Method and system for flexibly formulating product carbon footprint model based on LCA

Abstract

The invention discloses a method and system for flexibly formulating a product carbon footprint model based on LCA, and relates to the technical field of environmental sciences, and the method comprises the steps: determining a carbon footprint evaluation demand, and constructing an evaluation framework according to the demand; dividing the LCA data into data sets according to life cycle stages; collecting carbon emission data in a life cycle stage in real time; constructing an LSTM model for predicting the carbon emission trends of the product in different life cycle stages; and carrying out carbon footprint accounting, hotspot analysis, sensitivity analysis and uncertainty evaluation based on the LSTM model. According to the method for flexibly formulating the product carbon footprint model based on the LCA provided by the invention, flexible carbon footprint evaluation is realized through a modularized data platform and an adjustable system boundary; by combining the Internet of Things and the AI algorithm, the data acquisition accuracy and analysis efficiency are improved, manual intervention is reduced, carbon emission trend prediction and multi-dimensional analysis are realized, and the evaluation precision and decision support capability are improved.

Description

Technical Field

[0001] The present invention relates to the field of environmental science and technology, and in particular to a method and system for flexibly formulating a product carbon footprint model based on LCA. Background Art

[0002] Product carbon footprint assessment is an important means to assess the greenhouse gas emissions directly or indirectly generated by a product or service throughout its life cycle. Under existing technologies, the life cycle assessment (LCA) method is mainly used to assess product carbon footprint. LCA is a full-system analysis method that comprehensively considers energy consumption and environmental impact from the collection of raw materials, production and processing, transportation, use to waste disposal of products. During the assessment process, data from each stage, including energy consumption, material consumption, waste disposal and transportation, is collected, and the carbon emissions of each stage are calculated using relevant emission factors. Finally, the carbon emissions of each stage are summarized to obtain the total carbon footprint of the product. This method helps companies to fully understand the real impact of their products on the environment and provide data support for companies to formulate targeted emission reduction strategies. However, product carbon footprint assessment methods under existing technologies often use fixed frameworks and preset data sets, which are difficult to adapt to changes in different products and assessment needs, and often rely on manual operations in data processing and analysis, which is inefficient and prone to errors. Summary of the invention

[0003] In view of the above-mentioned problems, the present invention is proposed.

[0004] Therefore, the technical problem solved by the present invention is that the product carbon footprint assessment method under the existing technology often adopts a fixed framework and a preset data set, which is difficult to adapt to the changes in different products and assessment needs, and often relies on manual operations in data processing and analysis, which is inefficient and prone to errors.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for flexibly formulating a product carbon footprint model based on LCA, comprising:

[0006] Determine the carbon footprint assessment needs and build an assessment framework based on the needs;

[0007] Divide LCA data into data sets according to life cycle stages;

[0008] Real-time collection of carbon emission data during the life cycle;

[0009] Build an LSTM model to predict the carbon emission trend of products at different life cycle stages;

[0010] Carbon footprint accounting, hotspot analysis, sensitivity analysis and uncertainty assessment are performed based on the LSTM model.

[0011] As a preferred solution of the method for flexibly formulating a product carbon footprint model based on LCA described in the present invention, wherein: determining the carbon footprint assessment requirements includes identifying and determining specific products for which carbon footprint assessment is required; the selected products may be physical goods or services;

[0012] Determine the carbon footprint assessment needs of the product based on its attributes;

[0013] Determine the life cycle stages that need to be included and set emission reduction targets based on the assessment objectives and product characteristics;

[0014] The life cycle stages include raw material acquisition, production, transportation, use and waste treatment;

[0015] The construction of the assessment framework includes integrating the assessment scope and generating an assessment requirements document based on the identified product types, life cycle stages and set emission reduction targets;

[0016] Define the system boundaries based on the assessment requirements document, including the processes and activities that need to be included in the assessment at each life cycle stage;

[0017] Set up adjustable system boundary interfaces; determine unified functional units; select and define key environmental impact indicators;

[0018] The constructed assessment framework is documented, including descriptions of system boundaries, functional units, and environmental impact indicators.

[0019] As a preferred solution of the method for flexibly formulating a product carbon footprint model based on LCA described in the present invention, wherein: the data set includes a static data set and a dynamic data set;

[0020] The static data set includes data that does not change over time during the evaluation process;

[0021] The dynamic data set includes data that changes over time.

[0022] As a preferred solution of the method for flexibly formulating a product carbon footprint model based on LCA described in the present invention, wherein: the real-time collection of carbon emission data during the life cycle stage includes determining the location of deploying the Internet of Things device according to the life cycle stage;

[0023] Select IoT devices and plan the number of IoT devices to be deployed;

[0024] Deploy IoT devices according to the plan and configure the devices;

[0025] Build the overall architecture of the big data platform, including components of the data storage layer, processing layer, and analysis layer;

[0026] Determine the data storage solution and select the corresponding database and file system to store real-time data and static data.

[0027] As a preferred solution of the method for flexibly formulating a product carbon footprint model based on LCA described in the present invention, wherein: the construction of an LSTM model for predicting the trend of carbon emissions of products at different life cycle stages includes determining the prediction target of carbon emissions based on different life cycle stages of the product, analyzing the time series characteristics of historical data, including periodicity, seasonality and trend characteristics, and selecting a long short-term memory neural network;

[0028] Build a Python environment and install data processing libraries and deep learning libraries;

[0029] Use Pandas to read the original data set related to carbon emissions, clean the data, remove missing values ​​and outliers, and perform data standardization operations;

[0030] Perform feature extraction on the cleaned data, set the time step and window size, convert the historical data into a format suitable for time series analysis, and capture time dependencies;

[0031] Divide the extracted time series data into training set, validation set and test set;

[0032] Use TensorFlow or PyTorch to build an LSTM model, including an input layer, multiple hidden layers, and an output layer; configure hyperparameters, including learning rate, optimizer, and loss function, and initialize model weights;

[0033] Convert the training set data into the three-dimensional array format required by the LSTM model, including the number of samples, time steps, and features, and batch process the data;

[0034] Use the training set data to adjust the model weights through the back-propagation algorithm and stochastic gradient descent optimizer, and adopt a dynamic learning rate adjustment strategy;

[0035] The performance of the trained LSTM model is evaluated through the validation set, and the root mean square error and mean absolute error are calculated;

[0036] Encapsulate the verified LSTM model into an independent module and define the data input and output interfaces.

