An industrial enterprise carbon asset whole life cycle intelligent management system

Abstract

The present application relates to the technical field of industrial carbon asset management, in particular to an industrial enterprise carbon asset whole life cycle intelligent management system. The system generates high credible working condition characteristic data through working condition deterioration characteristic sensing and dynamic confidence weighted fusion; the carbon asset accounting algorithm module is internally provided with dynamic carbon factor endogenous accounting unit, state transition evolution unit, double-layer game collaborative optimization unit, etc., discards the fixed factor of national standard, and realizes accurate carbon emission right of five-level minimum unit; nine-state automatic transition and three-layer nested optimization architecture are adopted to generate production and carbon asset collaborative scheduling instructions and issue to generate control system closed loop correction; combined with block chain storage and range three traceability, the whole life cycle credible circulation and intelligent management of carbon assets are completed. The present application realizes quantifiable, rightable, circulatable and checkable carbon assets, and significantly improves the accounting accuracy and operation benefit.

Description

Technical Field

[0001] This invention relates to the field of industrial carbon asset management technology, and more specifically, to an intelligent management system for the entire life cycle of carbon assets in industrial enterprises. Background Technology

[0002] With the advancement of global carbon neutrality goals, industrial enterprises face increasingly stringent carbon emission control requirements. Carbon assets (including carbon emission allowances and certified emission reductions) have become an important new type of asset for enterprises. Currently, industrial enterprises typically rely on independent accounting software or trading systems for carbon management. Their main practices include: calculating carbon emissions using fixed emission factors published by the state or industry; recording carbon allowance holdings and compliance status through spreadsheets or databases; and trading allowances on the carbon trading market. Some advanced enterprises have begun to explore using IoT sensors to collect energy data and employing information systems for carbon emission data statistics. However, existing technologies still have many shortcomings in terms of accounting accuracy, asset status tracking, production collaborative scheduling, cross-entity transfer, and full lifecycle management. Therefore, this paper proposes a smart management system for the entire lifecycle of carbon assets for industrial enterprises.

[0003] The existing technology has the following technical defects, specifically: Existing technologies generally use national standard fixed emission factor lookup tables for calculation, ignoring the dynamic impact of real-time operating conditions such as equipment deterioration, raw material composition fluctuations, process coupling correlations, and production line load changes on actual emissions. This results in serious inaccuracies in carbon emission allocation for micro-units such as branch plants, production lines, processes, equipment, and work teams, and carbon asset confirmation lacks a reliable physical data foundation.

[0004] Traditional systems only record the total amount of carbon allowances held and consumed, without defining standardized states such as generation, storage, allocation, occupation, reserve, replacement, circulation, write-off, and revenue collection. They cannot automatically trigger state transitions based on production conditions, carbon price fluctuations, and policy milestones, making it difficult to achieve dynamic balance management of stock, flow, and surplus. Asset circulation efficiency is low and ownership is unclear.

[0005] Carbon management is independent of production scheduling and is mostly based on ex-post statistics, lacking collaborative optimization through a two-tiered game between production and carbon assets. At the same time, it lacks automated collection and aggregation of emissions across the supply chain, making it impossible to include carbon liabilities in the overall account. It also lacks the ability to adaptively deconstruct and hedge against policy texts and carbon price fluctuations, resulting in high compliance costs, low returns on carbon assets, and delayed compliance responses. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent management system for the entire life cycle of carbon assets in industrial enterprises, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention aims to provide an intelligent management system for the entire life cycle of carbon assets in industrial enterprises, comprising: a working condition deterioration characteristic perception module, used to collect in real time equipment deterioration parameters, raw material composition fluctuation data, process sequence operation data, and production line load dynamic data of industrial production, and output a multi-dimensional working condition characteristic dataset for endogenous correction of carbon factors.

[0008] The data fusion and trusted evidence storage module includes a dynamic confidence-weighted fusion unit and a blockchain evidence storage unit. The dynamic confidence-weighted fusion unit is used to collect multi-source heterogeneous carbon emission data, dynamically calculate the confidence weight of each data source based on the reliability level, timeliness, and consistency of the data source, and generate fused carbon emission data and fused operating condition characteristic data using a weighted fusion algorithm. The blockchain evidence storage unit is used to write key carbon asset operation records into the blockchain in hash form to form an immutable full life cycle traceability ledger.

[0009] The carbon asset kernel algorithm module incorporates a dynamic carbon factor endogenous accounting unit, a carbon asset full life cycle state transition evolution unit, a production-carbon asset two-layer game collaborative optimization unit, a multi-level carbon asset virtual ownership confirmation, splitting and transfer unit, a policy semantic deconstruction and carbon asset adaptive hedging deduction unit, as well as a multi-objective dynamic optimization unit and a carbon price prediction unit. Based on the fused carbon emission data and fused operating condition characteristic data, the carbon asset kernel algorithm module completes the core algorithm-driven processes of carbon accounting correction, asset state iteration, production carbon collaborative scheduling, asset ownership confirmation, carbon price prediction, and market policy hedging decisions.

[0010] The carbon asset ownership transfer platform module is coupled and linked with the carbon asset kernel algorithm module to realize standardized carbon asset status management, multi-level share splitting, ownership registration, on-chain clearing, and dynamic ledger evolution, thereby constructing a standardized asset system for carbon assets that is quantifiable, verifiable, transferable, and verifiable.

[0011] The full lifecycle business adaptation module, based on the mechanism output of the carbon asset kernel algorithm module, implements closed-loop business management of the entire process of carbon asset generation, storage, allocation, occupation, reserve, pledge, replacement, transfer, write-off, residual value recovery and revenue collection.

