Production workshop carbon flow twin mapping method

Abstract

The invention relates to a production workshop carbon flow twinborn mapping method, and belongs to the technical field of carbon emission optimization. The method comprises the steps of collecting production workshop data in real time to perform multi-modal data fusion; constructing a carbon flow dynamic accounting mechanism model, and calculating the real-time carbon emission intensity of the process; constructing a carbon flow intensity tensor model, extracting multi-granularity carbon flow features based on the carbon flow intensity tensor model, and realizing real-time digital twinborn deduction of carbon flow propagation by using an intelligent prediction algorithm; establishing a differential equation to describe double-flow real-time interaction of the carbon flow and the value flow, constructing a carbon value incidence matrix, and performing carbon flow value analysis; and on the basis of the carbon value incidence matrix, a space-time carbon chain-oriented cooperative adjustment strategy is dynamically generated through a multi-objective optimization algorithm, a double-layer topological optimization model is constructed, and a closed-loop feedback mechanism is introduced to drive the economic sustainability of the control strategy in an industrial environment. The workshop carbon footprint can be displayed in real time, the workshop carbon emission condition can be truly reflected, and the carbon emission data can be analyzed and optimized in time.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of carbon emission optimization, and relates to a production workshop carbon emission method, in particular to a production workshop carbon flow twin mapping method. BACKGROUND

[0002] In the field of industrial manufacturing, especially in discrete production workshops, precise carbon emission monitoring and management has become a key problem to be solved. However, the existing technology has significant limitations in addressing this challenge:

[0003] Firstly, traditional carbon emission accounting methods rely on static reports or periodic inventory, which cannot adapt to the characteristics of modern production workshops such as large data volume, fast update, and multiple modalities. The lack of real-time, multi-source data fusion capability for key parameters such as energy consumption, emissions, equipment status, and material flow in the production process makes it difficult to accurately capture transient carbon flow changes. At the same time, existing methods cannot build high-precision real-time carbon emission dynamic accounting models at the process level, and cannot effectively locate the spatio-temporal distribution characteristics and specific flow paths of carbon emissions (especially direct and indirect emissions), making carbon emission tracing ambiguous and limiting the accuracy of basic data.

[0004] Secondly, the existing technology cannot effectively establish a real-time and accurate correlation model between micro-dynamic data at the equipment level and macro-carbon flow state at the process level or even the whole link level. Carbon emissions in production workshops involve multiple physical scales such as equipment, processes, and production lines, and their flow has complex correlation. However, existing methods usually process emission data for equipment or processes in isolation, and lack effective means to dynamically couple real-time collected equipment-level operating parameters (such as power and status) with accounting-derived process-level carbon flow intensity. This makes it difficult to build mathematical models that can represent the cross-scale flow rules of "equipment-process-whole link" carbon emissions, and cannot clearly show the propagation path, interaction relationship, and intensity evolution of carbon emissions at different levels. Therefore, there is a lack of dynamic digital twin mapping capability to support real-time deduction of carbon flow propagation state and prediction of future evolution trend.

[0005] Thirdly, the existing technology fails to deeply analyze the real-time dynamic correlation between the physical flow path of carbon emissions and the value creation activities in the production process. Current carbon reduction measures are often disconnected from the core economic goals of enterprises (such as cost, resource utilization rate, and unit energy consumption revenue). Existing management methods generally lack a model framework to dynamically couple real-time carbon flow intensity data with multi-dimensional economic indicators. It is difficult to quantitatively describe how the two important flow streams, "carbon flow" and "value flow", interact (cooperate or conflict) in real time, such as revealing the immediate impact of changes in carbon emissions at a specific process on cost or revenue. This fragmentation makes it difficult to make precise carbon reduction decisions and optimize resource allocation while ensuring economic efficiency.