[0037] As a preferred solution of the method for flexibly formulating a product carbon footprint model based on LCA described in the present invention, the carbon footprint calculation and hot spot analysis include:

[0038] Obtaining predicted carbon emission data for each life cycle stage from the LSTM model, including dynamic trends in raw material acquisition, production, transportation, use, and waste treatment;

[0039] According to the prediction data of the LSTM model and the demand documents input by the user, parse the user needs and product life cycle information, and automatically analyze the current system boundaries; according to the analysis results, adjust the system boundaries;

[0040] The adjusting of the system boundary includes parsing and evaluating the requirement document to identify new life cycle activities; adjusting the system boundary to include new activities or exclude irrelevant activities according to data changes; and verifying the adjusted system boundary;

[0041] Based on the LSTM model prediction data, the carbon emissions of each life cycle stage are calculated within the adjusted system boundary, the emission data of each stage are classified, and the main carbon emission sources are identified.

[0042] As a preferred solution of the flexible formulation method of the product carbon footprint model based on LCA described in the present invention, the sensitivity analysis and uncertainty assessment include screening relevant static and dynamic data sets according to the adjusted system boundaries; screening includes parsing user demand documents and automatically screening data sets that meet the assessment requirements; setting data sampling frequency and time period; making adjustments based on LSTM prediction data to evaluate the impact of various parameters on carbon footprint;

[0043] Randomly sample the prediction results of the LSTM model to quantify the fluctuation range of the prediction data.

[0044] Another object of the present invention is to provide a flexible formulation system for product carbon footprint models based on LCA, which can solve the problems of lack of flexibility, incomplete data collection, low data processing efficiency, and inability to dynamically adjust system boundaries and data sets in existing product carbon footprint assessment methods by constructing a flexible formulation system for product carbon footprint models based on LCA.

[0045] To solve the above technical problems, the present invention provides the following technical solutions: a flexible formulation system of a product carbon footprint model based on LCA, comprising: a framework construction module, a stage division module, a data acquisition module, a model construction module and a prediction and analysis module; the framework construction module is used to determine the carbon footprint assessment needs and build an assessment framework according to the needs; the stage division module is used to divide the LCA data into data sets according to the life cycle stages; the data acquisition module is used to collect carbon emission data in the life cycle stages in real time; the model construction module is used to construct an LSTM model for predicting the carbon emission trend of products in different life cycle stages; the prediction and analysis module is used to perform carbon footprint accounting, hot spot analysis, sensitivity analysis and uncertainty assessment based on the LSTM model.

[0046] A computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the method for flexibly formulating a product carbon footprint model based on LCA are implemented.

[0047] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the method for flexibly formulating a product carbon footprint model based on LCA as described above.

[0048] Beneficial effects of the present invention: The flexible formulation method of the product carbon footprint model based on LCA provided by the present invention overcomes the problems of fixed evaluation framework, low data processing efficiency, insufficient accuracy of analysis results, etc. in the prior art. The present invention allows users to flexibly customize the LCA method according to different product characteristics and evaluation requirements by introducing adjustable system boundaries and modular data platforms, and can select suitable life cycle stages, data sets and analysis methods, thereby improving the flexibility and pertinence of the evaluation. By integrating the Internet of Things and big data technologies, the real-time collection of carbon emission data at each life cycle stage is realized, and the artificial intelligence algorithm module is used to clean, preprocess, classify and trend predict the data, which reduces manual intervention and significantly improves the efficiency and accuracy of data processing. Based on the LSTM model, the trend prediction of carbon emissions in the product life cycle is carried out, which provides data support for enterprises to formulate accurate emission reduction strategies. In addition, through sensitivity analysis and uncertainty assessment, the system boundaries and data set selection can be dynamically adjusted to ensure the reliability and operability of the evaluation results. In summary, the present invention is significantly superior to the prior art in terms of data collection coverage, processing efficiency and analysis accuracy, and can provide enterprises with more comprehensive and accurate carbon footprint assessment and emission reduction decision support, thereby helping to achieve green and low-carbon transformation. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0050] Figure 1 An overall flow chart of a method for flexibly formulating a product carbon footprint model based on LCA provided by an embodiment of the present invention.

[0051] Figure 2 An overall structural diagram of a flexible formulation system of a product carbon footprint model based on LCA provided by an embodiment of the present invention.

[0052] Figure 3A flow chart of a method for formulating a product carbon footprint model, which is a flexible method for formulating a product carbon footprint model based on LCA, is provided in accordance with an embodiment of the present invention.

[0053] Figure 4 A flowchart of an integrated artificial intelligence algorithm module for a flexible formulation method of a product carbon footprint model based on LCA provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0055] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0056] Example 1

[0057] Reference Figure 1 , which is an embodiment of the present invention, provides a method for flexibly formulating a product carbon footprint model based on LCA, comprising:

[0058] Determine the carbon footprint assessment needs and build an assessment framework based on the needs;

[0059] Divide LCA data into data sets according to life cycle stages;

[0060] Real-time collection of carbon emission data during the life cycle;

[0061] Build an LSTM model to predict the carbon emission trend of products at different life cycle stages;

[0062] Carbon footprint accounting, hotspot analysis, sensitivity analysis and uncertainty assessment are performed based on the LSTM model.