[0012] The multi-entity collaboration and production closed-loop module is used to connect with industrial production systems, park management and control platforms, carbon trading markets, government regulatory agencies and third-party verification agencies to achieve cross-entity carbon data collaboration, ownership transfer interaction and compliance reporting.

[0013] As a further improvement to this technical solution, the operating logic of the dynamic carbon factor endogenous accounting unit includes: pre-constructing an industrial multi-process carbon emission coupling transmission matrix and quantifying the carbon emission correlation influence coefficient between each production process.

[0014] Real-time equipment degradation is calculated based on equipment runtime, loss parameters, and energy efficiency decay rate; raw material component volatility is calculated based on raw material component detection data.

[0015] Using the equipment degradation degree, raw material composition fluctuation rate, and production line real-time load rate as correction variables, the baseline emission factor is dynamically weighted and corrected to generate a real-time endogenous dynamic emission factor.

[0016] By combining process production sequence data and capacity data, the carbon emissions of each smallest production unit are allocated level by level, and the ownership information of the production unit is linked simultaneously to complete the refined carbon asset registration.

[0017] As a further improvement to this technical solution, the standardized existence states defined by the carbon asset life cycle state transition evolution unit include nine states: generation state, storage state, allocation state, occupation state, reserve state, replacement state, circulation state, write-off state, and residual value income collection state.

[0018] The automatic state transition rules include: production conditions triggering the switch between the generation state and the occupancy state; compliance cycle nodes triggering the switch from the occupancy state to the write-off state; carbon prices exceeding a preset threshold triggering the switch between the reserve state and the circulation state; policy and rule updates triggering the switch between residual value recovery and re-confirmation states; carbon asset storage or purchase triggering the storage state; and internal allocation contract execution triggering the allocation state.

[0019] As a further improvement to this technical solution, the multi-objective dynamic optimization unit and the production-carbon asset dual-layer game collaborative optimization unit are integrated into a three-layer nested optimization architecture: The upper layer is the carbon asset return optimization layer, which aims to maximize the preservation of carbon assets, minimize compliance risks, and maximize carbon trading revenue. It generates carbon asset scheduling strategies based on the carbon price range output by the carbon price prediction unit and the carbon asset value assessment results.

[0020] The middle layer is the production-carbon game coordination layer, which aims to minimize production energy consumption, process costs, and capacity utilization. By introducing carbon quota constraint factors, carbon price fluctuation constraint factors, and production capacity constraint factors, a two-layer game equilibrium equation is constructed, and the optimal production schedule, energy load allocation, and carbon asset hedging scheme are iteratively solved.

[0021] The lower layer is the in-process closed-loop control layer, which sends the production parameter adjustment instructions solved by the middle layer to the industrial control system through the production collaborative execution interface, and receives the feedback real-time carbon emission data. When the deviation between the feedback data and the expected carbon emission exceeds a preset threshold, secondary optimization is triggered and the correction instructions are reissued.

[0022] As a further improvement to this technical solution, the operation process of the policy semantic deconstruction and carbon asset adaptive hedging simulation unit includes: intelligent semantic deconstruction of the real-time updated dual-carbon policy text, automatically identifying accounting boundaries, quota allocation rules, compliance cycles, and compliance verification standards, and structuring and storing them in the database; based on the carbon price prediction unit, the short-term and medium-term carbon market price fluctuation trends are simulated to predict the carbon asset quota gap and surplus scale of enterprises; and combined with policy compliance constraints and carbon price simulation results, a combination of hedging schemes including emission reduction technology upgrades, CCER replacement, carbon sink reserves, and carbon quota trading is automatically matched to achieve carbon asset risk avoidance and return maximization.

[0023] As a further improvement to this technical solution, the carbon price prediction unit adopts an LSTM-Transformer hybrid neural network combined with quantile regression. The input variables include the historical price series of the carbon market, the energy price index, the year-on-year growth rate of industrial added value, the paid allocation ratio of carbon quotas, and the change rate of total carbon market holdings. The output is a carbon price prediction range with confidence intervals.

[0024] As a further improvement to this technical solution, the carbon asset ownership transfer platform module is also configured to: automatically switch the carbon asset status from the transfer state to the mandatory reserve state when a liquidity shortage or abnormal price fluctuation is detected in the carbon market, and lock the operation permission of the carbon asset ownership transfer platform module until the abnormal state is resolved.

[0025] As a further improvement to this technical solution, the multi-entity collaboration and production closed-loop module also includes a scope-three collaborative traceability unit, which is used to drive supply chain enterprises to upload verified carbon footprint data through blockchain smart contracts.

[0026] Uploaded data is automatically verified and collected based on a standardized carbon footprint accounting protocol.

[0027] A range three emissions inventory is compiled and generated, and the carbon liabilities corresponding to range three emissions are included in the carbon asset master account management.

[0028] As a further improvement to this technical solution, the multi-entity collaboration and production closed-loop module includes a production collaboration execution interface, which is used to send production parameter adjustment instructions to the industrial control system, receive real-time carbon emission data after adjustment from the industrial control system, and send the data back to the carbon asset kernel algorithm module to form a closed-loop correction loop for carbon intensity control.

[0029] As a further improvement to this technical solution, the production parameter adjustment instructions issued by the production collaborative execution interface include at least one of the following: boiler fuel ratio adjustment value, electrolytic cell current density setting value, cement kiln alternative fuel ratio, steel rolling heating furnace air-fuel ratio correction value, or green electricity consumption ratio target value.

[0030] Compared with the prior art, the beneficial effects of the present invention are as follows: By combining the perception of deterioration characteristics of operating conditions with dynamic confidence weighted fusion, and integrating the dynamic carbon factor endogenous correction model, the fixed factors of the national standard are abandoned, and correction variables such as equipment deterioration degree, raw material volatility, and process coupling degree are generated in real time. This enables the accurate allocation and ownership binding of carbon emissions at the five smallest units: branch plant, production line, process, single equipment, and work group, providing a physical data foundation for the credible traceability of carbon assets.

