Zero-carbon comprehensive energy digital management method and equipment

Abstract

The invention discloses a zero-carbon comprehensive energy digital management method and equipment, belongs to the field of energy management, and is used for solving the problem of inaccurate energy digital management regulation and control. Obtaining a dynamic carbon flow model of carbon emission space-time distribution and a dynamic carbon sink model of carbon absorption space-time change; performing coupling calculation on the dynamic carbon flow model and the dynamic carbon sink model on a unified spatial-temporal scale to fit and generate a net carbon spatial-temporal tensor of the park in a scheduling period; the net carbon space-time tensor is used for quantifying the net carbon emission of each region in each time slice in the park; tensor decomposition is carried out on the net carbon space-time tensor, and target areas which are adjacent in space but opposite in function are selected; and regulating and controlling energy supply and load demands among the target regions based on the carbon balance levels and the time dynamic modes of the target regions so as to generate a collaborative optimization strategy.

Description

Technical Field

[0001] This application relates to the field of energy management technology, and in particular to a zero-carbon integrated energy digital management method and device. Background Technology

[0002] With the deepening implementation of dual-carbon goals, precise and intelligent carbon management has become an inevitable requirement for industrial parks, which are the core carriers of industrial agglomeration and energy consumption. To achieve carbon-neutral operation of industrial parks, the key lies not only in controlling the total amount of carbon emissions, but also in the refined perception and regulation of the spatial distribution and temporal dynamics of carbon emissions, as well as their complex interaction with the carbon absorption process.

[0003] Currently, carbon management at the industrial park level mainly relies on traditional carbon accounting methods. These methods typically perform ex-post calculations based on macro-level statistical data, such as the total annual or monthly carbon emissions at the park level. However, the operating status of carbon source equipment is constantly changing, making it impossible to reveal the specific spatial sources of carbon emissions within the park (e.g., which workshop or production line is a hotspot) and its dynamic changes over different periods. Therefore, the spatiotemporal granularity of energy management is coarse and cannot support precise regulation.

[0004] Furthermore, managing carbon emissions and carbon absorption as two separate systems lacks the ability to correlate them on a unified spatiotemporal scale. This results in the inability to quantify the net carbon environmental contribution of any region within the park at any given time, the inability to achieve local carbon balance through inter-regional collaboration, insufficient precision in energy regulation, and limited overall benefits. Summary of the Invention

[0005] To address the aforementioned issues, this application proposes a zero-carbon integrated energy digital management method, comprising: Carbon calculations are performed on multi-dimensional data of the park during the scheduling cycle to obtain a dynamic carbon flow model of the spatiotemporal distribution of carbon emissions and a dynamic carbon sink model of the spatiotemporal changes of carbon absorption. The dynamic carbon flow model and the dynamic carbon sink model are coupled and calculated on a unified spatiotemporal scale to fit and generate a net carbon spatiotemporal tensor of the park during the scheduling cycle. The net carbon spatiotemporal tensor is used to quantify the net carbon emissions of each region in the park at each time slice. Tensor decomposition is performed on the net carbon spatiotemporal tensor to select target regions that are spatially adjacent but functionally opposed. Based on the carbon balance level and temporal dynamic pattern of the target regions, the energy supply and load demand among the target regions are regulated to generate a collaborative optimization strategy.

[0006] In one example, the carbon calculation of multi-dimensional data of the park during the scheduling cycle to obtain the dynamic carbon flow model and dynamic carbon sink model of the park specifically includes: dividing the geographical area of ​​the park into grid units and establishing a mapping relationship between each grid and carbon source equipment and carbon sink resources in the park; calculating the carbon emissions of each grid in each time slice based on the operating parameters of energy demand equipment and the carbon emission factor labels of energy supply equipment to construct the dynamic carbon flow spatial distribution of the park; and calculating the carbon absorption of each grid in each time slice based on the carbon sequestration energy efficiency factor of carbon sink resources and the adjustment factor corresponding to environmental data to construct the dynamic carbon sink spatial distribution of the park.

[0007] In one example, the carbon emissions of each grid in each time slice are calculated based on the operating parameters of energy-demanding devices and the carbon emission factor labels of energy-supplying devices. Specifically, this includes: performing carbon accounting on the operating parameters of energy-demanding devices within the grid based on the carbon emission factor of energy-supplying devices to obtain the direct carbon emissions of energy-demanding devices in the current grid; compensating for the carbon emissions of adjacent grids in the previous time slice based on the distance-related diffusion coefficient and exponential decay function to obtain the carbon diffusion impact of adjacent grids; the carbon diffusion impact decreases exponentially with increasing distance and is related to the modulation effect of wind direction and terrain on carbon diffusion; and summing the direct carbon emissions with the carbon diffusion impact of all adjacent grids to obtain the carbon emissions of the current grid in the current time slice.

[0008] In one example, tensor decomposition is performed on the net carbon spatiotemporal tensor to select target regions that are spatially adjacent but functionally opposed. Specifically, this includes: constructing a net carbon spatiotemporal tensor within a scheduling period based on the spatial distribution of dynamic carbon flow and the spatial distribution of dynamic carbon sink; the net carbon spatiotemporal tensor includes a spatial grid dimension and a time series dimension; tensor decomposition is performed on the net carbon spatiotemporal tensor to extract target spatial emission pattern vectors representing typical carbon emission spatial distributions, target spatial absorption pattern vectors representing carbon absorption spatial distributions, and target temporal dynamic pattern vectors representing dynamic changes in carbon; based on the spatial relationship between the target spatial emission pattern vectors and the target spatial absorption pattern vectors, regions that are spatially adjacent but functionally opposed are identified to obtain the target regions.