[0063] Determining the need for carbon footprint assessment includes identifying and determining specific products for which carbon footprint assessment is required; the selected products may be physical goods or services;

[0064] Determine the carbon footprint assessment needs of the product based on its attributes;

[0065] Determine the life cycle stages that need to be included and set emission reduction targets based on the assessment objectives and product characteristics;

[0066] The life cycle stages include raw material acquisition, production, transportation, use and waste treatment;

[0067] The construction of the assessment framework includes integrating the assessment scope and generating an assessment requirements document based on the identified product types, life cycle stages and set emission reduction targets;

[0068] Define the system boundaries based on the assessment requirements document, including the processes and activities that need to be included in the assessment at each life cycle stage;

[0069] Set up adjustable system boundary interfaces; determine unified functional units; select and define key environmental impact indicators;

[0070] The constructed assessment framework is documented, including descriptions of system boundaries, functional units, and environmental impact indicators.

[0071] The data set includes a static data set and a dynamic data set;

[0072] The static data set includes data that does not change over time during the evaluation process;

[0073] The dynamic data set includes data that changes over time.

[0074] The real-time collection of carbon emission data during the life cycle stage includes determining the location of deploying the IoT device according to the life cycle stage;

[0075] Select IoT devices and plan the number of IoT devices to be deployed;

[0076] Deploy IoT devices according to the plan and configure the devices;

[0077] Build the overall architecture of the big data platform, including components of the data storage layer, processing layer, and analysis layer;

[0078] Determine the data storage solution and select the corresponding database and file system to store real-time data and static data.

[0079] The construction of the LSTM model for predicting the carbon emission trend of products at different life cycle stages includes determining the prediction target of carbon emissions based on the different life cycle stages of the products, analyzing the time series characteristics of historical data, including periodicity, seasonality and trend characteristics, and selecting a long short-term memory neural network;

[0080] Build a Python environment and install data processing libraries and deep learning libraries;

[0081] Use Pandas to read the original data set related to carbon emissions, clean the data, remove missing values ​​and outliers, and perform data standardization operations;

[0082] Perform feature extraction on the cleaned data, set the time step and window size, convert the historical data into a format suitable for time series analysis, and capture time dependencies;

[0083] Divide the extracted time series data into training set, validation set and test set;

[0084] Use TensorFlow or PyTorch to build an LSTM model, including an input layer, multiple hidden layers, and an output layer; configure hyperparameters, including learning rate, optimizer, and loss function, and initialize model weights;

[0085] Convert the training set data into the three-dimensional array format required by the LSTM model, including the number of samples, time steps, and features, and batch process the data;

[0086] Use the training set data to adjust the model weights through the back-propagation algorithm and stochastic gradient descent optimizer, and adopt a dynamic learning rate adjustment strategy;

[0087] The performance of the trained LSTM model is evaluated through the validation set, and the root mean square error and mean absolute error are calculated;

[0088] Encapsulate the verified LSTM model into an independent module and define the data input and output interfaces.

[0089] The carbon footprint calculation and hot spot analysis include

[0090] Obtaining predicted carbon emission data for each life cycle stage from the LSTM model, including dynamic trends in raw material acquisition, production, transportation, use, and waste treatment;

[0091] According to the prediction data of the LSTM model and the demand documents input by the user, parse the user needs and product life cycle information, and automatically analyze the current system boundaries; according to the analysis results, adjust the system boundaries;

[0092] The adjusting of the system boundary includes parsing and evaluating the requirement document to identify new life cycle activities; adjusting the system boundary to include new activities or exclude irrelevant activities according to data changes; and verifying the adjusted system boundary;

[0093] Based on the LSTM model prediction data, the carbon emissions of each life cycle stage are calculated within the adjusted system boundary, the emission data of each stage are classified, and the main carbon emission sources are identified.

[0094] The sensitivity analysis and uncertainty assessment include screening relevant static and dynamic data sets according to the adjusted system boundaries; screening includes parsing user requirement documents and automatically screening data sets that meet the assessment requirements; setting data sampling frequency and time period; making adjustments based on LSTM prediction data and assessing the impact of each parameter on the carbon footprint;

[0095] Randomly sample the prediction results of the LSTM model to quantify the fluctuation range of the prediction data.

[0096] Example 2

[0097] Reference Figure 2 , which is an embodiment of the present invention, provides a flexible formulation system of a product carbon footprint model based on LCA, including:

[0098] Framework construction module 100, stage division module 200, data collection module 300, model construction module 400 and prediction analysis module 500;

[0099] The framework building module 100 is used to determine the carbon footprint assessment requirements and build an assessment framework according to the requirements;

[0100] The stage division module 200 is used to divide the LCA data into data sets according to the life cycle stages;

[0101] The data collection module 300 is used to collect carbon emission data in the life cycle stage in real time;

[0102] The model building module 400 is used to build an LSTM model for predicting the carbon emission trend of products at different life cycle stages;

[0103] The prediction and analysis module 500 is used to perform carbon footprint calculation, hot spot analysis, sensitivity analysis and uncertainty assessment based on the LSTM model.

[0104] Example 3

[0105] An embodiment of the present invention is different from the first two embodiments in that:

[0106] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0107] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0108] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0109] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0110] Example 4

[0111] Reference Figure 3-Figure 4 , which is an embodiment of the present invention, provides a flexible formulation method of a product carbon footprint model based on LCA.