[0031] The system defines nine standardized existence states, which are automatically triggered by operating conditions, carbon prices, and policies, and the inventory, flow, and surplus ledgers are updated in real time. Based on the consortium blockchain, the multi-level virtual ownership confirmation and splitting transfer model supports dynamic splitting, internal transfer, and cross-entity on-chain settlement at the group, branch, and production line levels. Combined with blockchain notarization, ownership is solidified, significantly improving transfer efficiency and transparency.

[0032] Through a three-layer nested optimization architecture (upper layer carbon revenue optimization, middle layer production-carbon game coordination, and lower layer in-process closed-loop control), a two-layer game equilibrium equation is established to solve for the optimal production schedule and carbon asset hedging scheme. Production parameter adjustment instructions are then sent to the industrial control system to form feedback correction. The LSTM-Transformer carbon price prediction model and policy semantic deconstruction engine are integrated to automatically generate combined hedging schemes such as emission reduction technology upgrades, CCER replacement, and carbon quota trading, which significantly reduces compliance costs and increases carbon asset returns. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 This is a schematic diagram of the system structure connection of the present invention. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] Example: Please refer to Figure 1As shown, an intelligent management system for the entire life cycle of carbon assets in industrial enterprises is provided, including: a working condition deterioration characteristic perception module, which is used to collect in real time equipment deterioration parameters, raw material composition fluctuation data, process sequence operation data, and production line load dynamic data of industrial production, and output a multi-dimensional working condition characteristic dataset for endogenous correction of carbon factors.

[0037] The equipment energy efficiency decay parameters include equipment running time, loss parameters and energy efficiency decay rate; the raw material carbon content fluctuation parameters include real-time detection data of raw material composition and its carbon content fluctuation rate; the process cross-coupling energy consumption parameters include time-series correlation data between each production process and carbon emission mutual influence coefficient; and the production line variable load dynamic parameters include real-time production line load rate and load change curve. "Endogenous correction of carbon factors" is an original concept that differs from traditional fixed emission factor accounting methods. Its core meaning is as follows: Traditional methods use fixed emission factors published by the state or industry (such as how many tons of CO2 are emitted per ton of coal burned) to calculate carbon emissions. This method assumes that emission factors are stable and unchanging. However, in actual production, factors such as equipment aging, fluctuations in raw material composition, and load changes can significantly affect the actual emissions per unit of activity, leading to a systematic deviation between the calculated results and actual emissions.

[0038] The meaning of "endogenous modification": Endogenous: This refers to the correction being based on factors that do not rely on external static parameter tables, but rather are dynamically generated from the real-time operating status of the production system itself. Specifically, it utilizes equipment degradation parameters, raw material composition fluctuation data, process sequence data, and production line load dynamic data collected by the operating condition degradation feature sensing module as input variables for the emission correction factor.

[0039] Correction: Based on the baseline emission factor (such as theoretical design value or initial calibration value), the emission factor is dynamically weighted and adjusted according to real-time operating parameters such as equipment deterioration degree, raw material composition volatility, and load rate through the built-in dynamic carbon factor endogenous accounting unit, so as to generate a "real-time endogenous dynamic emission factor".

[0040] The multi-dimensional operating condition feature dataset for endogenous carbon factor correction is generated by the operating condition deterioration feature perception module through real-time collection of equipment energy efficiency degradation parameters (such as running time and loss rate), raw material carbon content fluctuation parameters (such as component detection values), process cross-coupling energy consumption parameters (such as time-series correlation data), and production line variable load dynamic parameters (such as load rate change curves) in industrial production. After the dynamic confidence weighted fusion unit performs reliability weighted fusion on the multi-source heterogeneous data, a highly reliable dataset containing feature vectors such as equipment deterioration degree, raw material component fluctuation rate, process coupling correlation degree, and real-time production line load rate is generated. This dataset is directly used as the key input of the dynamic carbon factor endogenous accounting unit for real-time correction of emission factors.

[0041] The data fusion and trusted evidence storage module includes a dynamic confidence-weighted fusion unit and a blockchain evidence storage unit. The dynamic confidence-weighted fusion unit is used to collect multi-source heterogeneous carbon emission data, dynamically calculate the confidence weight of each data source based on the reliability level, timeliness, and consistency of the data source, and generate fused carbon emission data and fused operating condition characteristic data using a weighted fusion algorithm. The blockchain evidence storage unit is used to write key carbon asset operation records into the blockchain in hash form to form an immutable full life cycle traceability ledger.

[0042] The implementation of the dynamic confidence-weighted fusion unit is as follows: First, the dynamic confidence-weighted fusion unit collects heterogeneous data from multiple sources, including IoT sensors, DCS control systems, manual reporting terminals, and satellite remote sensing monitoring, in parallel through the data access layer. For each data source, the system dynamically calculates the real-time confidence weight of each data source based on a preset reliability level base (e.g., 0.9 for sensors and 0.6 for manual reporting), a timeliness decay factor calculated from the difference between the data timestamp and the current system time, and a consistency coefficient characterized by the degree of deviation between the data source and other similar data sources at the same time. Subsequently, a weighted fusion algorithm (e.g., weighted arithmetic mean or adaptive fusion based on Bayesian estimation) is used to fuse the multi-source observations of the same physical quantity, outputting highly reliable fused carbon emission data and fused operating condition characteristic data. The original values, weights, and fusion results of each data source are recorded together for traceability.

