Personalized low-carbon manufacturing system and method

Abstract

The invention discloses a personalized low-carbon manufacturing system and method in the technical field of environmental protection manufacturing. The personalized low-carbon manufacturing system comprises distributed Internet of Things terminals deployed in manufacturing enterprises; the distributed Internet of Things terminal comprises a sensing matrix module used for collecting energy consumption, raw materials and process parameter data of each manufacturing enterprise. And the credible low-carbon information module is used for docking the dynamic update data of the ecological environment department of each manufacturing enterprise. The personalized carbon effect evaluation module is used for tracing carbon emission of each link of a supply chain based on a product carbon footprint dynamic decomposition algorithm and order parameters; the low-carbon scheme model is used for simulating carbon effect changes of different industrial combinations of each manufacturing enterprise; and the credibility analysis module is used for carrying out credibility scoring on the carbon emission of each manufacturing enterprise based on the scoring weight factors. According to the invention, through credible data management, dynamic carbon efficiency optimization and cross-chain privacy collaboration, the accuracy of data acquisition of small and medium-sized enterprises is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of environmental protection manufacturing, and specifically relates to a personalized low-carbon manufacturing system and method. Background Art

[0002] With the continuous growth of the global economy, the demand for resources and energy is also increasing continuously. However, the resources and energy on the earth are limited. Over-exploitation and consumption will not only lead to resource exhaustion, but also trigger a series of environmental problems. Therefore, how to utilize resources and energy efficiently and sustainably has become a major challenge facing mankind. As a new manufacturing model, low-carbon manufacturing aims to alleviate this pressure by improving the efficiency of energy and resource utilization and reducing waste and emissions. The concept of sustainable development emphasizes meeting the needs of the current generation without compromising the ability of future generations to meet their own needs. This concept requires that we must pay attention to environmental protection and resource conservation in the process of economic development. Low-carbon manufacturing is proposed based on this concept, which requires enterprises to not only pursue economic benefits in the production process, but also pay attention to the unity of social benefits and environmental benefits.

[0003] In the context of globalization and knowledge-based economy, market competition has further intensified. In a personalized consumption environment, more and more customers determine their consumption in completely different forms, and the market is required to meet their personalized products for specific purposes, specific environments and specific times. However, manufacturing enterprises will also face many new key problems when pursuing personalized low-carbon manufacturing. For example, it is difficult to obtain and organize low-carbon information. Especially for some small and medium-sized manufacturing enterprises with imperfect organizational structures, it is not only difficult to obtain their corresponding carbon emission index information, but also this information is particularly messy and filled with a large amount of false information, which will affect the subsequent utilization of information.

[0004] Therefore, it is necessary to propose a personalized low-carbon manufacturing system and method to solve the above problems. Summary of the Invention

[0005] In order to solve the above problems, the purpose of the present invention is to provide a personalized low-carbon manufacturing system and method, which can improve the accuracy of data collected from small and medium-sized enterprises through trusted data governance, dynamic carbon efficiency optimization, and cross-chain privacy collaboration.

[0006] To achieve the above purpose, the technical solution of the present invention is as follows: A personalized low-carbon manufacturing system includes distributed Internet of Things terminals deployed in each manufacturing enterprise;

[0007] The distributed Internet of Things terminals include:

[0008] A sensing matrix module for collecting energy consumption, raw material, and process parameter data of each manufacturing enterprise and defining them as initial parameters;

[0009] A data cleaning module for eliminating outliers and duplicate data in the initial parameters based on dual-channel filtering technology;

[0010] A trusted low-carbon information module for docking with the dynamically updated data of the ecological environment department of each manufacturing enterprise; protecting business privacy using zero-knowledge proof technology;

[0011] A personalized carbon efficiency assessment module for tracing carbon emissions in all links of the supply chain based on the dynamic decomposition algorithm of product carbon footprint and order parameters;

[0012] A low-carbon solution model for simulating the carbon efficiency changes of different industrial combinations of each manufacturing enterprise;

[0013] A credibility analysis module for performing credibility scoring on the carbon emissions of each manufacturing enterprise based on scoring weight factors and outputting the credibility score value of the carbon emissions of each manufacturing enterprise; the scoring weight factors include sensor calibration status, blockchain cross-verification matching degree, and historical data deviation rate.