[0009] In one example, the tensor decomposition of the net carbon spatiotemporal tensor specifically includes: when using CP decomposition, decomposing the net carbon spatiotemporal tensor into the sum of a number of rank-number components, each component including two spatial pattern vectors, one temporal dynamic pattern vector, and a scalar representing the component weights; when using Tucker decomposition, decomposing the net carbon spatiotemporal tensor into the product of a core tensor, two spatial factor matrices, and a temporal factor matrix; from the resulting vector group or factor matrix, extracting at least one spatial emission pattern vector representing the typical spatial distribution of carbon emissions, at least one spatial absorption pattern vector representing the spatial distribution of carbon absorption, and at least one temporal dynamic pattern vector representing the dynamic changes of carbon; calculating the sum of squares of the elements of each component weight in CP decomposition or the core tensor in Tucker decomposition to determine the contribution of each extracted pattern vector to the net carbon spatiotemporal tensor; and determining the pattern vectors corresponding to the top-ranked (preset number) component or core tensor elements as target pattern vectors.

[0010] In one example, identifying spatially adjacent but functionally opposing regions based on the spatial relationship between the target spatial emission pattern vector and the target spatial absorption pattern vector specifically includes: mapping the target spatial emission pattern vector and the target spatial absorption pattern vector back to two-dimensional geographic space to obtain typical carbon source distribution maps and typical carbon sink distribution maps; identifying strong carbon source regions with a weight higher than a preset emission threshold in the emission pattern and a weight lower than a preset absorption threshold in the absorption pattern, and identifying strong carbon sink regions with a weight higher than a preset absorption threshold in the absorption pattern and a weight lower than a preset emission threshold in the emission pattern; filtering out strong carbon source regions and strong carbon sink regions that meet the condition of being physically adjacent to each other based on spatial topological relationships; spatial proximity includes geometric adjacency determination and / or distance threshold determination; pairing regions that simultaneously meet the conditions of functional opposition and spatial proximity to identify spatially adjacent but functionally opposing regions.

[0011] In one example, based on the carbon balance level and time dynamic pattern of the target area, energy supply and load demand among target areas are regulated to generate a collaborative optimization strategy. Specifically, this includes: identifying transferable loads from carbon source areas within the target area; adjusting the operating schedule of transferable loads according to the peak carbon absorption period indicated by the corresponding time dynamic pattern vector of the target area, and / or the low carbon period indicated by the grid carbon emission factor label, so as to achieve optimal temporal matching between carbon emissions and carbon absorption; and / or determining the carbon balance level of the target area based on the carbon emissions and carbon absorption of the target area; and when the carbon balance level is lower than the preset level, regulating local clean energy to meet the load demand of carbon source areas within the target area.

[0012] In one example, after regulating energy supply and load demand among target areas based on the carbon balance level and time dynamic pattern of the target area to generate a collaborative optimization strategy, the method further includes: simulating and reasoning the collaborative optimization strategy in a digital twin model of the park to obtain the net carbon spatiotemporal distribution of the target area; for each time slice, aggregating the net carbon emissions of all grids in the net carbon spatiotemporal distribution to obtain the simulated comprehensive net carbon emissions of the target area; and generating an energy management report for the target area based on the comprehensive net carbon emissions and the collaborative optimization strategy.

[0013] On the other hand, embodiments of this application provide a zero-carbon integrated energy digital management device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute any of the above-described zero-carbon integrated energy digital management methods.

[0014] The above-described technical solutions adopted in the embodiments of this application can achieve the following beneficial effects: By integrating spatial modeling and multi-dimensional data, carbon calculations are brought down from the overall park to spatial units. This not only quantifies the park's overall carbon footprint, but also accurately traces the spatiotemporal hotspots of carbon emissions and carbon absorption, clearly revealing when and where carbon is generated and eliminated, providing an unprecedented data foundation for precise management.

[0015] By constructing a net carbon spatiotemporal tensor and performing data mining through tensor decomposition, stable carbon emission and carbon sink spatial patterns can be automatically and efficiently extracted from massive spatiotemporal data of the latest actual situation in the park. Furthermore, it can intelligently identify spatially adjacent but functionally opposing regions, thus achieving automated and intelligent discovery of targeted optimization areas.

[0016] Breaking away from the traditional model of independent operation of energy and carbon management systems, this approach uses carbon as a key link to establish a linkage between energy supply and demand in identified target areas. By spatially scheduling clean energy based on carbon balance levels and flexibly adjusting load demand according to dynamic carbon patterns over time, it maximizes the park's net carbon benefits at the system level.

[0017] In summary, by deconstructing the macro-level carbon neutrality target into a series of actionable and synergistic optimization strategies at specific temporal and spatial scales and for specific regions, the carbon emission reduction work of the park is made systematic and evidence-based, greatly enhancing the achievability and operational efficiency of the park-level carbon neutrality target. Attached Figure Description

[0018] To more clearly illustrate the technical solution of this application, some embodiments of this application will be described in detail below with reference to the accompanying drawings, in which: Figure 1 A flowchart illustrating a zero-carbon integrated energy digital management method provided in this application embodiment; Figure 2 This is a schematic diagram of the structure of a zero-carbon integrated energy digital management device provided in an embodiment of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] Some embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0021] Figure 1 This is a flowchart illustrating a zero-carbon integrated energy digital management method provided in an embodiment of this application. Certain input parameters or intermediate results in this process allow for manual adjustment to help improve accuracy.

[0022] The analysis method involved in the embodiments of this application can be implemented by a terminal device or a server, and this application does not impose any special limitations on it. For ease of understanding and description, the following embodiments are all described in detail using a server as an example.

[0023] It should be noted that the server can be a single device or a system composed of multiple devices, i.e., a distributed server. This application does not make any specific limitations on this.

[0024] Figure 1 The process includes the following steps: S101: Perform carbon calculations on multi-dimensional data of the park during the scheduling cycle to obtain a dynamic carbon flow model of the spatiotemporal distribution of carbon emissions and a dynamic carbon sink model of the spatiotemporal changes of carbon absorption.

[0025] The multi-dimensional data includes energy data, equipment operating parameters, environmental data, and carbon sink resource data.