[0043] The implementation of the blockchain-based evidence storage unit: Each time a key operation of a carbon asset occurs (including carbon asset registration, internal allocation contract signing, ownership transfer, transaction execution, performance verification, and residual asset cancellation and transfer), the blockchain-based evidence storage unit automatically calls a hash algorithm to generate a unique hash digest for the core elements of the operation event (operation type, timestamp, asset ID, quantity, participant digital signatures, and the hash value of the previous evidence storage record). This hash digest and necessary metadata are then packaged into an evidence storage transaction and submitted to multiple nodes in a pre-built consortium blockchain network for verification and confirmation through a consensus mechanism (such as a practical Byzantine fault-tolerant algorithm). Once consensus is reached, the transaction is written into a timestamped block and linked to previous blocks in chronological order, forming an immutable, fully traceable, and complete lifecycle ledger. Any subsequent query or audit of this ledger can verify the originality and integrity of the records through hash comparison.

[0044] The carbon asset kernel algorithm module incorporates a dynamic carbon factor endogenous accounting unit, a carbon asset full life cycle state transition evolution unit, a production-carbon asset two-layer game collaborative optimization unit, a multi-level carbon asset virtual ownership confirmation, splitting and transfer unit, a policy semantic deconstruction and carbon asset adaptive hedging deduction unit, as well as a multi-objective dynamic optimization unit and a carbon price prediction unit. Based on the fused carbon emission data and fused operating condition characteristic data, the carbon asset kernel algorithm module completes the core algorithm-driven processes of carbon accounting correction, asset state iteration, production carbon collaborative scheduling, asset ownership confirmation, carbon price prediction, and market policy hedging decisions.

[0045] The multi-level carbon asset virtual ownership confirmation, splitting, and transfer unit is based on a lightweight alliance chain to build a non-certificate ownership control mechanism; it supports dynamic splitting, internal transfer, and ownership change registration of multi-level carbon asset shares at the enterprise group level, branch plant level, and production line level; it has built-in cross-entity carbon asset transfer and settlement rules to realize on-chain ownership confirmation and automatic settlement of carbon asset pledge, replacement, performance compensation, and repurchase across enterprises and parks, and complete the technical solidification and reliable transfer of carbon asset ownership.

[0046] In one specific embodiment, the operating logic of the dynamic carbon factor endogenous accounting unit includes: pre-constructing an industrial multi-process carbon emission coupling transmission matrix and quantifying the carbon emission correlation influence coefficient between each production process.

[0047] When pre-constructing the carbon emission coupling transmission matrix for multiple industrial processes, the production process is first decomposed into several key processes (such as sintering, blast furnace ironmaking, converter steelmaking, and rolling). For each pair of upstream and downstream or parallel processes, correlation analysis of historical operating data and carbon emission monitoring data is conducted, and material and energy flow coupling simulation is performed (such as establishing a set of partial differential equations based on material flow balance and energy conservation). The sensitivity coefficient of upstream process carbon emission changes to downstream process emissions is calculated by combining the process mechanism model. Based on this, the sensitivity coefficients between all processes are organized into a square matrix, and the time delay factor and buffer effect correction term between processes are also included. Finally, a coupling transmission matrix that can characterize the transmission, superposition, and amplification of carbon emissions in the process network is formed, and each element in the matrix is ​​the carbon emission correlation influence coefficient between the corresponding processes.

[0048] First, the industrial production process is decomposed into several key processes (such as sintering, ironmaking, steelmaking, rolling, etc.) according to the process boundary. For each pair of process combinations, a correlation influence coefficient is defined. , indicating process The process caused by a change of 1 unit in carbon emission intensity The change in carbon emission intensity. The calculation formula is: in, Indicate process To the The quality distribution coefficient of intermediate material flow output (tons of material / tons of finished product). For the first Carbon emission factor of various material flows (tons) / ton of material), For the first Material flow input to process Consumption coefficient (tons of material input / tons of process j output). Indicates the number of the material flow. This indicates the total number of material flows. Indicate process To the Energy distribution coefficient (GJ / ton of finished product) of intermediate energy flow output. Indicates the first Carbon emission factor of intermediate energy flow (tons) ), Indicates the first Intermediate energy flow input to process Utilization coefficient (GJ input / ton of process j output). This represents the residual correction term obtained through regression analysis based on historical operating data, used to capture nonlinear coupling and time delay effects; The number representing the energy flow. This represents the total amount of energy flowing.

[0049] Real-time equipment degradation is calculated based on equipment runtime, loss parameters, and energy efficiency decay rate; raw material component volatility is calculated based on raw material component detection data.

[0050] Assume the equipment's design life is (Unit: hours or years), current cumulative runtime is Equipment loss parameters (Dimensionless, value range 0~1, reflecting maintenance level and wear accumulation), Energy efficiency degradation rate (Unit: % / 1000h, i.e., the percentage decrease in energy efficiency per thousand hours divided by 100), defining the real-time equipment degradation degree. ,for: in, , , This is a weighting coefficient, which can be determined based on historical fault data according to the equipment type (the default value is [value]). , , ), Ensure that the upper limit of degradation is 1 (complete failure state). The first term represents the basic aging related to lifespan, the second term represents the impact of maintenance and wear, and the third term represents the contribution of energy efficiency degradation to degradation.

[0051] Suppose a certain batch of raw materials was tested. Each sample is tested. Several key components (such as carbon content, sulfur content, volatile matter, etc.). For the first The component, the first The detection value of each sample Define the volatility of raw material components for: in, For the first The sample mean of each component For the first The weighting coefficients of each component Use very small positive numbers to avoid having a denominator of zero.

[0052] Using the equipment degradation degree, raw material composition fluctuation rate, and production line real-time load rate as correction variables, the baseline emission factor is dynamically weighted and corrected to generate a real-time endogenous dynamic emission factor.