[0112] Step 1: Determine the carbon footprint assessment needs, including specific products, life cycle stages and emission reduction targets; and build an assessment framework based on the needs, including adjustable system boundaries, functional units and key indicators; the assessment framework provides the basis and guidance for subsequent steps;

[0113] The specific steps to determine the need for carbon footprint assessment include:

[0114] aDetermination of specific products:

[0115] Identify and determine specific products that require carbon footprint assessment through market research, internal data analysis or customer feedback;

[0116] Products can be physical goods, such as electronic products, food, and clothing, or services, such as logistics and transportation, and travel services;

[0117] The selection of products should be based on factors such as their impact on the environment, the company's emission reduction goals or market attention;

[0118] b. Determination of life cycle stages:

[0119] Determine the life cycle stages that need to be included in the assessment based on the product characteristics and assessment objectives;

[0120] The life cycle stages usually include raw material acquisition, production, transportation, use, and waste disposal;

[0121] Each stage can be further subdivided according to the level of detail of the assessment, for example, the production stage can be subdivided into sub-stages such as processing, assembly, and packaging;

[0122] c Setting of emission reduction targets:

[0123] Set clear emission reduction targets in line with the company's sustainable development strategy, industry benchmarks or policy requirements;

[0124] Emission reduction targets can be either absolute reductions in carbon emissions or reduction ratios relative to a base year;

[0125] Goals should be specific, measurable, and trackable, with a reasonable time frame.

[0126] d. Build an evaluation framework:

[0127] Construct an assessment framework based on the specific products, life cycle stages and emission reduction targets identified;

[0128] The assessment framework should include adjustable system boundaries to clarify which processes and activities are included in the assessment scope;

[0129] Establishing functional units as a unified benchmark for measuring the environmental performance of products and services;

[0130] Determine key environmental impact indicators, such as carbon emissions and energy consumption;

[0131] The specific steps to build the evaluation framework include:

[0132] A1: Integrate the scope and objectives of the assessment, including specific products, selected life cycle stages, reduction targets and any constraints;

[0133] A2: According to the assessment requirements, determine which processes and activities should be included in the assessment scope and build a preliminary system boundary;

[0134] A3: Reserve adjustment interfaces for system boundaries so that they can be flexibly adjusted according to subsequent analysis requirements;

[0135] A4: Select a common benchmark unit for measuring the environmental performance of products and services;

[0136] A5: Determine key environmental impact indicators based on assessment objectives and product characteristics, mainly including carbon emissions and energy consumption;

[0137] A6: Document the constructed evaluation framework, including adjustable system boundaries, functional units, and key indicators.

[0138] Step 2: Divide the LCA data into multiple modules, including raw material acquisition, production, transportation, use, and waste treatment. Each module contains corresponding static and dynamic data sets, which provide a basis for subsequent data selection.

[0139] LCA data refers to all relevant data used for life cycle assessment, including but not limited to the various parameters and indicators mentioned in the above framework. These data are orderly divided into different life cycle stage modules such as raw material acquisition, production, transportation, use and waste treatment;

[0140] Step 3: Use the Internet of Things and big data technologies to build a real-time data integration module to support the real-time collection of carbon emission data at each life cycle stage. The data provided by the real-time data integration module is the basis for subsequent data processing and analysis.

[0141] Step 4: Integrate the artificial intelligence algorithm module for data processing, analysis and prediction, obtain a well-trained L-step TM model that can predict the carbon emission trend of products at different life cycle stages, and encapsulate it into a reusable intelligent algorithm module. The artificial intelligence algorithm module provides intelligent support for subsequent steps;

[0142] Step 5: Adjust the system boundaries according to the assessment requirements and select the life cycle stages that need to be included in the assessment; Select the life cycle stages that need to be included in the assessment, based on the assessment framework constructed in step 1;

[0143] Step 6: Select the corresponding static and dynamic data sets from the modular data platform, combine and configure them, and customize the LCA method according to product characteristics and evaluation objectives, including process LCA, input-output LCA and hybrid methods; the customized LCA method and the selected data set are based on the modular data in step 2;

[0144] Step 7: Use the artificial intelligence algorithm module integrated in step 4 to intelligently process the collected data, including cleaning, preprocessing, classification and trend prediction;

[0145] Step 8: Flexibly select analysis methods and tools according to assessment needs to conduct carbon footprint accounting, hot spot analysis, sensitivity analysis and uncertainty assessment; the analysis is based on the data after intelligent processing in step 7, and is based on the system boundary adjusted in step 5 and the customized LCA method in step 6;

[0146] Step 9: Use visualization tools to display and output the analysis results in the form of charts and animations; the visualization display and output are based on the comprehensive results of all the previous steps;

[0147] Among them, according to the evaluation requirements, adjust the system boundaries and select the life cycle stages that need to be included in the evaluation. This step is carried out with the support of the intelligent algorithm module obtained in step 4 to ensure the flexibility and pertinence of the evaluation;

[0148] Among them, the corresponding static and dynamic data sets are selected from the modular data platform, combined and configured, and LCA methods are customized according to product characteristics and evaluation objectives, including process LCA, input-output LCA and hybrid methods. This step uses the intelligent algorithm module in step 4 to optimize data processing and analysis;

[0149] Among them, the subsequent steps 7 to 9 are based on the data processing, analysis and prediction capabilities provided by the intelligent algorithm module obtained in step 4, as well as the customized evaluation framework and data set configuration of steps 5 and 6, to carry out carbon footprint accounting, hotspot analysis, sensitivity analysis, uncertainty assessment and visualization.