[0026] Energy data refers to the output and fuel consumption data of various energy supply sides (such as power grid, gas, photovoltaic power generation, wind power generation, and energy storage equipment) within the park, as well as the load data of the energy demand side (such as factories, office buildings, and data centers). It should be noted that the load data on the energy demand side includes the total energy consumption of a region or a building. For example, the total electrical load of Factory A at a certain time slice is 500kW.

[0027] Equipment operating parameters refer to the operating status, energy efficiency, start-up and shutdown times of key carbon source equipment (such as boilers, air compressors, and chillers). For example, the current operating frequency of air compressor A1 in the plant at a certain time slice is 45Hz, the outlet pressure is 0.7MPa, and the input power is 80kW.

[0028] Environmental data refers to information such as temperature, humidity, light intensity, wind speed, and wind direction from meteorological stations, and data from air quality monitoring stations. Concentration, etc. These data directly affect energy demand, renewable energy generation efficiency, and the absorption capacity of carbon sink resources.

[0029] Carbon sink resource data refers to spatial information such as the area, vegetation type, and growth status of ecological resources such as green space, woodland, and water bodies within the park.

[0030] In this example, after multi-dimensional data is input, data preprocessing and time slice alignment are required. Since the source data acquisition frequency and timestamp benchmark are inconsistent, a unified system time benchmark and fixed calculation time slices (e.g., 5 minutes, 15 minutes) are established to synchronize the heterogeneous data streams. For data at non-aligned time points, data interpolation or aggregation methods are used to generate a spatiotemporally aligned, consistent multi-dimensional data set for each time slice. This step ensures the accuracy and comparability of the subsequently constructed dynamic carbon flow model, dynamic carbon sink model, and net carbon spatiotemporal tensor in the time dimension, forming the basis for reliable simulation and optimization. For example, the system uses 5-minute intervals as time slices. For the slice from 14:00 to 14:05, the system takes the average of all load data within that time period as the load value for that slice. For illumination data, linear interpolation is performed using the values ​​at 14:00 and 14:05 to calculate the value at 14:02:30 as a representative value, or the value at 14:00 is directly used as an approximation for the entire slice.

[0031] Carbon computation refers to the process of quantifying all relevant carbon emissions and carbon absorption activities within a specific time period of a scheduling cycle. A dynamic carbon flow model is a quantitative model that reflects the dynamic changes in carbon emissions over time and space, while a dynamic carbon sink model is a quantitative model that reflects the dynamic changes in carbon absorption capacity over time and space. For example, dividing the geographical area of ​​a park into grid cells achieves spatial alignment.

[0032] It should be noted that the carbon calculation for carbon emissions includes energy data and equipment operating parameters within the park.

[0033] Carbon calculations for carbon absorption consider both environmental data and carbon sink resource data within the park. Environmental data serves as a dynamic adjustment factor to modify the carbon sequestration capacity of the carbon sink resource data. For example, under bright sunlight, plant photosynthesis is strong, resulting in a high carbon sequestration rate; in winter or at night, the carbon sequestration rate is low or even zero.

[0034] In summary, when constructing a dynamic carbon flow model, energy system data provides the total energy consumption distribution of the park on a spatiotemporal scale, which is the macroscopic basis for calculating carbon emissions. Meanwhile, equipment operating parameters enable a refined decomposition and source tracing of the total load, allowing carbon emissions to not only be located to subsequent geographic grids but also to be associated with specific equipment, providing data support for the subsequent generation of collaborative optimization strategies for specific equipment.

[0035] S102: Couple the dynamic carbon flow model and the dynamic carbon sink model on a unified spatiotemporal scale to fit and generate the net carbon spatiotemporal tensor of the park during the scheduling cycle; the net carbon spatiotemporal tensor is used to quantify the net carbon emissions of each region in the park at each time slice.

[0036] The unified spatiotemporal scale means that the carbon flow model and the carbon sink model are built on the same grid cell division and the same time slices. That is, both models use the same grid system. Each grid has unique spatial latitude and longitude coordinates, and both models use the same time slice sequence. For example, both are based on a 50m × 50m grid and a 30-minute time interval for calculations. This ensures that the data is comparable in time and space and can be directly used for mathematical operations.

[0037] The purpose of coupled computation is to synthesize carbon activity describing the same location and time slice in carbon flow and carbon sink models. It simulates the net environmental effect in the physical world where carbon emissions are partially offset by carbon absorption processes.

[0038] The net carbon spatiotemporal tensor is a multidimensional data array whose core dimensions include a spatial grid (such as a latitude and longitude grid) and a time series. These three dimensions are the spatial X-coordinate, spatial Y-coordinate, and time, respectively. For example, it can be imagined as a data cube, where the length and width of the cube represent the geographic space of the park, and the height represents the time axis. Each small cell in the cube stores a specific numerical value, namely the net carbon emissions at that spatial location at that point in time.

[0039] For ease of understanding, when only considering spatial location indexes and time series, the data form of this tensor can be regarded as a net carbon spatiotemporal matrix, where the rows of the matrix correspond to spatial units, the columns correspond to time points, and the element values ​​are net carbon emissions.

[0040] It should be noted that the coupled calculation can be performed as follows: First, for each time slice and each grid cell of the time slice sequence, net carbon emissions are calculated. For example, the arithmetic calculation formula is... =E( )-A( ), where E( A(i,j) refers to the carbon emissions of a grid (i,j) sliced ​​at time t, obtained from a dynamic carbon flow model. ) is the amount of carbon absorbed by the grid (i,j) at time t, obtained from the dynamic carbon sink model. It represents the net carbon emissions of grid (i, j) at time t.

[0041] Then, all the calculated net carbon emissions values ​​are organized into a three-dimensional data structure, namely the net carbon spatiotemporal tensor, according to their spatial and temporal coordinates. The three dimensions of this tensor correspond to the spatial X-axis, spatial Y-axis, and time axis, respectively. For example, the system initializes a three-dimensional array (i.e., a tensor) in memory with dimensions [X, Y, T], where X and Y are the dimensions of the spatial grid, and T is the total number of time points. Each calculated... The values ​​are filled into the corresponding (i, j, t) positions in this 3D tensor. This process is repeated across all grid points and time points until the entire tensor is filled.