[0053] The specific methods for generating real-time endogenous dynamic emission factors by the dynamic carbon factor endogenous accounting unit include: Obtaining baseline emission factors The equipment degradation degree calculated in real time Raw material component volatility and production line real-time load rate As three correction variables, they are respectively converted into corresponding correction coefficients through a preset nonlinear mapping function. , , The degree of equipment degradation Through a monotonically increasing function Mapped to The volatility of the raw material components Through the positive correlation function Mapped to The real-time load rate of the production line Through concave functions Mapped to ,and The minimum value is taken within the preset optimal load range, and it increases monotonically at both ends of the range. The baseline emission factor is calculated using a weighted product method or a weighted summation method. Dynamic correction is performed; when using a weighted product method, the correction formula is as follows: When using a weighted summation method, the corrected formula is as follows: ,in , , This represents the increment of each correction factor. , , Let be the weighting coefficient, satisfying Furthermore, it can be calibrated based on process characteristics or historical regression data; ultimately, it outputs a real-time endogenous dynamic emission factor adapted to the current operating conditions. .

[0054] By combining process production sequence data and capacity data, the carbon emissions of each smallest production unit are allocated level by level, and the ownership information of the production unit is linked simultaneously to complete the refined carbon asset registration.

[0055] In one specific embodiment, the standardized existence states defined by the carbon asset life cycle state transition evolution unit include nine states: generation state, storage state, allocation state, occupation state, reserve state, replacement state, circulation state, write-off state, and residual value income collection state.

[0056] The automatic state transition rules include: production conditions triggering the switch between the generation state and the occupancy state; compliance cycle nodes triggering the switch from the occupancy state to the write-off state; carbon prices exceeding a preset threshold triggering the switch between the reserve state and the circulation state; policy and rule updates triggering the switch between residual value recovery and re-confirmation states; carbon asset storage or purchase triggering the storage state; and internal allocation contract execution triggering the allocation state.

[0057] In one specific embodiment, the multi-objective dynamic optimization unit and the production-carbon asset two-layer game collaborative optimization unit are integrated into a three-layer nested optimization architecture: The upper layer is the carbon asset return optimization layer, which aims to maximize the preservation of carbon assets, minimize compliance risks, and maximize carbon trading revenue. It generates carbon asset scheduling strategies based on the carbon price range output by the carbon price prediction unit and the carbon asset value assessment results.

[0058] in, Indicates a time period index (such as day or week). Indicates the number of decision-making periods within the performance cycle. Let be the predicted selling price of carbon at time t. The upper quantile (e.g., the 80th quantile) of the carbon price prediction interval is used for a conservative estimate. Let t be the predicted carbon price purchase price. The lower quantile (e.g., the 20th percentile) of the predicted carbon price range is used to reduce carbon purchase costs. Let be the sales volume at time t. Let t be the purchase volume at time t. The carbon asset holding value coefficient (yuan / ton) represents the expected return on future appreciation of reserve carbon assets and can be estimated using forward curves or option pricing models. This represents the amount of carbon assets held at time t. This represents the risk aversion coefficient (dimensionless, >0), used to balance returns and performance risk. Indicates the discount factor (0 < ≤1), reflecting the time value of money, This represents the expected compliance penalty risk. The specific calculation method for the expected compliance penalty risk is as follows: First, based on production plans and operating condition forecasts, the cumulative sum of actual carbon emissions for each period during the entire compliance cycle is estimated. At the same time, the cumulative sum of the initial carbon asset holdings, the planned net purchase quotas (purchases minus sales) for the entire cycle, and the carbon emission reductions achieved through production adjustments (such as fuel switching, load optimization, etc.) is calculated. The cumulative emissions are subtracted from the cumulative available carbon asset holdings to obtain the quota gap (if the difference is positive, a gap exists; otherwise, the gap is zero). Since future carbon emissions, carbon price fluctuations, and emission reduction effects are all uncertain, the system generates multiple sets of possible emission and emission reduction results through Monte Carlo simulation or scenario analysis based on probability distribution. The penalty cost (gap multiplied by unit penalty price) is calculated for each scenario. Finally, the penalty costs for each scenario are weighted by probability to obtain the expected value of the expected compliance penalty risk.

[0059] Based on the carbon price range (including lower limit, upper limit, and median quantiles) output by the carbon price prediction unit and the carbon asset value assessment results (including holding value, trading value, and collateral value), the system first uses the carbon price range as an uncertainty input and constructs multiple carbon price evolution scenarios through random scenario generation or robust optimization methods. Then, for each scenario, combined with the current carbon asset status (such as holding quantity, remaining validity period, liquidity score, etc.), the system uses dynamic programming, model predictive control, or reinforcement learning algorithms to solve for the optimal scheduling strategy with the goal of maximizing expected return or risk-adjusted return. Finally, the system outputs specific carbon asset scheduling strategies, including but not limited to: buying to increase holdings when the carbon price is below the predicted lower limit, selling to reduce holdings when the carbon price is above the predicted upper limit, choosing to hold or pledge for financing when the carbon price is within the range, and deciding whether to use carbon assets as green credit collateral based on the collateral value assessment results, thereby achieving coordinated scheduling that maximizes carbon asset returns and minimizes risks.

[0060] The middle layer is the production-carbon game coordination layer, which aims to minimize production energy consumption, process costs, and capacity utilization. By introducing carbon quota constraint factors, carbon price fluctuation constraint factors, and production capacity constraint factors, a two-layer game equilibrium equation is constructed, and the optimal production schedule, energy load allocation, and carbon asset hedging scheme are iteratively solved.