[0006] In summary, the prior art is difficult to support precise carbon emission tracing, real-time state monitoring and collaborative optimization decision-making for complex dynamic production environments, and therefore an innovative method capable of high-precision carbon flow dynamic perception, cross-scale dynamic mapping modeling and carbon flow-value flow deep coupling analysis is urgently needed to provide core support for the refinement and intelligentization of workshop-level carbon management. SUMMARY

[0007] Therefore, the purpose of the present application is to provide a production workshop carbon flow mapping method to provide real-time and accurate mapping of carbon emissions for enterprises, help managers better understand production conditions, and at the same time help enterprises solve the problem of the fragmentation of traditional carbon reduction and economic goals, and achieve collaborative regulation of carbon reduction and gain.

[0008] To achieve the above purpose, the present application provides the following technical solutions:

[0009] A production workshop carbon flow twin mapping method, the method comprising:

[0010] Real-time acquisition of production workshop data, and fusion of the acquired multi-modal data; construction of a carbon flow dynamic accounting mechanism model to calculate real-time carbon emission intensity of a process;

[0011] Coupling of process-level anchor parameters and equipment-level dynamic data through carbon flow mapping modeling to construct a carbon flow intensity tensor model to represent the cross-scale correlation of carbon emissions; extraction of multi-granularity carbon flow features based on the carbon flow intensity tensor model to realize real-time digital twin deduction of carbon flow propagation using an intelligent prediction algorithm;

[0012] Dynamic coupling of carbon flow intensity and value flow intensity through multi-dimensional economic indicators to establish a differential equation to describe the real-time interaction of carbon flow and value flow, and to construct a carbon-value correlation matrix;

[0013] Carbon flow value flow analysis based on the carbon-value correlation matrix to construct a double-layer topological optimization model, and then dynamically generating a collaborative adjustment strategy for a time-space carbon chain through a multi-objective optimization algorithm; introduction of a closed-loop feedback mechanism to realize economic sustainability of the collaborative adjustment strategy in an industrial environment.

[0014] Further, the acquired production workshop data includes emission, energy consumption, equipment state and material flow data, and the production workshop data is acquired through sensors.

[0015] The non-time series data collected by the sensors is directly transmitted to a protocol analysis gateway for preprocessing, and the time series data collected by the sensors is transmitted to an ERP / MES system and uploaded to a time series database, and then the data is preprocessed through the protocol analysis gateway; the multi-modal data after preprocessing is fused through the protocol analysis gateway:

[0016] e j{Device ID, process number, logistics node}

[0017] where t i is the timestamp of the device, e j is the carbon emission entity identifier, v k is the original consumption.

[0018] Further, a process-level carbon flow dynamic accounting mechanism model is constructed in combination with production system process parameters, the production system process parameters including process instantaneous power and process duration, and the carbon flow dynamic accounting mechanism model is represented as:

[0019]

[0020] where a is a carbon emission factor, P k (t) is the instantaneous power of process k, φ ik (t) is the consumption of process k to energy i, e i (t) is the real-time carbon emission intensity of energy i, M jk (t) is the equivalent carbon value of raw material j in process k. Δt is the time interval, t is the time, β(t) is a time-varying adjustment factor of material carbon emission, C k (t) is the total carbon emission of process k at time t.

[0021] Further, the process-level anchor parameters are coupled with the device-level dynamic data through carbon flow mapping modeling to construct a carbon flow intensity tensor model, and the model is represented as:

[0022]

[0023] where a i (t) is the carbon emission coefficient of energy, E i is the carbon influence factor of the device, L i is the logistics path topology vector, N is the total set of carbon flow paths, τ t (t) is the time carbon flow tensor, P i is the process chain energy transfer matrix, and m is the number of processes; in the matrix P i , the row direction is the output process , the column direction is the input process C(t); the elements p i in the matrix P rs are updated by the following formula:

[0024]

[0025] where L i→k is the flow path weight of the logistics between process i and process k within time t.

[0026] Further, based on the carbon flow intensity tensor model, multi-granularity carbon flow features are extracted, and real-time digital twin deduction of carbon flow propagation is realized through an LSTM-Transformer hybrid network. The LSTM is used to capture the long-period carbon flow evolution trend, the Transformer is used to learn the multi-device carbon flow interaction, and the attention mechanism is used to calculate the carbon influence weight between devices.