[0042] Clearly, the net carbon spatiotemporal tensor describes the net carbon environmental status of each location within the park at each time slice. It is the result of positive and negative values ​​canceling each other out, with positive values ​​indicating net carbon emission areas and negative values ​​indicating net carbon absorption areas.

[0043] In summary, the net carbon spatiotemporal tensor not only focuses on the total net emissions of the park, but also on the spatial distribution density of emissions and absorption, thus enabling precise identification of areas with the greatest environmental pressure. It assigns a clear and comparable net carbon indicator to each region and each time slice within the park, transforming the management objective from vague emissions reduction to precisely reducing the net carbon emissions of a specific region during a specific period. That is, it no longer focuses on the vague total emissions of the park, but on the net carbon emissions of each region and each time slice, directly revealing which areas within the park represent ongoing environmental liabilities (net carbon sources) and which represent environmental assets (net carbon sinks).

[0044] S103: Perform tensor decomposition on the net carbon spatiotemporal tensor and select target regions that are spatially adjacent but functionally opposed.

[0045] In this example, the technical challenge of automatically and accurately identifying combinations of regions with high optimization potential in existing park carbon management technologies can be solved. By using high-order tensor decomposition, essential and stable spatial and temporal patterns can be extracted from massive and dynamic net carbon data. Based on these patterns, regions that are geographically close but conflicting in terms of carbon function can be intelligently identified, providing clear targets for subsequent precise and coordinated optimization.

[0046] Based on this, the complex primitive tensor is decomposed into a series of low-dimensional pattern vectors with clear physical meaning.

[0047] The model vectors include the target space emission model vector, the target space absorption model vector, and the target time dynamic model vector.

[0048] The target spatial emission pattern vector has a length equal to the total number of spatial grids. The values ​​in this vector, whether high or low, identify which spatial regions are typical and stable carbon emission hotspots. It reflects the spatial distribution characteristics of the most core and persistent carbon sources within the park.

[0049] The target spatial absorption pattern vector has a length equal to the total number of spatial grids. The values ​​in this vector, whether high or low, identify which spatial regions are typical and highly efficient carbon absorption core areas. It reflects the spatial distribution characteristics of the most core and persistent carbon sinks within the park.

[0050] The target time dynamic pattern vector length is equal to the total number of time points. This vector describes the common patterns of spatial emission and absorption patterns over time (e.g., the patterns of variation with production shifts and solar radiation intensity).

[0051] Functional opposition refers to a situation where one region has a high weight in the emission pattern vector, while another region has a high weight in the absorption pattern vector, forming an opposing source-sink relationship in terms of carbon function. Spatial proximity refers to two functionally opposed regions being physically adjacent to each other or very close in distance, for example, less than a preset distance threshold. The system automatically pairs regions that simultaneously meet the conditions of functional opposition and spatial proximity to obtain the target region.

[0052] S104: Based on the carbon balance level and time dynamic pattern of the target area, regulate the energy supply and load demand among the target areas to generate a collaborative optimization strategy.

[0053] In this example, the traditional, isolated operating model between energy supply and demand is broken, and carbon is used as the key link and optimization target to coordinate and regulate energy flow and consumption between regions in both time and space dimensions.

[0054] The carbon balance level of a target region is a quantitative assessment of its overall net carbon offsetting capacity. Based on a comprehensive analysis of regional carbon emissions and carbon absorption, this capacity determines the spatial boundaries of collaborative optimization and the priority of local resource utilization. For example, if a paired region has a strong carbon balance level (i.e., the carbon sink region's absorption capacity is sufficient to offset most of the emissions from the carbon source region), the optimization strategy should prioritize achieving energy balance within the region, such as directly supplying local photovoltaic power to the carbon source region to reduce the input of high-carbon grid electricity. Conversely, if the carbon balance level is weak, the strategy may need to consider introducing more external clean energy at the industrial park level.

[0055] A time-dynamic model is a vector extracted from tensor decomposition that characterizes the changes in carbon emissions and carbon sinks over time. It reveals the tidal patterns of carbon behavior in the industrial park, such as peak carbon absorption in the afternoon and peak carbon emissions during production shifts. This model determines the time window for collaborative optimization. It guides the system on when and what actions to take. Optimization strategies attempt to shift energy demand (load) from carbon-intensive periods to low-carbon periods.

[0056] In other words, the directive prioritizes local clean energy (such as solar and wind power) to meet the load demand in carbon hotspot areas. This essentially involves the targeted delivery of green carbon sink energy to high-carbon-source areas, achieving spatial carbon balance on the energy supply side. Furthermore, based on the peak carbon absorption periods indicated by the time-dynamic model (when the park's net carbon pressure is lowest) and the low-carbon periods indicated by the grid's carbon emission factor label, the directive adjusts transferable loads within the carbon-source area (such as electric vehicle charging, water pumps, and some production processes) to operate during these periods, thereby achieving temporal carbon balance on the energy demand side.

[0057] In addition, the above-mentioned control measures can be combined. For example, during the peak carbon absorption period of 12:00-14:00, the backup charging piles in carbon source area A can be activated to charge vehicles in the factory area while disconnecting the power grid and switching to direct power supply from the photovoltaic power station in area B. The pre-cooling time of the central air conditioning in carbon source area C can be adjusted from 08:00 (peak carbon emission period) to 04:00 (low carbon period of the power grid).

[0058] In summary, through the aforementioned targeted scheduling of energy supply in the spatial dimension and flexible adjustment of load demand in the temporal dimension, the system automatically generates a series of collaborative optimization strategies aimed at maximizing the region's net carbon benefits.

[0059] It should be noted that, although the embodiments in this application are based on... Figure 1 Steps S101 to S104 will be described sequentially, but this does not mean that steps S101 to S104 must be performed in a strict order. The reason this embodiment follows this order is... Figure 1The order in which steps S101 to S104 are described is provided to facilitate understanding of the technical solutions of the embodiments of this application by those skilled in the art. In other words, in the embodiments of this application, the order of steps S101 to S104 can be appropriately adjusted according to actual needs.