[0061] in, Indicates the first The load factor of each production unit (dimensionless, between 0 and 1). Indicates the first Energy consumption adjustment factors for each production unit (such as fuel ratio, current density, etc., dimensionless). Indicates the first Carbon asset hedging amount obtained by each production unit Indicates energy prices, Indicates the first Baseline energy consumption (GJ / ton of product) for each production unit. Indicates the first Maximum capacity of each production unit (tons of product / period). Indicates the first Fixed process cost per production unit (RMB / ton of product). Constraints Carbon quota constraint factor: Total emissions ≤ Available carbon assets + Hedge amount.

[0062] Carbon asset hedging total limit: the hedging amount shall not exceed the reserve amount.

[0063] Carbon price volatility constraint factor: Remaining reserves must cover the risk of carbon price volatility.

[0064] Production capacity constraint factors: Load rate and energy consumption adjustment factor are within the safe range.

[0065] Production demand constraint: Total production must meet orders.

[0066] Carbon quota constraint factor: Based on the total available quota in the current carbon asset account (initial quota plus net purchased quota minus occupied quota), combined with the real-time emission prediction of the production process by the dynamic carbon factor accounting model, the system uses the total available quota as the emission upper limit on the right side of the inequality, thus forming the carbon quota constraint factor to ensure that the cumulative carbon emissions of the production unit do not exceed this upper limit in any period.

[0067] Carbon price volatility constraint factor: The system calculates a risk buffer multiplier based on the carbon price volatility (such as historical volatility or implied volatility) output by the carbon price prediction unit and the risk tolerance coefficient preset by the enterprise. This multiplier is then multiplied by the square root (or standard deviation) of the expected carbon emissions to obtain the minimum carbon asset reserve that must be maintained. This constitutes the carbon price volatility constraint factor, requiring that the remaining carbon reserves after hedging are not lower than this value to resist adverse carbon price fluctuations.

[0068] Production capacity constraint factors: Based on the rated power of equipment, process safety boundaries, maintenance cycles, and historical operating data of each production unit, the system calibrates the minimum and maximum allowable values ​​of load rate, as well as the feasible range of energy consumption adjustment factors (such as fuel ratio, current density, etc.). These boundary parameters are directly used as the upper and lower limits of decision variables, thus forming the production capacity constraint factors. The lower layer is the in-process closed-loop control layer, which sends the production parameter adjustment instructions solved by the middle layer to the industrial control system through the production collaborative execution interface, and receives the feedback real-time carbon emission data. When the deviation between the feedback data and the expected carbon emission exceeds a preset threshold, secondary optimization is triggered and the correction instructions are reissued.

[0069] In one specific embodiment, the operation of the policy semantic deconstruction and carbon asset adaptive hedging simulation unit includes: intelligently deconstructing the real-time updated dual-carbon policy text, automatically identifying accounting boundaries, quota allocation rules, compliance cycles, and compliance verification standards, and storing them in a structured database; based on the carbon price prediction unit, simulating short-term and medium-term carbon market price fluctuation trends, and predicting the carbon asset quota gap and surplus scale of enterprises; and combining policy compliance constraints and carbon price simulation results to automatically match a combination of hedging schemes including emission reduction technology upgrades, CCER replacement, carbon sink reserves, and carbon quota trading, thereby achieving carbon asset risk avoidance and return maximization.

[0070] In one specific embodiment, the carbon price prediction unit uses an LSTM-Transformer hybrid neural network combined with quantile regression. The input variables include historical carbon market price series, energy price index, year-on-year growth rate of industrial added value, paid allocation ratio of carbon quotas, and change rate of total carbon market holdings. The output is a carbon price prediction range with confidence intervals.

[0071] In one specific embodiment, the carbon asset ownership transfer platform module is further configured to: automatically switch the carbon asset status from the transfer state to the mandatory reserve state when a liquidity shortage or abnormal price fluctuation is detected in the carbon market, and lock the operation permissions of the carbon asset ownership transfer platform module until the abnormal state is resolved.

[0072] The carbon asset ownership transfer platform module is coupled and linked with the carbon asset kernel algorithm module to realize standardized carbon asset status management, multi-level share splitting, ownership registration, on-chain clearing, and dynamic ledger evolution, thereby constructing a standardized asset system for carbon assets that is quantifiable, verifiable, transferable, and verifiable.

[0073] The carbon asset ownership transfer platform module constructs a standardized asset system in the following ways: First, based on the nine standardized states of the carbon asset's entire life cycle (generation, storage, allocation, occupation, reserve, replacement, transfer, write-off, and residual value collection), a unified state machine model is established for each carbon asset to achieve standardized management of asset states; Second, multi-level share splitting technology is used to dynamically break down and virtually confirm the ownership of enterprise group-level carbon assets according to organizational structure (branch plant, production line, work group), and a unique identifier is generated for each minimum share. First, ownership identification enables the quantification and confirmation of carbon assets. Second, key operations such as ownership registration, share transfer, and transaction settlement are deployed on the consortium blockchain in the form of smart contracts, automatically completing on-chain clearing across entities and levels, ensuring that the transfer process is open, transparent, and tamper-proof. Finally, a dynamic ledger evolution engine synchronizes events such as the generation, occupation, replacement, and write-off of carbon assets in real time, automatically updating the holdings, frozen amounts, and available amounts of each entity, forming an asset ledger covering the entire life cycle, thereby ultimately realizing the transferability and write-offability of carbon assets.

[0074] The full lifecycle business adaptation module, based on the mechanism output of the carbon asset kernel algorithm module, implements closed-loop business management of the entire process of carbon asset generation, storage, allocation, occupation, reserve, pledge, replacement, transfer, write-off, residual value recovery and revenue collection.

[0075] The multi-entity collaboration and production closed-loop module is used to connect with industrial production systems, park management and control platforms, carbon trading markets, government regulatory agencies and third-party verification agencies to achieve cross-entity carbon data collaboration, ownership transfer interaction and compliance reporting.