[0014] Furthermore, the dual-channel filtering technology includes:

[0015] A dual-threshold filtering mechanism, where the hard threshold is based on device technical parameters and the soft threshold is dynamically adjusted through the LSTM prediction interval;

[0016] A perception completion algorithm for intelligently filling in from historical data according to equipment working condition similarity when key data is missing;

[0017] A traceability marking unit for attaching credibility labels to the cleaned data.

[0018] Furthermore, the trusted low-carbon information module is also used for storing conventional data fraud factors; the conventional data fraud factors include characteristics of meter pulse interference and traces of non-linear tampering of gas flow; it is also used for performing energy consumption and output material conservation verification on newly stored data, detecting data mutations simultaneously, and making horizontal comparisons with data of the same region and the same industry.

[0019] Furthermore, in the perception completion algorithm, unstructured data provided by suppliers is parsed through NLP, carbon emissions are allocated as needed for different scenarios, and the confidence interval of carbon data is calculated using Monte Carlo simulation.

[0020] Furthermore, in the low-carbon solution model, the multi-objective optimization space of different process routes is visualized, docked with the global low-carbon material database and updated in real time, and then the compatibility threshold between the solution and the existing production line is verified.

[0021] Furthermore, the distributed IoT terminal also includes a crowdsourcing verification module; the crowdsourcing verification module is used for randomly assigning 3 non-competing enterprises for blind data review.

[0022] Further, in the credibility analysis module, when the industry benchmark data is updated, the scoring weights are automatically recalculated, and the LSTM is used to predict the data deviation trend in the next 10 - 15 days, and the scoring strategy is adjusted in advance.

[0023] Further, it also includes a dynamic optimization execution module, which is used to generate process parameter adjustment instructions based on the real-time credibility score and carbon efficiency evaluation results through a reinforcement learning algorithm, and use digital twin technology to simulate the carbon efficiency gain after the execution of the instructions. When the gain threshold exceeds 10% - 15%, the production line controller is automatically triggered to implement low-carbon strategy iteration.

[0024] Further, the dynamic optimization execution module is connected to the cross-chain verification gateway. When multi-link collaborative optimization of the supply chain is involved, the data trust labels of upstream and downstream enterprises are synchronously verified through cross-chain smart contracts, and multi-party secure computing technology is used to achieve privacy-protected joint solution of cross-enterprise carbon emission optimization schemes.

[0025] Further, a personalized low-carbon manufacturing method is as follows:

[0026] Step 1: Through the sensing matrix module of the distributed IoT terminal, collect the energy consumption, raw materials, and process parameter data of each manufacturing enterprise to generate initial parameters;

[0027] Step 2: Eliminate outliers through hard and soft thresholds, use the perception completion algorithm to fill in the missing key data according to the equipment working condition similarity, and attach traceability credibility labels to the cleaned data;

[0028] Step 3: Connect to the dynamically updated data of the Ministry of Ecology and Environment, and verify the data integrity in combination with zero-knowledge proof technology; detect fraud factors, and verify the data authenticity through energy consumption-output conservation verification and horizontal comparison with the same industry;

[0029] Step 4: Based on the dynamic decomposition algorithm of product carbon footprint, trace the carbon emissions of each link in the supply chain; use Monte Carlo simulation to calculate the carbon emission confidence interval, generate a multi-objective optimization plan in combination with the global low-carbon material database, and verify the compatibility threshold between the optimization plan and the existing production line;

[0030] Step 5: Generate process parameter adjustment instructions through a reinforcement learning algorithm, use digital twin to simulate carbon efficiency gain, trigger production line iteration when the gain exceeds 10% - 15%, and at the same time, the cross-chain verification gateway synchronously verifies the trust labels of upstream and downstream enterprises, and realizes privacy-protected cross-enterprise joint optimization based on multi-party secure computing.