[0060] pass Figure 1 The method, through spatial modeling and multi-dimensional data fusion, extends carbon calculation from the overall park to spatial units. This not only quantifies the overall carbon footprint of the park, but also accurately traces the spatiotemporal hotspots of carbon emissions and carbon absorption, clearly revealing when and where carbon is generated and eliminated, providing an unprecedented data foundation for precise management.

[0061] By constructing a net carbon spatiotemporal tensor and performing data mining through tensor decomposition, stable carbon emission and carbon sink spatial patterns can be automatically and efficiently extracted from massive spatiotemporal data of the latest actual situation in the park. Furthermore, it can intelligently identify spatially adjacent but functionally opposing regions, thus achieving automated and intelligent discovery of targeted optimization areas.

[0062] Breaking away from the traditional model of independent operation of energy and carbon management systems, this approach uses carbon as a key link to establish a linkage between energy supply and demand in identified target areas. By spatially scheduling clean energy based on carbon balance levels and flexibly adjusting load demand according to dynamic carbon patterns over time, it maximizes the park's net carbon benefits at the system level.

[0063] In summary, by deconstructing the macro-level carbon neutrality target into a series of actionable and synergistic optimization strategies at specific temporal and spatial scales and for specific regions, the carbon emission reduction work of the park is made systematic and evidence-based, greatly enhancing the achievability and operational efficiency of the park-level carbon neutrality target.

[0064] based on Figure 1 In addition to the method described herein, this application also provides some specific implementation schemes and extended schemes of the method, which will be further described below.

[0065] In one example, through grid-based management and multi-source data fusion, the activities of the park are dynamically quantified from the macro-statistical level down to the fine spatiotemporal units.

[0066] Based on this, carbon calculations are performed on the park's multi-dimensional data during the scheduling cycle, specifically including: First, the geographical area of ​​the park is divided into grid units, and a mapping relationship is established between each grid and the carbon source equipment and carbon sink resources in the park.

[0067] The park's geographical area can be discretized into several grid units with a preset precision (e.g., 50m x 50m). Each grid unit becomes the smallest unit for calculation and management. A correspondence is established between each grid unit and its internal carbon source equipment (e.g., boilers, air compressors, process units) and carbon sink resources (e.g., green spaces, forests, water bodies).

[0068] Then, based on the operating parameters of energy demand equipment and the carbon emission factor labels of energy supply equipment, the carbon emissions of each grid in each time slice are calculated to construct the dynamic spatial distribution of carbon flow in the park.

[0069] The operating parameters include activity level data such as power, energy consumption, and start / stop status. The carbon emission factor labels of energy supply equipment define the carbon density of different energy sources. For example, grid electricity has a carbon emission factor that varies over time, natural gas has a fixed carbon emission factor, and photovoltaic power has a near-zero carbon emission factor.

[0070] The calculation process may include: determining the carbon source devices contained in each grid based on the mapping relationship between grids and devices; for each grid, at each time slice, multiplying and summing the energy consumption data of all carbon source devices within it with the corresponding carbon emission factor labels to obtain the carbon emissions of that grid in that time slice.

[0071] It should be noted that carbon calculations can be based on total emissions or on direct monitoring.

[0072] Under the total load calculation method: the total electricity load of the grid in each time slice is obtained from the electricity meter at the main entrance of the park. Based on the carbon emission factor label corresponding to the grid and the total electricity load, the total carbon emissions of the grid in each time slice are calculated. For example, if the average electricity load of the office building in a certain time slice (30 minutes) is 100kW, then the energy consumption in that time slice is 50kWh, and the grid carbon emission factor is 0.5kg. Therefore, the carbon emissions of the office building sliced ​​at this time are 25kg. .

[0073] It should be noted that electrical load is a unit of power, representing instantaneous power, while carbon emission factor represents the amount of carbon dioxide emitted for every kilowatt-hour of electricity consumed.

[0074] In the direct monitoring approach: when all key carbon source devices within the grid have independent metering capabilities, in each time slice, the carbon emissions of each device are obtained by multiplying the energy consumption (e.g., electricity consumption data) of each device by the carbon emission factor label corresponding to the grid. Then, these are accumulated to obtain the total carbon emissions of the grid in each time slice.

[0075] It should be noted that energy consumption may include electricity data, fossil fuel data, heat data, etc.

[0076] Based on this, direct monitoring can be prioritized to obtain the highest accuracy, and when the data is incomplete, it can be automatically and seamlessly switched to total amount calculation, thereby ensuring that the dynamic carbon flow model can be reliably constructed in various application scenarios.

[0077] Finally, based on the carbon sequestration efficiency factor of carbon sink resources and the adjustment factor corresponding to environmental data, the carbon absorption of each grid in each time slice is calculated to construct the dynamic spatial distribution of carbon sink in the park.

[0078] The carbon sequestration efficiency factor of carbon sink resources serves as a baseline for their capacity; for example, the annual carbon sequestration volume per unit area of ​​lawn or forest under standard conditions. Adjustment factors corresponding to environmental data are crucial for achieving dynamic monitoring. Environmental data (such as light intensity, temperature, humidity, wind speed, and season) significantly influence the photosynthetic rate of carbon sinks such as plants.

[0079] Based on the physiological characteristics of carbon sink vegetation, a time-scale conversion model is established. The calculation process may include: first, converting the annual carbon sequestration efficiency factor into a baseline absorption rate corresponding to the time slice; then, obtaining the dynamic absorption rate based on the environmental adjustment coefficient and the baseline absorption rate; and finally, calculating the carbon absorption within the time slice based on the dynamic absorption rate and the grid carbon sink area. For example, the dynamic absorption rate can be obtained by multiplying the environmental adjustment coefficient and the baseline absorption rate, and the carbon absorption can be obtained by multiplying the dynamic absorption rate and the grid carbon sink area.

[0080] It should be noted that the carbon sink type and carbon sink area within each grid are determined based on the mapping relationship between the grid and carbon sink resources.