[0076] The carbon asset life cycle state transition and evolution unit defines multiple standardized existence states of carbon assets. By preset operating condition trigger thresholds, policy compliance nodes, and carbon price fluctuation thresholds, it realizes automatic transition and iteration of various states of carbon assets, and constructs a life cycle evolution mechanism of dynamic balance between stock, flow, and surplus.

[0077] In one specific embodiment, the multi-entity collaboration and production closed-loop module further includes a scope-three collaborative traceability unit, which is used to drive supply chain enterprises to upload verified carbon footprint data through blockchain smart contracts.

[0078] The specific implementation method is as follows: The system deploys a smart contract on the consortium blockchain that includes data upload, format verification, and signature verification logic, and assigns a unique digital identity to each supply chain enterprise; after the enterprise completes the product carbon footprint calculation, it calls the upload interface of the smart contract to submit a carbon footprint report (including emission source data, activity level, and emission factor) verified by a third party. The smart contract automatically verifies the validity of the data format, timestamp, and digital signature, and stores the data hash on the chain. At the same time, the original data is encrypted and distributed to authorized regulatory nodes to ensure that the data source is credible and non-repudiable.

[0079] Uploaded data is automatically verified and collected based on a standardized carbon footprint accounting protocol.

[0080] The specific implementation method is as follows: The system has a built-in standardized accounting protocol template that conforms to international standards and industry-specific rules (such as accounting guidelines for industries such as steel and cement); when the uploaded data is received, the system first automatically compares whether key fields such as accounting boundaries, emission factor sources, and activity data collection methods are consistent with the protocol template through the rule engine; secondly, it uses cross-validation algorithms (such as material flow balance and energy balance) to check the data self-consistency; finally, the carbon emission data that has passed the verification is automatically collected into a unified data pool according to the enterprise number, product category and time dimension, and the verification status is marked (passed / pending review / failed).

[0081] A range three emissions inventory is compiled and generated, and the carbon liabilities corresponding to range three emissions are included in the carbon asset master account management.

[0082] The specific implementation method is as follows: The system accumulates the collected carbon footprint data of each enterprise according to the upstream and downstream relationships of the supply chain (such as raw material procurement, logistics and transportation, product use and waste disposal, etc.) to generate a scope three emission list covering the entire supply chain (displaying emissions and proportions by category); then, based on the enterprise's preset internal carbon pricing or external carbon market price, the scope three emissions are converted into carbon liabilities (i.e., additional carbon allowances or emission reductions that need to be purchased), and the liability is recorded in the "liabilities" sub-ledger of the enterprise's total carbon asset account through the carbon asset ownership transfer platform module, and managed in conjunction with the enterprise's own scope one and scope two emissions, so as to simultaneously consider the carbon cost of the supply chain in compliance scheduling and strategy optimization.

[0083] In one specific embodiment, the multi-entity collaboration and production closed-loop module includes a production collaboration execution interface, which is used to send production parameter adjustment instructions to the industrial control system, receive real-time carbon emission data after adjustment from the industrial control system, and send the data back to the carbon asset kernel algorithm module to form a closed-loop correction loop for carbon intensity control.

[0084] In one specific embodiment, the production parameter adjustment instructions issued by the production collaborative execution interface include at least one of the following: boiler fuel ratio adjustment value, electrolyzer current density setting value, cement kiln alternative fuel ratio, steel rolling heating furnace air-fuel ratio correction value, or green electricity consumption ratio target value.

[0085] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention.