[0031] Beneficial effects of adopting this solution: 1. This solution significantly improves the authenticity of carbon emission data and the reliability of decision-making through multi-dimensional data trusted governance and intelligent cleaning technology. In the traditional manufacturing system, data collection is vulnerable to equipment errors, human tampering, and data silos, resulting in distorted carbon emission accounting. This solution uses dual-channel filtering technology (hard threshold + dynamic LSTM soft threshold) to accurately eliminate outliers, and combines a perception completion algorithm to fill in key missing data to ensure data integrity; the traceability marking unit and credibility label further establish a trusted link throughout the data life cycle. At the same time, the trusted low-carbon information module has a fraud feature library such as electricity meter pulse interference and gas tampering traces, and blocks data fraud from the source through energy consumption-output conservation verification and industry horizontal comparison.

[0032] 2. This solution realizes dynamic traceability and personalized optimization of the carbon footprint of the entire supply chain, breaking through the locality and lag of traditional carbon management. Traditional carbon management relies on static accounting models and cannot track the carbon emissions of complex supply chain links in real time. This solution is based on the dynamic decomposition algorithm of product carbon footprint, combines order parameters to parse unstructured data of suppliers in real time (such as NLP parsing of logistics documents), and uses Monte Carlo simulation to quantify the confidence interval of carbon emissions, realizing the traceability of the carbon footprint of the entire chain from raw material procurement, production to logistics. The low-carbon solution model is connected to the global low-carbon material database to generate multi-objective optimization solutions (such as process route replacement, low-carbon material matching), and ensures the feasibility of the solutions through compatibility threshold verification.

[0033] 3. This solution integrates dynamic execution and cross-chain privacy collaboration to build a closed-loop low-carbon ecosystem, promoting enterprises to shift from passive compliance to active innovation. Traditional low-carbon transformation relies on manual experience and single-point optimization, making it difficult to achieve cross-enterprise collaboration and dynamic response. This solution uses a dynamic optimization execution module to generate process parameter adjustment instructions in real time based on reinforcement learning algorithms, and uses digital twin simulation to verify carbon efficiency gains. When the gain exceeds 15%, the production line is automatically triggered for iteration, realizing a closed loop of "perception - decision - execution". For example, a steel enterprise optimized blast furnace parameters through digital twin simulation, with a 22% increase in carbon efficiency and a 30% reduction in the production line iteration cycle. In addition, the cross-chain verification gateway and multi-party secure computing technology break data barriers: synchronously verify the trusted labels of upstream and downstream data through cross-chain smart contracts, and jointly solve the optimization solution for privacy protection. Brief Description of the Drawings

[0034] Figure 1 It is a flowchart of an embodiment of the personalized low-carbon manufacturing system of the present invention.

[0035] Figure 2 It is a schematic diagram of an embodiment of the personalized low-carbon manufacturing method of the present invention. Detailed Description of the Invention

[0036] The following is a further detailed description through specific embodiments:

[0037] Example 1:

[0038] Basically as shown in the appendix Figure 1 : A personalized low-carbon manufacturing system, including distributed Internet of Things terminals deployed in each manufacturing enterprise;

[0039] The distributed Internet of Things terminals include: a sensing matrix module, used to collect energy consumption, raw material, and process parameter data of each manufacturing enterprise and define them as initial parameters; a data cleaning module, used to eliminate outliers and duplicate data in the initial parameters based on dual-channel filtering technology.

[0040] Among them, the dual-channel filtering technology includes: a dual-threshold filtering mechanism, where the hard threshold is based on device technical parameters, and the soft threshold is dynamically adjusted through the LSTM prediction interval; for example, in the hard threshold setting, the rated power of a certain injection molding machine is 50kW, and a ±10% hard threshold (45 - 55kW) is set, and data outside the range is directly eliminated. In the dynamic adjustment of the soft threshold, the LSTM model predicts the daily power fluctuation interval based on the electricity consumption data of the past 30 days (the confidence level of the prediction interval is 95%). If the actual power on a certain day reaches 58kW but is within the prediction interval (55 - 60kW), it is retained. A perception completion algorithm, when key data is missing, intelligently fills it from historical data according to the similarity of equipment working conditions; when the temperature sensor in a stamping workshop fails and causes data loss: 1. Retrieve the historical working condition library (similar working conditions with pressure ≥ 200 tons and rotational speed ≥ 1200 rpm); 2. Select the temperature data under the same working conditions in the recent 7 days (average 365℃ ± 8℃); 3. Generate a filled value (368℃ ± 5℃) using Bayesian regression; 4. Attach a yellow traceability label (confidence level 85%). A traceability marking unit attaches a credibility label to the cleaned data.