[0027] Further, the carbon flow intensity and value flow intensity are dynamically coupled through multi-dimensional economic indicators, and a differential equation is established to describe the real-time interaction of carbon flow and value flow. The differential equation is represented as:

[0028]

[0029] In the formula, β k (t) is the value conversion efficiency factor of process k, U k is an external variable, is an energy coupling matrix, C k,i (t) is a dynamic coupling coefficient, λ k is an external control factor, V k is the real-time carbon flow value of the kth node.

[0030] Further, a carbon-value correlation matrix is constructed, which is represented as:

[0031]

[0032] In the formula, the row elements in the matrix are carbon emission types, and the column elements are value categories.

[0033] Further, the carbon-value correlation matrix is analyzed in terms of carbon flow and value flow to construct a double-layer topology optimization model. In the constructed double-layer topology optimization model, the upper-layer objective is to minimize the total entropy value of the carbon flow topology network:

[0034]

[0035] In the formula, C e is the carbon flow of a single path e, C total is the total carbon flow, and E is the carbon flow transmission path.

[0036] The lower-layer objective is to satisfy the process beat time constraint:

[0037]

[0038] In the formula, T k is the actual beat time of the Kth process, T spec is the maximum beat time threshold allowed by the process, and K is the set of process steps in the whole process.

[0039] The present application has the advantages of:

[0040] (1) The present application provides a production workshop carbon flow twin mapping method aiming at the characteristics of complex operation, fast data generation speed and large data volume of discrete production manufacturing workshops. Through the method, real-time and accurate mapping of carbon emissions can be provided for enterprises, helping managers better understand the production status.

[0041] (2) The present application can better help enterprises solve the problem of traditional carbon reduction and economic target fragmentation by coupling the carbon flow intensity and value flow benefit of the process chain in real time, and realize the coordinated regulation of carbon reduction and gain.

[0042] (3) The present application supports precise carbon emission tracing and intervention from local device level to whole process link level by deeply analyzing the carbon emission flow path and space-time distribution characteristics between processes, and generates flexible strategies such as carbon emission path reorganization and process priority adjustment based on dynamic industrial environment, verifies the regulation strategy, iteratively optimizes model parameters and decisions, and forms a full life cycle sustainable capability of "perception-optimization-verification-evolution".

[0043] Other advantages, objects and features of the present application will be set forth in part in the following specification, and in part will become apparent to those skilled in the art upon examination of the following specification, or can be learned by practice of the present application. The objects and other advantages of the present application can be realized and attained by the methods and instrumentalities particularly pointed out in the following description. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to make the purpose, technical scheme and advantages of the present application clearer, the preferred detailed description of the present application will be combined with the drawings as follows, wherein:

[0045] Fig. 1 The flowchart of the production workshop carbon flow twin mapping method provided by an embodiment of the present application is shown in the figure;

[0046] Fig. 2 The production workshop carbon flow twin mapping model framework diagram is shown in the figure;

[0047] Fig. 3 The carbon flow-value flow analysis model framework diagram is shown in the figure. DETAILED DESCRIPTION

[0048] The present application is illustrated by way of example and not limitation in the figures of the accompanying drawings, in which like references indicate similar elements, and in which: BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Wherein, the drawings are only used for example description, the representation is only schematic diagram, not real object drawing, and cannot be understood as the limitation of the present application; in order to better illustrate the embodiments of the present application, some components of the drawings are omitted, enlarged or reduced, and do not represent the size of the actual product; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings can be omitted.

[0050] The same or similar reference numerals in the drawings of the embodiments of the present application correspond to the same or similar components; in the description of the present application, it should be understood that if the terms 'upper', 'lower', 'left', 'right', 'front', 'back' and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, only for the convenience of describing the present application and simplifying the description, and not to indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, therefore the positional relationship described in the drawings is only used for example description, and cannot be understood as the limitation of the present application, for those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0051] Please refer to Figs. 1-3 A carbon flow twin mapping method for a production workshop is provided, the method comprising:

[0052] 1. Carbon flow dynamic data anchoring

[0053] Deploying Internet of Things terminals in the production workshop, collecting production workshop data in real time, mainly including emission, energy consumption, device state, material flow data, etc.; fusing the collected multi-modal data; constructing a dynamic accounting mechanism model of carbon flow, calculating the real-time carbon emission intensity of the process, and using twin technology to construct a process-level carbon flow dynamic atlas, positioning the spatio-temporal distribution of carbon emission, so as to track the flow path and intensity of direct and indirect carbon emission in the production system in real time.

