processing parameters, and scrap copper treatment parameters are used as input features, and the recycling phase carbon footprint, transportation phase carbon footprint, storage phase, stretching phase, processing phase carbon footprint, and waste treatment phase carbon footprint are used as output features. A neural network model is trained using artificial intelligence methods to find the corresponding relationship between the parameters and the carbon footprint. The trained model is used to predict the future carbon footprint using future planned production parameters.

[0167] Carbon footprint calculation based on the trained model:

[0168] Scrap copper recycling stage:

[0169] C 回收 =0.9MJ / kg*1000 tons*1000kg / ton*0.07kgCO2-eq / MJ=63000kgCO2-eq.

[0170] Transportation stage:

[0171] C 运输 =60km**1000 tons*0.1kg CO2-eq / ton·km=6000kg CO2-eq.

[0172] Storage stage:

[0173] C 存储 =0.5MW*2days*86400seconds / day*0.05kgCO2-eq / MJ=4320kgCO2-eq.

[0174] Stretching phase

[0175] C 拉伸 =1000 tons*90kWh / ton*0.45kg CO2-eq / kWh=40500kg CO2-eq.

[0176] Processing stage

[0177] C 加工 =1000 tons*80kWh / ton*0.5kg CO2-eq / kWh=40000kg CO2-eq.

[0178] Waste disposal stage: Assuming that the waste accounts for 6% of the copper rod production, that is, 60 tons, all of which are landfilled,

[0179] C 处理 =60 tons*1000kg / ton*0.5kg CO2-eq / kg=30000kg CO2-eq.

[0180] 3. Total carbon footprint calculation:

[0181] Ctotal=63000+6000+4320+40500+40000+30000=183820kg CO2-eq.

[0182] 4. Dynamic carbon footprint accounting:

[0183] Assume ΔOperation Efficiency = -0.05 (indicating a 5% increase in operation efficiency); ΔRaw Material Quality = 0.06 (indicating a 6% decrease in raw material quality); ΔEquipment Wear = 0.02 (indicating a 2% increase in equipment wear).

[0184]

[0185] Therefore, based on the above example, when processing 1,000 tons of scrap copper, Company A's total carbon footprint is 185,658.2 kg CO₂-eq. Carbon emissions at each stage can be further reduced by optimizing recycling and transportation methods, improving energy efficiency, and improving waste disposal methods, thereby further optimizing and reducing the carbon footprint. Despite improved operational efficiency, declining raw material quality and equipment wear still resulted in an increase in the carbon footprint. Ultimately, the dynamically adjusted carbon footprint increased by 1% compared to the static carbon footprint, reaching 185,658.2 kg CO₂-eq. Adjustment and optimization strategies in actual production can be further adjusted based on specific changes.

[0186] Thus, the embodiment of the present invention establishes a parameter and carbon footprint model for preparing electrolytic copper from scrap copper, and simultaneously accurately predicts future short-term carbon emission levels, thereby improving the effectiveness of emission reduction strategies.

[0187] Obviously, the above embodiments of the present invention are merely examples for the purpose of clearly illustrating the technical solutions of the present invention, and are not intended to limit the specific implementation methods of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the claims of the present invention shall be included within the scope of protection of the claims of the present invention.

Claims

1. A method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence, characterized in that: The method comprises the following steps: S1: Obtaining scrap copper recycling process parameters, transportation route parameters, scrap copper storage parameters, scrap copper stretching parameters, scrap copper processing parameters, scrap copper treatment parameters, and weather parameters; obtaining the carbon footprint of the scrap copper recycling stage, transportation stage, storage stage, stretching stage, processing stage, and waste treatment stage; and pre-processing the above-obtained data; S2: Build a TCN-Transformer model and input the data obtained after S1 preprocessing into the model for training, so as to train a neural network model containing the corresponding relationship between parameters and carbon footprint; S3: Some parameters of each stage in the future planned process of preparing electrolytic copper from scrap copper are used as input features and input into the model trained in step S2 for prediction, so as to obtain the carbon footprint data of each stage in the future. Finally, the carbon footprint data of each stage are added up and calculated to obtain the total carbon footprint of the entire life cycle of electrolytic copper prepared from scrap copper.

2. The method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence according to claim 1, characterized in that: In step S2, constructing the TCN-Transformer model includes: S21: Construct a multi-layer temporal convolutional network (TCN) to extract features from the data input to the model, and perform feature concatenation and fusion on the extracted feature data; S22: The Transformer network receives the fused features and performs position encoding on the input sequence, and then performs sequence prediction to obtain the output data matrix of the carbon footprint, thereby training a carbon footprint calculation model that includes the corresponding relationship between the parameters and the corresponding carbon footprint for each parameter of the scrap copper in step S1. The total carbon footprint outputted is: C 总 =C 回收 +C 运输 +C 存储 +C 拉伸 +C 加工 +C 处理 Among them, C 总 Represents the total carbon footprint over the entire life cycle.

3. The method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence according to claim 2, characterized in that: In step S21, the formula for feature extraction is: Where X is the input data matrix, To extract the feature matrix.

4. The method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence according to claim 2, characterized in that: In step S22, performing position coding of the input sequence includes: constructing a position code and adding it to the input embedding to obtain an input sequence with time position information; wherein the position code is: PE is the positional encoding, t is the time step index, i is the position in the embedding dimension, and d model is the dimension of the input vector.