[0041] In the perception completion algorithm, unstructured data provided by suppliers is parsed through NLP, and carbon emissions are allocated on demand for different scenarios, while the confidence interval of carbon data is calculated using Monte Carlo simulation.

[0042] Specifically, for example, the "Production Process Description" in PDF format provided by an automotive parts supplier contains unstructured text: "This batch of wheels uses the low-pressure casting process. The melting furnace is heated to 720 ± 15℃ using natural gas and switched to biomass fuel during the insulation stage. Each molding takes 45 minutes, and the scrap rate is controlled below 3%". The specific explanation process is as follows:

[0043] Step 1, entity extraction, using the BERT + BiLSTM model to identify key parameters:

[0044]

[0045] Step 2, structured mapping, generating a standardized data table:

[0046] Parameter Value Unit Confidence level Melting fuel type Natural gas - 88％ Heat preservation fuel type Biomass fuel - 91％ Melting temperature 720 ℃ 89％ Temperature fluctuation ±15 ℃ 89％ Single mold time 45 Minute 95％

[0047] Step 3: Conflict resolution. When the same parameter appears multiple times (such as the fuel type described in different paragraphs), the following rules are adopted:

[0048]

[0049] The trustworthy low-carbon information module is used to dock with the dynamically updated data of the ecological environment department of each manufacturing enterprise; the zero-knowledge proof technology is adopted to protect business privacy;

[0050] The credibility analysis module is used to conduct a credibility score on the carbon emissions of each manufacturing enterprise based on the scoring weight factors and output the credibility score value of the carbon emissions of each manufacturing enterprise; the scoring weight factors include the sensor calibration status, the blockchain cross-verification matching degree, and the historical data deviation rate. The trustworthy low-carbon information module is also used to store the conventional data fraud factors; the conventional data fraud factors include fraud factors such as the electric meter pulse interference characteristics and the non-linear tampering traces of the gas flow rate. It is also used to perform energy consumption and output material conservation verification on the newly stored data, detect sudden data drops at the same time, and make horizontal comparisons with the data of the same region and the same industry. In the credibility analysis module, when the industry benchmark data is updated, the scoring weight is automatically recalculated, and the LSTM is used to predict the data deviation trend in the next 10-15 days and adjust the scoring strategy in advance.

[0051] For the verification of credibility, for example, in the analysis of the electric meter pulse characteristics, the normal pulse interval standard deviation should be <0.02 seconds, and an abnormal fluctuation of 0.15 seconds appears in the data of a certain enterprise, triggering a tampering alarm; in the conservation verification, a certain steel plant reports a comprehensive energy consumption of 580 kgce per ton of steel, but according to the hot metal ratio (0.92) and the scrap ratio (0.15), the theoretical value should be ≥605 kgce, and the data is judged to be abnormal. Then, blockchain cross-verification is carried out. The VOCs emission data (hash value 0x7d3...) uploaded by a certain chemical enterprise is compared with the data on the environmental protection department's chain, and it is found that the timestamp offset is >30 minutes, and the credibility score is downgraded (from level B to level C).

[0052] The distributed IoT terminal also includes a crowdsourcing verification module; the crowdsourcing verification module is used to randomly assign 3 non-competing enterprises for blind data review. When a data mutation is detected (such as a unit energy consumption drop of >25% within 24 hours), a three-level verification mechanism is started: (1) local sensor re-inspection; (2) comparison with the data of the same region and the same industry (Euclidean distance ≤0.3); (3) crowdsourcing blind review (independent verification by 3 non-competing enterprises).