[0150] As a priority, determine the carbon footprint assessment needs, including specific products, life cycle stages and emission reduction targets; and build an assessment framework based on the needs, including adjustable system boundaries, functional units and key indicators; the specific steps are as follows:

[0151] Step 101: Select a common benchmark unit for measuring the environmental performance of products and services based on the specific products and services to be evaluated;

[0152] Step 102: Integrate the assessment scope and objectives, including specific products, life cycle stages, emission reduction targets and constraints, construct an adjustable system boundary, and identify the processes and activities included in the assessment scope;

[0153] Step 103: Determine key environmental impact indicators, including carbon emissions and energy consumption, based on the assessment objectives and product characteristics;

[0154] Step 104: Document the constructed assessment framework, including system boundaries, functional units, and key indicators.

[0155] As a preference, the LCA data is divided into multiple modules, including raw material acquisition, production, transportation, use, and waste treatment, each module containing corresponding static and dynamic data sets; the specific steps are as follows:

[0156] Step 201: Define the classification standard of LCA data and classify it according to the life cycle stage, including raw material acquisition, production, transportation, use, and waste treatment;

[0157] Step 202: Collect LCA related data from various data sources including internal enterprise databases, external databases, and literature;

[0158] Step 203: traverse the collected data and assign the data to the corresponding life cycle stage according to the classification criteria;

[0159] Step 204: Create independent folders and database tables for data in each life cycle stage;

[0160] Step 205: In the data set of each life cycle stage, a static data set and a dynamic data set are further distinguished. The static data set includes average energy consumption and emission factors, and the dynamic data set includes real-time energy consumption data and transportation distance.

[0161] Step 206: The static data set is stored in the database, and the dynamic data set is connected to the IoT device in real time.

[0162] As a preferred method, the Internet of Things and big data technologies are used to build a real-time data integration module to support the real-time collection of carbon emission data at each life cycle stage; the specific steps are as follows:

[0163] Step 301: Determine the specific location where the IoT device needs to be deployed based on the criticality of the life cycle stage;

[0164] Step 302: Select IoT devices and plan their deployment quantity and location. The IoT devices include temperature sensors, pressure sensors, RFID tags, and GP step trackers.

[0165] Step 303: deploy IoT devices according to the plan and configure the devices so that the devices can work normally and collect data in real time;

[0166] Step 304: construct the overall architecture of the big data platform, including components of the data storage layer, the processing layer, and the analysis layer;

[0167] Step 305: Determine a data storage solution and select a corresponding database and file system to store real-time data and static data;

[0168] Among them, static data sets include data that do not change or change little over time during the evaluation process, such as average energy consumption and emission factors; dynamic data sets include data that change over time, such as real-time energy consumption data and transportation distance;

[0169] Real-time data refers to data collected in real time through IoT devices that reflects the actual operating status of the product at each stage of its life cycle, including but not limited to real-time energy consumption data, real-time emission data, etc.;

[0170] Static data refers to data that already exists and is relatively stable before the carbon footprint assessment is conducted, including but not limited to the average energy consumption of product raw materials, average emission factors, etc.;

[0171] The step 305 of determining a data storage scheme and selecting a corresponding database and file system to store real-time data and static data specifically refers to: selecting appropriate storage methods for real-time data and static data respectively. Due to its timeliness and dynamism, real-time data is usually stored in a database or file system that can read and write quickly and supports streaming processing; and static data can be stored in a traditional relational database or file system due to its stability;

[0172] Step 306: Plan the data processing flow, including data cleaning, conversion, and aggregation steps;

[0173] Step 307: Build the hardware and software environment of the big data platform according to the architecture; install the database management system, the big data processing framework Hadoop and the data analysis tool Tableau, and configure the network and security policies;

[0174] Step 308: Receive the real-time data stream from the IoT device through the real-time data receiving interface, and use the message queue Kafka and the stream processing framework Apache Flink to manage the reception and distribution of the real-time data stream;

[0175] Step 309: Clean the received real-time data, remove noise and outliers, and convert the data format;

[0176] Step 310: Integrate the preprocessed real-time data with the static data set to form a complete data set;

[0177] Step 311: Use data warehouse and data lake technologies to store and manage data sets, and build data indexing and partitioning strategies;

[0178] Step 312: Verify the integrated data, including data integrity check and data consistency check.

[0179] As a preferred option, an artificial intelligence algorithm module is integrated for data processing, analysis and prediction to provide intelligent support; the specific steps are as follows:

[0180] Step 401: Determine that the prediction target is the carbon emission trend of the product at different life cycle stages, analyze the time series characteristics of the data, including periodicity, seasonality, and trend, and select L-step TM as the trend prediction algorithm;

[0181] Step 402: Install Python and its necessary libraries, where NumPy and Pandas are used for data processing, and TensorFlow and PyTorch are used for deep learning model construction;

[0182] Step 403: Use Panda steps to read and preprocess data, including missing value processing, outlier detection and correction;

[0183] Step 404: extracting features of the time series data, including time step and window size;

[0184] Step 405: Divide the data set into a training set, a validation set, and a test set;

[0185] Step 406: Use TensorFlow and PyTorch to define the L-step TM model structure, including input layer, hidden layer, and output layer; configure model parameters, including learning rate, optimizer, and loss function;

[0186] Step 407: Convert the training set data into a format acceptable to the model, i.e., a three-dimensional array of [number of samples, time steps, number of features];

[0187] Step 408: Use the training set data to train the L-step TM model and adjust the model parameters through the back propagation algorithm;

[0188] Step 409: Use the validation set to evaluate the generalization ability of the model, and use the test set to test the final model;

[0189] Step 410: Encapsulate the trained L-step TM model into independent modules and define the data exchange interface between the modules;

[0190] Step 411: When new real-time data arrives, it is passed to the L-step TM module through the data interface;

[0191] Step 412: The L-step TM module processes the real-time data and outputs a prediction result, that is, the carbon emission trend for a period of time in the future.