[0081] In addition, a dynamic adjustment coefficient for each time slice is obtained by looking up a table (e.g., the coefficient is 1.2 when the light intensity is strong and the temperature is suitable, and the coefficient is 0.2 when it is night or there is no light).

[0082] Based on this, by introducing environmental adjustment factors to make real-time corrections to the basic carbon sequestration energy efficiency, the limitations of traditional static accounting are broken through, and dynamic perception of carbon absorption capacity is realized.

[0083] In summary, by precisely pinpointing the abstract total carbon emissions or absorption of the park to a specific geographical location, it becomes clear which devices generate carbon or which resources absorb it within each grid.

[0084] Furthermore, the actual carbon emissions of a grid are also significantly affected by the carbon diffusion of surrounding grids. By introducing an atmospheric diffusion model, the dynamic simulation of the spatiotemporal propagation of carbon emissions can be achieved, greatly improving the accuracy and realism of the model.

[0085] Based on this, by adopting the above-mentioned direct monitoring or total calculation method, and after performing carbon accounting on the operating parameters of energy demand equipment within the grid based on the carbon emission factor of energy supply equipment, the direct carbon emissions of energy demand equipment in the current grid are obtained.

[0086] Then, based on the distance-dependent diffusion coefficient and the exponential decay function, the carbon emissions of adjacent grids in the previous time slice are compensated to obtain the carbon diffusion influence of adjacent grids. The carbon diffusion influence decreases exponentially with increasing distance and is related to the modulation effect of wind direction and topography on carbon diffusion.

[0087] This can be achieved by simulating the decay of carbon concentration with distance based on an exponential decay function. The larger the distance *d*, the smaller the impact, and the decay coefficient controls the decay rate. For example, the exponential decay function... , This refers to the attenuation coefficient. This refers to the distance between adjacent grid cells.

[0088] Regarding directional diffusion, the distance-dependent diffusion coefficient is determined by the following factors: based on the prevailing wind direction frequency statistically derived from historical meteorological data, the diffusion coefficient of upwind grids is higher than that of downwind grids; based on slope data from the digital elevation model, the diffusion coefficient of flat areas is higher than that of complex terrain areas. Based on layout data from the building information model, the diffusion coefficient of densely built-up areas is adjusted accordingly to simulate the blocking effect. For example, the upwind correction factor is greater than 1, the downwind correction factor is less than 1, the flat terrain correction factor is equal to 1, the complex terrain correction factor is less than 1, the densely built-up area correction factor is less than 1, and the building ventilation corridor correction factor is greater than 1. It should be noted that upwind refers to adjacent grids being upstream of the current grid in the wind, and downwind refers to adjacent grids being downstream of the current grid in the wind.

[0089] That is, the comprehensive diffusion coefficient is obtained by multiplying the preset base diffusion coefficient, wind direction correction factor, terrain correction factor and building correction factor.

[0090] Based on this, the carbon diffusion influence of adjacent grids is obtained by multiplying the comprehensive diffusion coefficient, the exponential decay function value, and the carbon emissions of adjacent grids in the previous time slice.

[0091] Finally, the direct carbon emissions are summed with the carbon diffusion effects of all adjacent grids to obtain the carbon emissions of the current grid in the current time slice.

[0092] In summary, spatial correlation modeling of carbon emissions has been achieved. The carbon emissions in one grid not only affect itself but also contribute to the downwind grid under the influence of wind and topography. This allows for dynamic simulation of the diffusion paths and cumulative effects of carbon emissions within the park, providing precise spatial insights for identifying severely carbon-affected areas and generating coordinated control strategies.

[0093] In one example, its inherent, stable core features are extracted, and regions with high optimization potential are intelligently discovered based on these features.

[0094] Based on this, the net carbon spatiotemporal tensor is decomposed into tensors, selecting target regions that are spatially adjacent but functionally opposed, specifically including: First, based on the spatial distribution of dynamic carbon flow and the spatial distribution of dynamic carbon sink, a net carbon spatiotemporal tensor is constructed within the scheduling period. This net carbon spatiotemporal tensor includes both spatial grid dimensions and time series dimensions.

[0095] In this process, the dynamic carbon flow and the spatial distribution of carbon sinks under each time slice within the scheduling cycle are coupled, and the net carbon emission data of all time slices are stacked in the time dimension to form a three-dimensional data cube, namely the net carbon spatiotemporal tensor.

[0096] Then, the net carbon spatiotemporal tensor is decomposed to extract the target spatial emission pattern vector representing the typical spatial distribution of carbon emissions, the target spatial absorption pattern vector representing the spatial distribution of carbon absorption, and the target temporal dynamic pattern vector representing the dynamic changes of carbon.

[0097] Specifically, using a specific tensor decomposition algorithm (such as CP decomposition or Tucker decomposition), the net carbon spatiotemporal vector is approximated as a combination of several low-dimensional model vectors. Regions in the target spatial emission model vector with values ​​exceeding a preset threshold are considered core carbon source areas within the scheduling cycle, while regions in the target spatial absorption model vector with values ​​exceeding a preset threshold are considered core carbon sink areas within the scheduling cycle. The target temporal dynamic model vector characterizes the temporal variation patterns (such as daily and weekly variations) commonly followed by the aforementioned spatial models.

[0098] Finally, based on the spatial relationship between the target space emission pattern vector and the target space absorption pattern vector, regions that are spatially adjacent but functionally opposed are identified to obtain the target region.

[0099] Specifically, the target spatial emission pattern vector and target spatial absorption pattern vector are mapped back to two-dimensional geographic space to form typical carbon source distribution maps and typical carbon sink distribution maps. Two types of regions are identified on these maps: first, strong carbon source regions with a weight higher than a preset emission threshold in the emission pattern and a weight lower than a preset absorption threshold in the absorption pattern; and second, strong carbon sink regions with a weight higher than a preset absorption threshold in the absorption pattern and a weight lower than a preset emission threshold in the emission pattern. These two types constitute an opposing relationship between carbon sources and sinks in terms of carbon function. Based on spatial topological relationships, strong carbon source regions and strong carbon sink regions that are physically adjacent to each other are selected. Regions that simultaneously meet the conditions of functional opposition and spatial proximity are paired as the final target regions.