[0053] The personalized carbon efficiency assessment module is used to trace the carbon emissions of each link in the supply chain based on the product carbon footprint dynamic decomposition algorithm and order parameters. Among them, a hierarchical traceability model is adopted to decompose the product carbon footprint into:

[0054]

[0055] Among them, E proc : Process energy consumption (kWh), which collects the power of the machine tool servo motor in real time through the OPC-UA protocol (±0.5% accuracy); E trans : Transportation energy consumption, calculated based on GPS trajectory data and load factor, and the load compensation factor ξi = 1 + 0.05 (Wactual / Wmax - 1) ξi = 1 + 0.05 (Wactual / Wmax - 1); C mat : Embodied carbon in materials, docking with the global material database (including LCA data of more than 5,000 materials), and the matching error rate ≤ 1.2%; ρ: Process correction coefficient, dynamically adjusted through Monte Carlo simulation (iterated 5,000 times).

[0056] For the supply chain map empowered by blockchain, each node contains: Process parameter hash value: Generate an irreversible digest using the SHA-3 algorithm; Carbon data trusted label: Based on the endorsement strategy of Hyperledger Fabric (requiring 2 / 3 node verification).

[0057] The low-carbon solution model is used to simulate the carbon efficiency changes of different industrial combinations of each manufacturing enterprise. In the low-carbon solution model, visualize the multi-objective optimization space of different process routes, dock with the global low-carbon material database and update it in real time, and then verify the compatibility threshold between the solution and the existing production line. For example, an aluminum alloy wheel manufacturing enterprise needs to select among three process routes: traditional casting (A), semi-solid forming (B), and 3D printing (C), aiming to achieve the optimal balance among carbon emissions (CE), cost (Cost), and yield (Yield). First, call the global low-carbon material database (such as MaterialProject) to obtain candidate material parameters and establish a three-dimensional decision space. The optimization algorithm in it can choose the non-dominated sorting genetic algorithm (population size: 200, crossover probability: 0.85, mutation probability: 0.02, iteration times: 1000, specific parameters are determined according to the actual situation). Finally, conduct solution compatibility verification, and thus select a suitable process plan according to the verification results.

[0058] The distributed IoT terminal also includes a dynamic optimization execution module, which is used to generate process parameter adjustment instructions through a reinforcement learning algorithm based on real-time credibility scores and carbon efficiency evaluation results, and use digital twin technology to simulate the carbon efficiency gain after the execution of the instructions. When the gain threshold exceeds 15%, the production line controller is automatically triggered to implement low-carbon strategy iteration. The dynamic optimization execution module is connected to the cross-chain verification gateway. When it comes to collaborative optimization of multiple supply chain links, it synchronously verifies the data credibility tags of upstream and downstream enterprises through cross-chain smart contracts, and uses multi-party secure computing technology to achieve privacy-protected joint solution of cross-enterprise carbon emission optimization schemes.

[0059] Embodiment 2:

[0060] The difference from the above embodiment is that, as Figure 2 shown, a personalized low-carbon manufacturing method is as follows:

[0061] Step 1: Through the sensing matrix module of the distributed IoT terminal, collect the energy consumption, raw materials, and process parameter data of each manufacturing enterprise to generate initial parameters; the detection sensors are configured according to the actual situation. For example:

[0062] Injection molding machine unit: Type: Piezoelectric vibration sensor (PCB352C33), Sampling rate: 10kHz, Parameters: Clamping force (0 - 3000kN), Barrel temperature (0 - 400°C); Energy consumption monitoring: Equipment: Three-phase power quality analyzer (Fluke435-II), Parameters: Instantaneous power (kW), Harmonic distortion rate (%).

[0063] After data collection, align the corresponding equipment status (OPCUA protocol) and energy consumption (ModbusTCP) data in time.

[0064] The example table of initial parameters is as follows:

[0065] Timestamp Clamping force (kN) Molding temperature (℃) Instantaneous power (kW) 2023-09-2014:05:23.450 2850 235 184.7

[0066] Step 2: Eliminate outliers through hard and soft thresholds, use the perception completion algorithm to fill in the missing key data according to the equipment working condition similarity, and attach traceability credibility tags to the cleaned data.