[0192] As a preference, adjust the system boundaries and select the life cycle stages that need to be included in the assessment according to the assessment requirements; the specific steps are as follows:

[0193] Step 501: Input user requirement document, system boundary document, and product life cycle information;

[0194] Step 502: Analyze the user requirements document to clarify the specific goals, scope and requirements of the assessment;

[0195] Step 503: Review the current system boundary document and identify the life cycle phases and activities included;

[0196] Step 504: Determine whether to add or delete specific life cycle stages and activities based on user needs;

[0197] Step 505: Update the assessment scope description;

[0198] Step 506: Input the updated assessment scope description and the original system boundary document;

[0199] Step 507: Redraw and adjust the system boundary map according to the updated evaluation scope;

[0200] Step 508: Identify the processes and activities that need to be included in the assessment scope and mark them in the system boundary diagram;

[0201] Step 509: Remove processes and activities that are no longer within the scope of the assessment;

[0202] Step 510: Update the system boundary document to describe in detail the new system boundary and the processes and activities it contains.

[0203] As a preference, select corresponding static and dynamic data sets from the modular data platform, combine and configure them, and customize LCA methods according to product characteristics and evaluation objectives, including process LCA, input-output LCA and hybrid methods; the specific steps are as follows:

[0204] Step 601: input the modular data platform and the updated system boundary document;

[0205] Step 602: Access the modular data platform to obtain all available static and dynamic data sets;

[0206] Step 603: Filter out data sets directly related to the evaluation scope according to the updated system boundary;

[0207] Step 604: Mark the filtered data set as "selected"

[0208] Step 605: input the selected data set and evaluation requirements document;

[0209] Step 606: Analyze the evaluation requirements document and clarify the specific requirements for data configuration, including sampling frequency and time period;

[0210] Step 607: configure the selected data set, set an appropriate data sampling frequency and time period; and verify whether the configured data set meets the evaluation requirements;

[0211] Step 608: Input product feature description, evaluation target document, and LCA method library;

[0212] Step 609: Analyze the product feature description to identify its unique environmental impact characteristics and assessment difficulties;

[0213] Step 610: Browse the LCA method library to understand the advantages and disadvantages and applicable scope of the LCA method, and select the most appropriate LCA method based on product characteristics and evaluation objectives;

[0214] Step 611: input the selected LCA method, assessment requirements document, and product feature description;

[0215] Step 612: Identify the parts that need customized adjustment based on the evaluation requirements document and product feature description, including calculation methods and analysis steps;

[0216] Step 613: Customize the selected LCA method and verify whether the customized LCA method can effectively solve the assessment problem.

[0217] As a preferred method, AI algorithms are used to intelligently process the collected data, including cleaning, preprocessing, classification and trend prediction; the specific steps are as follows:

[0218] Step 701: Start the intelligent algorithm module and load the original data related to the product life cycle, including data on raw material procurement, production and manufacturing, transportation and distribution, use stage, and waste treatment stage;

[0219] Step 702: Based on the characteristics of carbon footprint data, a statistical method of threshold detection based on industry standards and historical data is used to identify abnormal values ​​in the data;

[0220] Step 703: Delete the abnormal values ​​and replace them with reasonable values;

[0221] Step 704: searching and identifying duplicate data records through a hash function data comparison algorithm, and deleting duplicate data items;

[0222] Step 705: storing the cleaned data back to the data platform and marking it as cleaned carbon footprint related data;

[0223] Step 706: Loading cleaned carbon footprint related data from the data platform;

[0224] Step 707: convert the data into a unified format, normalize the data that needs to be compared or comprehensively calculated, and process the data in segments according to different stages of the product life cycle;

[0225] Step 708: storing the pre-processed data back to the data platform and marking it as pre-processed carbon footprint related data;

[0226] Step 709: Loading the pre-processed carbon footprint related data from the data platform;

[0227] Step 710: extracting features related to carbon emissions from the data, including energy consumption data at different stages, types of raw materials, and production processes;

[0228] Step 711: Select a cluster analysis classification algorithm to classify carbon emission sources; select a time series analysis trend prediction algorithm, and use historical carbon emission data to train the trend prediction model, optimize model parameters, and predict future carbon emission trends;

[0229] Step 712: Classify and predict the trend of the preprocessed data to identify key emission sources and carbon emission trends;

[0230] Step 713: Verify the classification and prediction results and output the classification and prediction results.

[0231] As a priority, flexibly select analysis methods and tools according to assessment needs to conduct carbon footprint accounting, hot spot analysis, sensitivity analysis and uncertainty assessment; the specific steps are as follows:

[0232] Step 801: Determine the analysis objectives, including identifying emission reduction potential and optimizing the supply chain;

[0233] Step 802: Evaluate the applicability and advantages and disadvantages of different analysis methods under the LCA framework, including carbon footprint accounting, hot spot analysis, and sensitivity analysis;

[0234] Step 803: Select an analysis method based on evaluation requirements and data characteristics;

[0235] Step 804: Select an LCA software analysis tool according to the selected analysis method;

[0236] Step 805: configure the selected analysis tool, set parameters, including calculation boundaries, emission factors, etc., and import cleaned and pre-processed data;

[0237] Step 806: Execute the LCA carbon footprint assessment task using the configured analysis tool;

[0238] Step 807: Collating the analysis results, including calculating carbon emissions at each stage, identifying key emission sources, and evaluating emission reduction potential;

[0239] Step 808: Output the analysis results.