[0054] The specific way of data acquisition is: on the one hand, the required data is collected by various sensors and transmitted to the protocol analysis gateway for data preprocessing, and on the other hand, the time series data collected by the sensor is transmitted to the ERP / MES system and uploaded to the time series database and then preprocessed by the protocol analysis gateway. The preprocessed data is fused by the protocol analysis gateway, the multi-source heterogeneous data is integrated, the model input accuracy is improved, and the dynamic update of the tensor matrix is driven:

[0055] e j ∈{device ID, process number, logistics node}

[0056] Wherein, t i is the timestamp of the device, e j is the carbon emission entity identifier, v k is the original consumption.

[0057] A process-level carbon flow dynamic accounting mechanism model is constructed combined with production system process parameters to locate the spatio-temporal distribution of carbon emissions and provide high-precision input for carbon emission accounting. The production system process parameters mainly include process instantaneous power, process duration, etc. The constructed carbon flow dynamic accounting mechanism model is as follows:

[0058]

[0059] Wherein, α is the carbon emission factor, P k (t) is the instantaneous power of process k, φ ik (t) is the consumption of process k to energy i, ε i (t) is the real-time carbon emission intensity of energy i, M jk (t) is the equivalent carbon value of raw material j in process k. Δt is the time interval, t is the time, β(t) is the time-varying adjustment factor of material carbon emission, C k (t) is the total carbon emission of process k at time t.

[0060] 2. Carbon flow dynamic data mapping

[0061] Through carbon flow mapping modeling, the process-level anchor parameters are coupled with the device-level dynamic data, a carbon flow intensity tensor model is constructed, and the device-process-full-link carbon emission mapping is used to represent the cross-scale correlation of carbon emission; based on the carbon flow intensity tensor model, multi-granularity carbon flow features are extracted to realize real-time digital twin deduction of carbon flow propagation by using intelligent prediction algorithm. The twin mapping modeling is the dynamic data bidirectional interaction between the physical workshop and the virtual model, which inputs the historical operation data and process parameters into the twin model.

[0062] Firstly, the process chain energy transfer matrix P i: Carbon flow network with equipment as nodes, process chain energy transfer matrix calculates carbon flow transfer according to time advancement, process chain energy transfer input C(t) is expressed as:

[0063]

[0064] Where, direct s is direct carbon emission of process s, Δ rs is the material transfer time from process r to s, P rs is the process matrix from process r to s, is the energy output of process r. Where, r is the process before s.

[0065] Process chain carbon flow output is expressed as:

[0066]

[0067] Where, η s is the carbon retention rate of process s.

[0068] Node carbon flow equipment level carbon emission data α i (t)·E i is weighted and transmitted to process matrix P rs :

[0069]

[0070] L i→k is the flow path weight of the material flow between process i and process k at time t, which determines the position of the equipment in the whole process chain through the path weight, and realizes carbon source tracing.

[0071] Matrix P i is expressed as:

[0072]

[0073] Where, m is the number of processes, r = 1, 2, …, m, s = 1, 2, …, m. In the matrix, the row direction is the output process and the column direction is the input process C(t).

[0074] Then, the carbon flow intensity tensor model is constructed, which is expressed as:

[0075]

[0076] Where, α i (t) is the carbon emission coefficient of energy, E i is the carbon influence factor of equipment, L i is the material flow path topology vector. N is the total set of carbon flow paths, τ t(t) is the time carbon flow tensor.

[0077] Finally, the real-time digital twin deduction of carbon flow propagation is realized by using intelligent prediction algorithms. Specifically, the real-time digital twin deduction of carbon flow propagation is realized by using LSTM-Transformer hybrid network. The long-period carbon flow evolution trend is captured by the LSTM module, the multi-device carbon flow interaction is learned by the Transformer module, and the carbon influence weight between devices is calculated by the attention mechanism.