[0100] It should be noted that spatial proximity can include geometric adjacency determination and distance threshold determination.

[0101] Geometric adjacency determination refers to determining whether two regions are adjacent if their corresponding grid cells satisfy either shared edge adjacency or shared corner adjacency, based on a unified spatial grid coordinate system. Shared edge adjacency means that the two grid cells share a common boundary, while shared corner adjacency means that the two grid cells share a common vertex, i.e., they are diagonally connected.

[0102] Distance threshold determination refers to the Euclidean distance between the geometric centers of two grids being less than a preset distance threshold. For example, a circular buffer zone is drawn with the geometric center of one grid as the center and a preset effective cooperative radius R. If the geometric center of another grid falls within this buffer zone, it is determined to be adjacent. The distance threshold R can be set according to the size of the park and the economics of energy transmission technology, for example, 500 meters.

[0103] In summary, by performing tensor decomposition, spatial emission patterns, spatial absorption patterns, and temporal dynamic patterns were extracted. This process achieved dimensionality reduction and noise reduction of the data, revealing the intrinsic core structure driving carbon change in the industrial park. By analyzing the geographical distribution of spatial pattern vectors, the system intelligently identified spatially proximate but functionally opposed carbon source-carbon hole pairs. Overcoming the limitations of manual analysis, it can quickly and accurately locate key region combinations with the greatest potential for synergistic optimization due to geographical proximity, providing clear decision targets for subsequent generation of spatially executable and functionally matched synergistic optimization strategies.

[0104] Furthermore, the net carbon spacetime tensor is decomposed, including the following steps: When using CP decomposition, the net carbon spatiotemporal tensor is decomposed into the sum of rank-number components. Each component includes two spatial mode vectors, one temporal dynamic mode vector, and a scalar representing the component weights.

[0105] One spatial pattern vector may characterize the spatial distribution of carbon emissions, while another spatial pattern vector may characterize the spatial distribution of carbon absorption. The temporal dynamic pattern vector may characterize the dynamic variation of the component, and the scalar weights represent the relative importance of the component in the entire tensor.

[0106] When using Tucker decomposition, the net carbon spatiotemporal tensor is decomposed into the product of a core tensor, two spatial factor matrices, and a time factor matrix.

[0107] The tensor is decomposed into a core tensor and the product of multiple factor matrices. The column vectors of each factor matrix are the pattern vectors for the corresponding dimensions. The core tensor determines how different pattern vectors interact and their magnitudes. The larger the value of an element in the core tensor, the more important the combination of its corresponding spatial and temporal pattern vectors.

[0108] Then, from the decomposed vector group or factor matrix, extract at least one spatial emission pattern vector representing the typical spatial distribution of carbon emissions, at least one spatial absorption pattern vector representing the spatial distribution of carbon absorption, and at least one temporal dynamic pattern vector representing the dynamic changes of carbon.

[0109] From the component set of CP decomposition or from the factor matrix of Tucker decomposition, the system examines all generated spatial and temporal pattern vectors. By analyzing the correlation between these vectors and the pre-calculated spatial distributions of dynamic carbon flux and dynamic carbon sink, the system can intelligently identify and extract spatial emission pattern vectors (vectors whose shape is highly similar to the spatial distribution of carbon emissions), spatial absorption pattern vectors (vectors whose shape is highly similar to the spatial distribution of carbon absorption), and temporal dynamic pattern vectors (time series vectors describing common changing trends).

[0110] Then, the weights of each component in the CP decomposition or the sum of squares of each element in the core tensor in the Tucker decomposition are calculated to determine the contribution of each extracted mode vector to the net carbon spatiotemporal tensor.

[0111] The absolute value of the scalar weight of each component directly reflects its contribution to the net carbon spatiotemporal tensor. A larger weight indicates a higher contribution. For Tucker decomposition, the sum of squares of the absolute values ​​of each element in the core tensor is calculated; the mode vector corresponding to the element with the larger value has a higher contribution.

[0112] Finally, the pattern vectors corresponding to the top-ranked number of components or core tensor elements in terms of contribution are determined as the target pattern vectors.

[0113] In other words, all components are sorted from highest to lowest contribution, and the top N components are selected. The pattern vectors corresponding to these selected components are then used as the target pattern vectors. Similarly, the core tensor elements are sorted from highest to lowest contribution, and the top M core tensor elements are selected. The set of pattern vectors corresponding to these selected elements is then used as the target pattern vectors.

[0114] For example, suppose we have a simplified campus divided into a 2x2 grid, and we observe net carbon emissions over three time slices (T1, T2, T3). This gives us a 2x2x3 net carbon spatiotemporal tensor.

[0115] This tensor can be visualized as three carbon maps: net carbon emissions for grids A, B, C, and D at time T1; net carbon emissions for grids A, B, C, and D at time T2; and net carbon emissions for grids A, B, C, and D at time T3. Grids A and B consistently show positive values, representing typical carbon source areas (such as factories and office buildings), while grids C and D consistently show negative values, representing typical carbon sink areas (such as green spaces and forests). From T1 to T2, the absolute values ​​of all grids increase, potentially corresponding to increased daytime activity.

[0116] Taking CP decomposition as an example, the system performs CP decomposition on this 2×2×3 tensor. Assume that the decomposition yields two core components, i.e., rank R=2. Component 1 has a contribution weight of 0.9, and its spatial pattern vector (emission pattern) is [0.95, 0.3, 0.1, 0.05], corresponding to grids A, B, C, and D. It can be concluded that grid A is a very strong and stable carbon emission source, grid B is the next strongest, while grids C and D emit almost no carbon. The spatial pattern vector (absorption pattern) is [0.05, 0.1, 0.85, 0.4], corresponding to grids A, B, C, and D. It can be concluded that grid C is a very strong and stable carbon sink, grid D is the next strongest, while grids A and B absorb almost no carbon. The time dynamic pattern vector is [0.4, 0.9, 0.7], corresponding to times T1, T2, and T3. It can be concluded that this component pattern reaches its peak at T2 (e.g., noon), which is synchronized with the main carbon emission and absorption activities throughout the day.