5. The method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence according to claim 2, characterized in that: The recycling process parameters include the mass of collected scrap copper, the energy consumption per unit mass of collected scrap copper and its first carbon emission factor, the mass of sorted scrap copper, the energy consumption per unit mass of sorted scrap copper and its second carbon emission factor; In step S2, a corresponding relationship between the energy consumption per unit mass of scrap copper collected and its first carbon emission factor and the carbon footprint of the collection is established, as well as a corresponding relationship between the energy consumption per unit mass of scrap copper sorted and its second carbon emission factor and the carbon footprint of the sorting is established. The obtained carbon footprint calculation model includes a collection carbon emission calculation function and a sorting carbon emission calculation function. Therefore, the collected carbon footprint of the scrap copper can be calculated based on the collected scrap copper mass and the collected carbon emission calculation function. The collected carbon emission calculation function is as follows: Among them, C1 represents the collection of carbon footprint, D m M represents the energy consumption required to collect unit mass of type m scrap copper. m Indicates the mass of the mth type of scrap copper, E Dm represents the first carbon emission factor of type m of scrap copper, and h represents the number of scrap copper types; The sorting carbon footprint of the scrap copper is calculated based on the mass of the sorted scrap copper and the sorting carbon emission calculation function. The sorting carbon emission calculation function is as follows: Among them, C2 represents the sorting carbon footprint, S m M represents the energy consumption required to sort the mth type of scrap copper per unit mass. m Indicates the mass of the mth type of scrap copper, E Sm represents the second carbon emission factor of type m scrap copper; The collection carbon footprint and the sorting carbon footprint are summed up to obtain the carbon footprint of the scrap copper in the recycling stage.

6. The method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence according to claim 2, characterized in that: The transport route parameters include the transport route distance, the mass of the transported scrap copper, and its third carbon emission factor; a corresponding relationship between the third carbon emission factor and the carbon footprint of the transport stage is established in step S2, and the obtained carbon footprint calculation model includes a transport carbon emission calculation function; Thus, the carbon footprint of the transportation stage of the scrap copper is calculated based on the transportation route distance, the mass of the transported scrap copper, and the transportation carbon emission calculation function. The transportation carbon emission calculation function is as follows: Among them, C 运输 represents the carbon footprint of the transportation stage, D j represents the transport route distance of the jth transport route, F j represents the mass of scrap copper transported on the jth transport route, EY j represents the third carbon emission factor of the jth transport route, and g represents the total number of transport routes.

7. The method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence according to claim 2, characterized in that: The scrap copper storage parameters include the storage time of the storage method, the energy consumption of the equipment per unit time, and the fourth carbon emission factor; in step S2, a corresponding relationship between the energy consumption of the equipment per unit time and the fourth carbon emission factor and the carbon footprint of the storage stage is established, and the obtained carbon footprint calculation model includes a storage carbon emission calculation function; Therefore, based on the storage time of the storage method and the storage carbon emission calculation function, the carbon footprint of the storage stage of the scrap copper is calculated. The storage carbon emission calculation function is as follows: Among them, C 存储 represents the carbon footprint of the storage stage, T t Indicates the storage time of the t-th storage method, P t Indicates the energy consumption per unit time of the t-th storage mode, EC t represents the fourth carbon emission factor of the t-th storage method, and f represents the number of storage methods.

8. The method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence according to claim 2, characterized in that: The scrap copper stretching parameters include the mass of the stretched scrap copper, the energy usage per unit mass of the stretched scrap copper, and its fifth carbon emission factor; through step S2, a corresponding relationship is established between the energy usage per unit mass of the stretched scrap copper, its fifth carbon emission factor, and the carbon footprint of the stretching stage, and the obtained carbon footprint calculation model includes a stretching carbon emission calculation function; Therefore, based on the mass of the stretched scrap copper and the stretching carbon emission calculation function, the carbon footprint of the stretching stage of the scrap copper is calculated. The stretching carbon emission calculation function is as follows: Among them, C 拉伸 represents the carbon footprint of the stretching stage, M k represents the mass of the kth type of scrap copper, U k The unit mass tensile energy consumption of the kth type of scrap copper, EL k represents the fifth carbon emission factor of the kth type of scrap copper, and h represents the number of scrap copper types.

9. The method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence according to claim 2, characterized in that: The scrap copper processing parameters include the mass of the scrap copper processed, the energy usage per unit mass of the scrap copper processed, and its sixth carbon emission factor; in step S2, a corresponding relationship is established between the energy usage per unit mass of the scrap copper processed, its sixth carbon emission factor, and the carbon footprint of the processing stage, and the obtained carbon footprint calculation model includes a processing carbon emission calculation function; Therefore, based on the mass of the processed scrap copper and the processing carbon emission calculation function, the carbon footprint of the processing stage of the scrap copper is calculated. The processing carbon emission calculation function is as follows: Among them, C 加工 represents the carbon footprint of the processing stage, M k represents the mass of the kth type of scrap copper, Q k It represents the unit mass processing energy consumption of the kth type of scrap copper, EJ k represents the sixth carbon emission factor of the kth type of scrap copper, and h represents the number of scrap copper types.

10. The method for predicting the carbon footprint of electrolytic copper from scrap copper based on artificial intelligence according to claim 2, characterized in that: The copper scrap processing parameters include waste mass and the seventh carbon emission factor; a corresponding relationship between the seventh carbon emission factor and the carbon footprint of the waste processing stage is established through step S2, and the obtained carbon footprint calculation model includes a processing carbon emission calculation function; Thus, the carbon footprint of the waste treatment stage of the scrap copper is calculated based on the waste mass and the treatment carbon emission calculation function. The treatment carbon emission calculation function is as follows: Among them, C 处理 represents the carbon footprint of the waste treatment stage, W v represents the waste mass of type v waste, EC v represents the seventh carbon emission factor of the vth type of waste, and r represents the number of waste types.

Images