[0078] 3. Carbon flow-value flow analysis

[0079] The physical flow path of carbon emissions is deeply associated with the value flow, the carbon flow intensity and the value flow intensity are dynamically coupled by multi-dimensional economic indicators, the real-time interaction of carbon flow-value flow is described by differential equations, the carbon-value correlation matrix is constructed, and the carbon flow value analysis is performed. The multi-dimensional economic indicators include process cost, resource utilization rate, unit energy consumption benefit, etc.

[0080] First, based on the discrete time simulation of the production workshop, the carbon flow intensity and the value flow intensity are dynamically coupled, the real-time interaction of carbon flow and value flow is described by differential equations, and the differential equations are as follows:

[0081]

[0082] where β k (t) is the value conversion efficiency factor of process k, which is related to the comprehensive efficiency of the device, U k is an external variable (such as carbon price, which reflects the influence of external market quotation on value). is the energy coupling matrix, C k,i (t) is the dynamic coupling coefficient, λ k is an external control factor, V k is the real-time carbon flow value of the kth node.

[0083] Then, the carbon-value correlation matrix is constructed, the parameters of the carbon-value correlation matrix are dynamically adjusted by reinforcement learning, and the form of the carbon-value correlation matrix is as follows:

[0084]

[0085] where the row elements in the matrix are carbon emission types (direct emission, indirect emission, and implicit carbon), and the column elements are value categories (output value and material loss).

[0086] 4. Carbon flow twin topology optimization

[0087] Based on the carbon-value correlation matrix, carbon flow value flow analysis is carried out to construct a double-layer topological optimization model. According to the optimization model, a collaborative adjustment strategy for time and space carbon chain is dynamically generated by a multi-objective optimization algorithm. A closed-loop feedback mechanism is introduced to realize the economic sustainability of the collaborative adjustment strategy in the industrial environment, and to realize the global dynamic balance of carbon emissions and value flow.

[0088] The multi-objective optimization algorithm can use deep reinforcement learning, adaptive model predictive control, etc.

[0089] The collaborative adjustment strategy mainly includes process priority reset, physical path low-carbon reorganization, etc.

[0090] The closed-loop feedback mechanism mainly includes real-time carbon efficiency-economic loss and profit verification, strategy correction, etc.

[0091] Based on the carbon flow value flow correlation analysis model, a double-layer topological optimization model is constructed. The upper-layer topological optimization inputs the process node, carbon flow characteristic matrix and value flow constraint, and establishes the upper-layer target as minimizing the total entropy value of the carbon flow topological network:

[0092]

[0093] Wherein, C e is the carbon flow rate of a single path e, C total is the total carbon flow, and E is the carbon flow transmission path.

[0094] The lower-layer parameter optimization inputs the real-time data of the equipment, the carbon emission factor and the upper-layer topological decision, and establishes the lower-layer target as satisfying the process beat time constraint:

[0095]

[0096] Wherein, T k is the actual beat time of the kth process, T spec is the maximum beat time threshold allowed for the process, and K is the collection of process steps in the whole process.

[0097] Through the double-layer topological optimization model, the upper-layer outputs the optimal topological structure and key carbon flow path, and the lower-layer outputs the optimal process parameters and dynamic control strategy.

[0098] Finally, it should be pointed out that the above examples are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the purpose and scope of the present technical solutions, which should be covered by the scope of the claims of the present application.

Claims

1. A carbon flow twin mapping method for a production workshop, characterized in that: The method includes: Collect production workshop data in real time and fuse the collected multimodal data; build a carbon flow dynamic accounting mechanism model to calculate the real-time carbon emission intensity of the process; Through carbon flow mapping modeling, process-level anchoring parameters are coupled with equipment-level dynamic data to construct a carbon flow intensity tensor model to characterize the cross-scale correlation of carbon emissions. Based on the carbon flow intensity tensor model, multi-granularity carbon flow characteristics are extracted and intelligent prediction algorithms are used to achieve real-time digital twin deduction of carbon flow propagation. By dynamically coupling carbon flow intensity and value flow intensity through multi-dimensional economic indicators, a differential equation is established to describe the real-time interaction between carbon flow and value flow, and a carbon-value correlation matrix is ​​constructed; Based on the carbon-value association matrix, the carbon flow value stream is analyzed, a two-layer topology optimization model is constructed, and then a collaborative adjustment strategy for the spatiotemporal carbon chain is dynamically generated through a multi-objective optimization algorithm; a closed-loop feedback mechanism is introduced to achieve the economic sustainability of the collaborative adjustment strategy in an industrial environment.