[0067] Step 3: Connect to the dynamically updated data of the Ministry of Ecology and Environment, and verify the data integrity in combination with zero-knowledge proof technology; detect fraud factors, and verify the data authenticity through energy consumption-output conservation verification and horizontal comparison with the same industry. An enterprise declares an annual output of Q = 120,000 pieces. Without disclosing the actual data, verify the authenticity, and further verify through industry comparative analysis. For example:

[0068]

[0069] Step 4: Based on the dynamic decomposition algorithm of product carbon footprint, trace the carbon emissions of each link in the supply chain; use Monte Carlo simulation to calculate the confidence interval of carbon emissions, generate a multi-objective optimization plan in combination with the global low-carbon material database, and verify the compatibility threshold between the optimization plan and the existing production line.

[0070] Supply chain decomposition case: Carbon footprint decomposition of a new energy vehicle battery pack:

[0071]

[0072] Monte Carlo simulation parameters:

[0073] Variable Distribution type Parameter Data source Positive electrode material carbon intensity Normal distribution μ = 28kg / kg, σ = 3.2 Supplier ESG report Transportation distance Uniform distribution 800 - 1200km Logistics GPS data Yield fluctuation Beta distribution α＝2.5,β＝1.8 Production quality record

[0074] Multi-objective optimization results:

[0075] Solution Carbon emission (kg) Cost (10,000 yuan) Production cycle (days) Compatibility score A 2850 120 45 92％ B 2630 135 38 87％ C 2415 158 42 79％

[0076] Step 5: Generate process parameter adjustment instructions through the reinforcement learning algorithm, use digital twin to simulate carbon efficiency gain, trigger production line iteration when the gain exceeds 15%, and at the same time, the cross-chain verification gateway synchronously verifies the trusted labels of upstream and downstream enterprises, and realizes cross-enterprise joint optimization with privacy protection based on multi-party secure computing.

[0077] The above are only embodiments of the present invention. Specific structures and common knowledge such as characteristics well known in the art are not described in detail here. Those of ordinary skill in the art know all the common general technical knowledge in the technical field to which the invention belongs before the application date or the priority date, can know all the existing technologies in this field, and have the ability to apply the conventional experimental means before this date. Those of ordinary skill in the art can, under the inspiration given in this application, combine their own abilities to improve and implement this solution. Some typical well-known structures or well-known methods should not become an obstacle for those of ordinary skill in the art to implement this application. It should be noted that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can still be made, and these should also be regarded as the protection scope of the present invention, and these will not affect the implementation effect of the present invention and the practicality of the patent. The protection scope required by this application should be based on the content of its claims, and the specific implementation manners and the like recorded in the specification can be used to interpret the content of the claims.

Claims

1. A personalized low-carbon manufacturing system, characterized in that, Including distributed Internet of Things terminals deployed in each manufacturing enterprise; The distributed Internet of Things terminals include: A sensing matrix module, which is used to collect data on energy consumption, raw materials, and process parameters of each manufacturing enterprise and define them as initial parameters; A data cleaning module, which is used to eliminate outliers and duplicate data in the initial parameters based on dual-channel filtering technology; A trusted low-carbon information module, which is used to interface with the dynamically updated data of the ecological environment department of each manufacturing enterprise; and uses zero-knowledge proof technology to protect business privacy; A personalized carbon efficiency assessment module, which is used to trace the carbon emissions of each link in the supply chain based on the product carbon footprint dynamic decomposition algorithm and order parameters; A low-carbon solution model, which is used to simulate the carbon efficiency changes of different industrial combinations of each manufacturing enterprise; A credibility analysis module, which is used to perform a credibility score on the carbon emissions of each manufacturing enterprise based on scoring weight factors and output the credibility score value of the carbon emissions of each manufacturing enterprise; the scoring weight factors include sensor calibration status, blockchain cross-verification matching degree, and historical data deviation rate.