[0240] As a preference, visualization tools are used to display and output the analysis results in the form of charts and animations; the specific steps are as follows:

[0241] Step 901: Plan a visualization scheme based on the analysis results and data characteristics, including chart type, color matching, and layout design;

[0242] Step 902: Import the analysis results into a visualization tool;

[0243] Step 903: Creating visualization elements of charts and animations in a visualization tool;

[0244] Step 904: Construct the structure and content of the visualization report according to the assessment requirements and communication goals;

[0245] Step 905: Filling the visualization elements and analysis results into the report;

[0246] Step 906: Export the report into PDF and PPT formats.

[0247] The process of assessing the carbon footprint of computer products involves multiple links, including the acquisition of raw materials, production and manufacturing, transportation and distribution, use, and waste disposal. Traditional methods often have problems such as incomplete data collection, low processing efficiency, and inaccurate analysis results when processing these data.

[0248] The data comparison results of the carbon footprint assessment of computer products using the method of the present invention and the prior art are shown in Table 1.

[0249] Table 1 Experimental data comparison table

[0250] Method of the present invention Prior art Data Collection 98％ 70％ Data processing efficiency 30% improvement - Accuracy of analysis results Error <5% Error about 15% Identification of key emission sources Accurate to specific parts Rough classification Carbon footprint accounting Detailed accounting of the entire life cycle Partial stage accounting Emission reduction potential assessment Propose specific emission reduction measures -

[0251] In summary, the method of the present invention can collect 98% of the relevant data in the entire life cycle of computer products through a modular data platform and a real-time data integration module, while the existing technology can only collect about 70% of the data, and there is a problem of data missing; the method of the present invention integrates an artificial intelligence algorithm module, which can realize intelligent processing of data, including cleaning, preprocessing, classification and trend prediction, etc., and the processing efficiency is improved by more than 30% compared with the existing technology; the method of the present invention adopts advanced LCA methods and tools for carbon footprint accounting, and the error of the analysis result is less than 5%, while the error of the existing technology often reaches 15% due to the limitations of data processing and analysis methods. The method of the present invention can accurately identify the key emission sources in the entire life cycle of computer products, even to a specific component or production link, while the existing technology can only roughly classify and identify them; the method of the present invention can perform detailed carbon footprint accounting for the entire life cycle of computer products, including raw material acquisition, production and manufacturing, transportation and distribution, use stage and waste treatment, while the existing technology can only calculate some stages; the method of the present invention can propose specific emission reduction measures and suggestions based on the analysis results, helping enterprises to reduce carbon footprints and achieve sustainable development of the environment, while the existing technology can only talk about it in general terms, lacking pertinence and operability;

[0252] Through the above data comparison, we can intuitively see the advantages of the method of the present invention in the carbon footprint assessment of computer products; it not only improves the accuracy and efficiency of the assessment, but also provides companies with more targeted emission reduction measures and suggestions, which helps promote green and low-carbon transformation.

[0253] Compared with the product carbon footprint assessment methods under the existing technology, which often use fixed frameworks and preset data sets, it is difficult to adapt to the changes of different products and assessment needs, and often rely on manual operations in data processing and analysis, which is inefficient and prone to errors. This invention allows users to flexibly customize the LCA method according to specific product characteristics and assessment needs through a modular data platform and adjustable system boundary design, including selecting life cycle stages, data sets and assessment methods, thereby greatly improving the flexibility and pertinence of the assessment. At the same time, it integrates artificial intelligence algorithm modules to realize intelligent processing, analysis and prediction of data, reduce manual intervention, improve data processing efficiency and accuracy, and can predict future carbon emission trends, providing strong support for decision-making;

[0254] By using the Internet of Things and big data technologies to build a real-time data integration module, it supports the real-time collection of carbon emission data at each life cycle stage; integrates artificial intelligence algorithm modules for data processing, analysis and prediction to provide intelligent support; adjusts the system boundaries according to the assessment needs, and selects the life cycle stages that need to be included in the assessment; selects the corresponding static and dynamic data sets from the modular data platform, combines and configures them, and customizes the LCA method according to product characteristics and assessment objectives, including process LCA, input-output LCA and hybrid methods; thus, through the modular data platform and adjustable system boundary design, allows users to flexibly customize the LCA method according to specific product characteristics and assessment needs, including selecting life cycle stages, data sets and assessment methods, thereby greatly improving the flexibility and pertinence of the assessment;

[0255] By determining the prediction target as the carbon emission trend of the product at different life cycle stages, analyze the time series characteristics of the data, including periodicity, seasonality, and trend, and select LSTM as the trend prediction algorithm; install Python and its necessary libraries, including NumPy and Pandas for data processing, TensorFlow and PyTorch for deep learning model construction; use Pandas to read and preprocess data, including missing value processing, outlier detection and correction; extract the characteristics of time series data, including time step and window size; divide the data set into training set, validation set and test set; use TensorFlow and PyTorch to define the LSTM model structure, including input layer, hidden layer, output layer; configure model parameters, including learning rate, optimizer, loss function, etc. Loss function; convert the training set data into a format acceptable to the model, that is, a three-dimensional array of [number of samples, time steps, number of features]; use the training set data to train the LSTM model and adjust the model parameters through the back propagation algorithm; use the validation set to evaluate the generalization ability of the model, and use the test set to test the final model; encapsulate the trained LSTM model into an independent module and define the data exchange interface between modules; when new real-time data arrives, pass it to the LSTM module through the data interface; the LSTM module processes the real-time data and outputs the prediction result, that is, the carbon emission trend in the future; thereby realizing intelligent processing, analysis and prediction of data, reducing human intervention, improving data processing efficiency and accuracy, and being able to predict future carbon emission trends, providing strong support for decision-making.