Claims

1. An industrial enterprise carbon asset life cycle intelligent management system, characterized in that, include: The working condition deterioration feature perception module is used to collect equipment deterioration parameters, raw material composition fluctuation data, process sequence operation data, and production line load dynamic data in real time in industrial production, and output a multi-dimensional working condition feature dataset for endogenous correction of carbon factors. The data fusion and trusted evidence storage module includes a dynamic confidence-weighted fusion unit and a blockchain evidence storage unit; The dynamic confidence-weighted fusion unit is used to collect multi-source heterogeneous carbon emission data. Based on the reliability level, timeliness and consistency of the data source, it dynamically calculates the confidence weight of each data source and uses a weighted fusion algorithm to generate fused carbon emission data and fused operating condition characteristic data. The blockchain storage unit is used to write key operation records of carbon assets into the blockchain in hash form to form an immutable full life cycle traceability ledger. The carbon asset kernel algorithm module incorporates a dynamic carbon factor endogenous accounting unit, a carbon asset full life cycle state transition evolution unit, a production-carbon asset two-layer game collaborative optimization unit, a multi-level carbon asset virtual ownership confirmation, splitting and transfer unit, a policy semantic deconstruction and carbon asset adaptive hedging deduction unit, as well as a multi-objective dynamic optimization unit and a carbon price prediction unit. Based on the fused carbon emission data and fused operating condition characteristic data, the carbon asset kernel algorithm module completes the core algorithm-driven processes of carbon accounting correction, asset state iteration, production carbon collaborative scheduling, asset ownership confirmation, carbon price prediction, and market policy hedging decisions. The carbon asset ownership transfer platform module is coupled and linked with the carbon asset kernel algorithm module to realize standardized carbon asset status management, multi-level share splitting, ownership registration, on-chain clearing, dynamic ledger evolution, and to build a standardized carbon asset system. The full lifecycle business adaptation module, based on the mechanism output of the carbon asset kernel algorithm module, implements closed-loop business management of the entire carbon asset process; The multi-entity collaboration and production closed-loop module is used to connect with industrial production systems, park management and control platforms, carbon trading markets, government regulatory agencies and third-party verification agencies to achieve cross-entity carbon data collaboration, ownership transfer interaction and compliance reporting. 2.The industrial enterprise carbon asset whole life cycle intelligent management system according to claim 1, characterized in that: The operating logic of the dynamic carbon factor endogenous accounting unit includes: A pre-constructed carbon emission coupling transmission matrix for multiple industrial processes was used to quantify the carbon emission correlation coefficients between various production processes. Real-time equipment degradation is calculated based on equipment runtime, loss parameters, and energy efficiency decay rate; raw material component volatility is calculated based on raw material component detection data. The equipment degradation degree, raw material composition fluctuation rate and production line real-time load rate are used as correction variables to dynamically weight and correct the baseline emission factor, thereby generating a real-time endogenous dynamic emission factor. By combining process production sequence data and capacity data, the carbon emissions of each smallest production unit are allocated level by level, and the ownership information of the production unit is linked simultaneously to complete the refined carbon asset registration. 3.The industrial enterprise carbon asset whole life cycle intelligent management system according to claim 1, characterized in that: The standardized existence states defined by the carbon asset life cycle state transition and evolution unit include nine states: generation state, storage state, allocation state, occupation state, reserve state, replacement state, circulation state, write-off state, and residual value income collection state. The automatic state transition rules include: production conditions triggering the switch between the generation state and the occupancy state; compliance cycle nodes triggering the switch from the occupancy state to the write-off state; carbon prices exceeding a preset threshold triggering the switch between the reserve state and the circulation state; policy and rule updates triggering the switch between residual value recovery and re-confirmation states; carbon asset storage or purchase triggering the storage state; and internal allocation contract execution triggering the allocation state. 4.The industrial enterprise carbon asset whole life cycle intelligent management system according to claim 1, characterized in that: The multi-objective dynamic optimization unit and the production-carbon asset dual-layer game collaborative optimization unit are integrated into a three-layer nested optimization architecture: The upper layer is the carbon asset return optimization layer, which aims to maximize the preservation of carbon assets, minimize compliance risks, and maximize carbon trading revenue. It generates a carbon asset scheduling strategy based on the carbon price range output by the carbon price prediction unit and the carbon asset value assessment results. The middle layer is the production-carbon game coordination layer, with the optimization objectives of minimizing production energy consumption, minimizing process costs, and maximizing capacity utilization. By introducing carbon quota constraint factors, carbon price fluctuation constraint factors, and production capacity constraint factors, a two-layer game equilibrium equation is constructed, and the optimal production schedule, energy load allocation, and carbon asset hedging scheme are iteratively solved. The lower layer is the in-process closed-loop control layer, which sends the production parameter adjustment instructions solved by the middle layer to the industrial control system and receives the feedback real-time carbon emission data. When the deviation between the feedback data and the expected carbon emission exceeds a preset threshold, secondary optimization is triggered and the correction instructions are reissued.

5. The industrial enterprise carbon asset lifecycle intelligent management system according to claim 1, characterized in that: The operation process of the policy semantic deconstruction and carbon asset adaptive hedging simulation unit includes: The system performs intelligent semantic deconstruction on real-time updated dual-carbon policy texts, automatically identifies accounting boundaries, quota allocation rules, compliance cycles, and compliance verification standards, and stores them in a structured database. Based on the carbon price prediction unit, it extrapolates short- and medium-term carbon market price fluctuation trends and predicts the carbon asset quota gap and surplus scale of enterprises. Combining policy compliance constraints and carbon price extrapolation results, it automatically matches a combination of hedging schemes, including emission reduction technology upgrades, CCER replacement, carbon sink reserves, and carbon quota trading, to achieve carbon asset risk avoidance and maximum returns.

6. The industrial enterprise carbon asset lifecycle intelligent management system according to claim 1, characterized in that: The carbon price prediction unit uses an LSTM-Transformer hybrid neural network combined with quantile regression. The input variables include historical carbon market price series, energy price index, year-on-year growth rate of industrial added value, paid allocation ratio of carbon quotas, and change rate of total carbon market holdings. The output is a carbon price prediction range with confidence intervals.

7. The industrial enterprise carbon asset lifecycle intelligent management system according to claim 1, characterized in that: The carbon asset ownership transfer platform module is also configured to automatically switch the carbon asset status from the transfer state to the mandatory reserve state when a liquidity shortage or abnormal price fluctuation is detected in the carbon market, and lock the operation permissions of the carbon asset ownership transfer platform module until the abnormal state is resolved. 8.The industrial enterprise carbon asset whole life cycle intelligent management system according to claim 1, characterized in that: The multi-entity collaboration and production closed-loop module also includes a scope three collaborative traceability unit, which is used to drive supply chain enterprises to upload verified carbon footprint data through blockchain smart contracts; The uploaded data is automatically verified and collected based on a standardized carbon footprint accounting protocol. A comprehensive emission inventory of Scope 3 will be generated, and the carbon liabilities corresponding to Scope 3 emissions will be included in the carbon asset master account management. 9.The industrial enterprise carbon asset whole life cycle intelligent management system according to claim 1, characterized in that: The multi-entity collaboration and production closed-loop module includes a production collaboration execution interface, which is used to send production parameter adjustment instructions to the industrial control system, receive real-time carbon emission data after adjustment from the industrial control system, and send the data back to the carbon asset kernel algorithm module to form a closed-loop correction loop for carbon intensity control. 10.The industrial enterprise carbon asset whole life cycle intelligent management system according to claim 1, characterized in that: The production parameter adjustment instructions issued by the production collaboration execution interface include at least one of the following: boiler fuel ratio adjustment value, electrolytic cell current density setting value, cement kiln alternative fuel ratio, steel rolling heating furnace air-fuel ratio correction value, or green electricity consumption ratio target value.

Images