2. The personalized low-carbon manufacturing system according to claim 1, characterized in that: The dual-channel filtering technology includes: A dual-threshold filtering mechanism, where the hard threshold is based on device technical parameters and the soft threshold is dynamically adjusted through the LSTM prediction interval; A perception completion algorithm, which intelligently fills in missing key data from historical data according to the similarity of device operating conditions; A traceability marking unit, which attaches credibility tags to the cleaned data.

3. The personalized low-carbon manufacturing system according to claim 2, characterized in that: The trusted low-carbon information module is also used to store conventional data fraud factors; the conventional data fraud factors include characteristics of meter pulse interference and traces of non-linear tampering of gas flow; it is also used to perform energy consumption and output material conservation verification on newly stored data, detect data mutations at the same time, and make horizontal comparisons with data of the same region and industry.

4. The personalized low-carbon manufacturing system according to claim 3, wherein: In the perception completion algorithm, unstructured data provided by suppliers is parsed through NLP, and carbon emissions are allocated on demand for different scenarios. At the same time, Monte Carlo simulation is used to calculate the confidence interval of carbon data.

5. The personalized low-carbon manufacturing system according to claim 4, wherein: In the low-carbon solution model, the multi-objective optimization space of different process routes is visualized, connected to the global low-carbon material database and updated in real time, and then the compatibility threshold between the solution and the existing production line is verified.

6. The personalized low-carbon manufacturing system according to claim 5, characterized in that: The distributed Internet of Things terminal also includes a crowdsourcing verification module; the crowdsourcing verification module is used to randomly assign 3 non-competing enterprises for blind data review.

7. The personalized low-carbon manufacturing system according to claim 6, characterized in that: In the credibility analysis module, when the industry benchmark data is updated, the scoring weights are automatically recalculated, and the LSTM is used to predict the data deviation trend in the next 10-15 days and adjust the scoring strategy in advance.

8. The personalized low-carbon manufacturing system according to claim 7, wherein: The distributed Internet of Things terminal also includes a dynamic optimization execution module, which is used to generate process parameter adjustment instructions based on real-time credibility scores and carbon efficiency assessment results through a reinforcement learning algorithm, and use digital twin technology to simulate the carbon efficiency gain after the execution of the instructions. When the gain threshold exceeds 10%-15%, it automatically triggers the production line controller to implement low-carbon strategy iteration.

9. The personalized low-carbon manufacturing system according to claim 8, wherein: The dynamic optimization execution module is connected to the cross-chain verification gateway. When multi-link collaborative optimization of the supply chain is involved, it synchronously verifies the data credibility tags of upstream and downstream enterprises through cross-chain smart contracts and uses multi-party secure computing technology to achieve privacy-protected joint solution of cross-enterprise carbon emission optimization schemes.

10. A personalized low-carbon manufacturing method, based on the personalized low-carbon manufacturing system according to any one of claims 1-9, characterized in that, The specific steps are as follows: Step 1: Collect the energy consumption, raw material, and process parameter data of each manufacturing enterprise through the sensing matrix module of the distributed Internet of Things terminal to generate initial parameters. Step 2: Eliminate outliers through hard and soft thresholds, use the perception completion algorithm to fill in the missing key data according to the similarity of equipment operating conditions, and attach traceability credibility labels to the cleaned data. Step 3: Connect to the dynamically updated data of the Ministry of Ecology and Environment, and verify the data integrity in combination with zero-knowledge proof technology. Detect fraud factors, and verify the data authenticity through energy consumption-output conservation verification and horizontal comparison with the same industry. Step 4: Trace the carbon emissions of each link in the supply chain based on the dynamic decomposition algorithm of product carbon footprint; calculate the carbon emission confidence interval using Monte Carlo simulation, generate a multi-objective optimization plan in combination with the global low-carbon material database, and verify the compatibility threshold between the optimization plan and the existing production line. Step 5: Generate process parameter adjustment instructions through the reinforcement learning algorithm, use digital twin to simulate carbon efficiency gain, trigger production line iteration when the gain exceeds 10%-15%, and at the same time, the cross-chain verification gateway synchronously verifies the credible labels of upstream and downstream enterprises, and realizes cross-enterprise joint optimization with privacy protection based on multi-party secure computing.

Images