[0256] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A flexible formulation method for a product carbon footprint model based on LCA, characterized in that: include: Determine the carbon footprint assessment needs and build an assessment framework based on the needs; Divide LCA data into data sets according to life cycle stages; Real-time collection of carbon emission data during the life cycle; Build an LSTM model to predict the carbon emission trend of products at different life cycle stages; Carbon footprint accounting, hotspot analysis, sensitivity analysis and uncertainty assessment are performed based on the LSTM model.

2. The method for flexibly formulating a product carbon footprint model based on LCA as claimed in claim 1, characterized in that: Determining the need for carbon footprint assessment includes identifying and determining specific products for which carbon footprint assessment is required; the selected products may be physical goods or services; Determine the carbon footprint assessment needs of the product based on its attributes; Determine the life cycle stages that need to be included and set emission reduction targets based on the assessment objectives and product characteristics; The life cycle stages include raw material acquisition, production, transportation, use and waste treatment; The construction of the assessment framework includes integrating the assessment scope and generating an assessment requirements document based on the identified product types, life cycle stages and set emission reduction targets; Define the system boundaries based on the assessment requirements document, including the processes and activities that need to be included in the assessment at each life cycle stage; Set up adjustable system boundary interfaces; determine unified functional units; select and define key environmental impact indicators; The constructed assessment framework is documented, including descriptions of system boundaries, functional units, and environmental impact indicators.

3. The method for flexibly formulating a product carbon footprint model based on LCA as claimed in claim 2, characterized in that: The data set includes a static data set and a dynamic data set; The static data set includes data that does not change over time during the evaluation process; The dynamic data set includes data that changes over time.

4. The method for flexibly formulating a product carbon footprint model based on LCA as claimed in claim 3, characterized in that: The real-time collection of carbon emission data during the life cycle stage includes determining the location of deploying the IoT device according to the life cycle stage; Select IoT devices and plan the number of IoT devices to be deployed; Deploy IoT devices according to the plan and configure the devices; Build the overall architecture of the big data platform, including components of the data storage layer, processing layer, and analysis layer; Determine the data storage solution and select the corresponding database and file system to store real-time data and static data.

5. The method for flexibly formulating a product carbon footprint model based on LCA as claimed in claim 4, characterized in that: The construction of the LSTM model for predicting the carbon emission trend of products at different life cycle stages includes determining the prediction target of carbon emissions based on the different life cycle stages of the products, analyzing the time series characteristics of historical data, including periodicity, seasonality and trend characteristics, and selecting a long short-term memory neural network; Build a Python environment and install data processing libraries and deep learning libraries; Use Pandas to read the original data set related to carbon emissions, clean the data, remove missing values ​​and outliers, and perform data standardization operations; Perform feature extraction on the cleaned data, set the time step and window size, convert the historical data into a format suitable for time series analysis, and capture time dependencies; Divide the extracted time series data into training set, validation set and test set; Use TensorFlow or PyTorch to build an LSTM model, including an input layer, multiple hidden layers, and an output layer; configure hyperparameters, including learning rate, optimizer, and loss function, and initialize model weights; Convert the training set data into the three-dimensional array format required by the LSTM model, including the number of samples, time steps, and features, and batch process the data; Use the training set data to adjust the model weights through the back-propagation algorithm and stochastic gradient descent optimizer, and adopt a dynamic learning rate adjustment strategy; The performance of the trained LSTM model is evaluated through the validation set, and the root mean square error and mean absolute error are calculated; Encapsulate the verified LSTM model into an independent module and define the data input and output interfaces.

6. The method for flexibly formulating a product carbon footprint model based on LCA as claimed in claim 5, characterized in that: The carbon footprint calculation and hot spot analysis include Obtaining predicted carbon emission data for each life cycle stage from the LSTM model, including dynamic trends in raw material acquisition, production, transportation, use, and waste treatment; According to the prediction data of the LSTM model and the demand documents input by the user, parse the user needs and product life cycle information, and automatically analyze the current system boundaries; according to the analysis results, adjust the system boundaries; The adjusting of the system boundary includes parsing and evaluating the requirement document to identify new life cycle activities; adjusting the system boundary to include new activities or exclude irrelevant activities according to data changes; and verifying the adjusted system boundary; Based on the LSTM model prediction data, the carbon emissions of each life cycle stage are calculated within the adjusted system boundary, the emission data of each stage are classified, and the main carbon emission sources are identified.

7. The method for flexibly formulating a product carbon footprint model based on LCA according to claim 6, characterized in that: The sensitivity analysis and uncertainty assessment include screening relevant static and dynamic data sets according to the adjusted system boundaries; screening includes parsing user requirement documents and automatically screening data sets that meet the assessment requirements; setting data sampling frequency and time period; making adjustments based on LSTM prediction data and assessing the impact of each parameter on the carbon footprint; Randomly sample the prediction results of the LSTM model to quantify the fluctuation range of the prediction data.

8. A system using the flexible formulation method of a product carbon footprint model based on LCA as claimed in any one of claims 1 to 7, characterized in that: include: A framework construction module (100), a stage division module (200), a data collection module (300), a model construction module (400) and a prediction analysis module (500); The framework building module (100) is used to determine carbon footprint assessment requirements and build an assessment framework according to the requirements; The stage division module (200) is used to divide the LCA data into data sets according to life cycle stages; The data collection module (300) is used to collect carbon emission data in the life cycle stage in real time; The model building module (400) is used to build an LSTM model for predicting the carbon emission trend of products at different life cycle stages; The prediction analysis module (500) is used to perform carbon footprint calculation, hot spot analysis, sensitivity analysis and uncertainty assessment based on the LSTM model.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for flexibly formulating a product carbon footprint model based on LCA according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for flexibly formulating a product carbon footprint model based on LCA according to any one of claims 1 to 7 are implemented.

Images