2. The method according to claim 1, characterized in that The collected production workshop data includes emissions, energy consumption, equipment status and material flow data, and the production workshop data is collected through sensors.

3. The method according to claim 2, characterized in that The non-time series data collected by the sensors is directly transmitted to the protocol parsing gateway for preprocessing. The time series data collected by the sensors is transmitted to the ERP / MES system and uploaded to the time series database, where it is preprocessed by the protocol parsing gateway. The preprocessed multimodal data is fused by the protocol parsing gateway: D raw ={(t i ,e j ,v k )},e j ∈{equipment ID, process number, logistics node} Where, t i is the timestamp of the device, e j is the carbon emission entity identifier, v k The original consumption.

4. The method according to claim 1, wherein The process-level carbon flow dynamic accounting mechanism model is constructed in combination with the production system process parameters. The production system process parameters include the process instantaneous power and process duration. The carbon flow dynamic accounting mechanism model is expressed as: Where α is the carbon emission factor, P k (t) is the instantaneous power of process k, φ ik (t) is the energy consumption of process k for energy i, ε i (t) is the real-time carbon emission intensity of energy source i, M jk (t) is the equivalent carbon value of raw material j in process k. Δt is the time interval, t is time, β(t) is the time-varying adjustment factor of material carbon emissions, C k (t) is the total carbon emission of process k at time t.

5. The method according to claim 1, wherein Through carbon flow mapping modeling, the process-level anchoring parameters are coupled with the equipment-level dynamic data to construct a carbon flow intensity tensor model, which is expressed as: Where, α i (t) is the carbon emission coefficient of energy, E i is the carbon impact factor of the equipment, L i is the logistics path topology vector, N is the total number of carbon flow paths, τ t (t) is the time carbon flow tensor, P i is the energy transfer matrix of the process chain, m is the number of processes; the matrix P i In the row direction, the output process The column direction is the input process C(t); the matrix P i The element p in rs Update it with the following formula: Where, L i→k is the flow path weight of the logistics between process i and process k within time t.

6. The method according to claim 5, characterized in that Based on the carbon flow intensity tensor model, multi-granularity carbon flow features are extracted, and real-time digital twin deduction of carbon flow propagation is realized through the LSTM-Transformer hybrid network; among them, LSTM is used to capture the long-term carbon flow evolution trend, Transformer is used to learn the interactivity of multi-device carbon flow, and the attention mechanism is used to calculate the carbon impact weights between devices.

7. The method according to claim 1, characterized in that By dynamically coupling the carbon flow intensity and the value flow intensity through multi-dimensional economic indicators, a differential equation is established to describe the real-time interaction between the carbon flow and the value flow. The differential equation is expressed as: Where, β k (t) is the value conversion efficiency factor of process k, U k is an external variable, is the energy coupling matrix, C k,i (t) is the dynamic coupling coefficient, λ k is an external regulatory factor, V k is the real-time carbon flow value of the k-th node.

8. The method according to claim 7, characterized in that Constructing a carbon-value correlation matrix: Among them, the matrix The row elements in are carbon emission types, and the column elements are value categories.

9. The method according to claim 8, characterized in that The carbon-value association matrix is ​​analyzed by carbon flow value flow to construct a two-layer topology optimization model. In the constructed two-layer topology optimization model, the upper layer goal is to minimize the total entropy of the carbon flow topology network: Where C e is the carbon flow rate of a single path e, C total is the total carbon flow, E is the carbon flow transfer path; The lower-level goal is to meet the process cycle time constraints: Where, T k is the actual cycle time of the kth process, T spec is the maximum takt time threshold allowed for this process, and K is the set of process steps in the entire process.

Images