[0117] The time-dynamic pattern vector reflects the quantitative changes in the source-sink synergy pattern. In time slice T2, not only do carbon emissions peak, but absorption also peaks, resulting in the highest carbon cycle volume for the entire system. For example, in time slice T2, grid A experiences its highest carbon emissions, while grid C experiences its highest carbon absorption. For instance, the carbon emission intensity of grid A in time slice T1 is the product of 0.38, 0.95, and 0.4.

[0118] Component 2 has a contribution weight of 0.2 and a spatial pattern vector of [0.2, 0.9, 0.05, 0.1], representing specific emissions from grid B. Its spatial pattern vector is [0.1, 0.1, 0.3, 0.95], representing specific absorption from grid D, with a temporal dynamic pattern vector of [0.8, 0.2, 0.4], indicating a different temporal behavior from Component 1. For example, it might reveal the unique activity patterns of a specific group. This group might include: large equipment requiring preheating (peaking in time slice T1), workshops implementing staggered production (time slice T3: activity continues in the evening after get off work), areas primarily illuminated at night (beginning to become active in T3), and specific light-sensitive vegetation (whose carbon absorption is strongest during specific periods), etc.

[0119] Since the weight of component 1 is greater than that of component 2, if one is selected, the three vectors included in component 1 with the highest contribution will be selected as the target pattern vector.

[0120] In one example, based on the carbon balance level and time dynamics of the target region, the energy supply and load demand among the target regions are regulated, including the following steps: On the one hand, the transferable loads in carbon source areas are identified, and the operating schedules of the transferable loads are adjusted according to the peak carbon absorption period indicated by the time dynamic pattern vector corresponding to the target area and / or the low carbon period indicated by the grid carbon emission factor label, so as to achieve optimal time matching between carbon emissions and carbon absorption.

[0121] Identifying transferable loads involves finding equipment within the carbon source area of ​​a target region whose electricity usage time is adjustable without affecting core production or services. Examples include electric vehicle charging stations, pre-cooling / pre-heating systems for central air conditioning, water pumps, and some discontinuous production processes.

[0122] Adjusting the load operation schedule aims to shift load from high-carbon periods to low-carbon periods. Based on the target time dynamic pattern vector, the periods with the strongest carbon sequestration capacity in the carbon sink area are identified (typically the afternoon with abundant sunshine and strong photosynthesis). Operating loads during these periods maximizes real-time absorption of carbon emissions, thereby reducing the park's instantaneous net carbon emissions. Furthermore, based on the dynamic changes in the grid's carbon emission factor label, periods with high overall grid cleanliness are identified (typically periods with high wind and solar power output). Electricity consumption during these periods has lower carbon costs.

[0123] Based on this, it achieves precise timing matching of emissions when carbon is absorbed most quickly or electricity consumption when external electricity is cleanest, using the time difference to optimize the net carbon performance of the entire system.

[0124] On the other hand, the carbon balance level of the target area is determined based on the carbon emissions and carbon absorption of the target area; when the carbon balance level is lower than the preset level, local clean energy is adjusted to meet the load demand of the carbon source area within the target area.

[0125] The calculation involves the ratio of carbon absorption to carbon emissions in the target area; a higher ratio corresponds to a stronger carbon balance rating. The carbon balance rating quantifies the carbon offsetting potential within the target area.

[0126] When the carbon balance level is lower than the preset level, it indicates a severe carbon imbalance within the target area and a weak self-regulation capacity. Therefore, clean energy generated within the park (such as solar and wind power) can be prioritized for transmission to carbon source equipment (such as factories and workshops) within the area.

[0127] Based on this, clean energy resources within the park are invested in areas that generate the greatest marginal emission reduction benefits, achieving intelligence and efficiency in achieving the overall carbon neutrality goal and maximizing the overall net carbon benefits of the park.

[0128] It should be noted that when the target area meets the policy conditions, both of the above policies can be executed simultaneously.

[0129] In one example, a carbon energy digital twin is used as a decision-making laboratory to conduct periodic simulations of collaborative optimization strategies. A comprehensive quantitative indicator, net carbon emissions, is used to uniformly measure the overall merits of the strategies in terms of economic and environmental benefits, thereby supporting the scientific nature and global optimality of decision-making.

[0130] The carbon energy digital twin is a high-fidelity dynamic mapping of the physical entities of the park (including energy equipment, loads, carbon sink resources, etc.) in virtual space. It is a system simulation platform that can synchronize, simulate and predict in real time.

[0131] For example, each generated collaborative optimization strategy (such as adjusting the charging load in area A to the afternoon peak photovoltaic period) can be used as an input command for simulation within a digital twin. The twin will then simulate and adjust the operating status of relevant equipment, the direction and flow of energy, and deduce the dynamic response of the energy and carbon systems within the target area over a complete scheduling cycle, based on the strategy commands.

[0132] Based on this, and taking into account the carbon balance level and time dynamics of the target region, the energy supply and load demand among the target regions are regulated to generate a collaborative optimization strategy, which includes the following steps: First, the collaborative optimization strategy is simulated and reasoned in a digital twin model to obtain the spatiotemporal distribution of net carbon in the target region.

[0133] The system simulates the operating state of the energy system within the scheduling cycle according to policy instructions, thereby obtaining the carbon flow model and carbon sink model for the target region based on the aforementioned carbon calculation model. Finally, based on the carbon flow model and carbon sink model, the system obtains the net carbon spatiotemporal distribution of the target region.

[0134] Then, for each time slice, the net carbon emissions of all grids in the spatiotemporal distribution of net carbon are aggregated to obtain the overall net carbon emissions of the target region.

[0135] For each time slice in the simulation results, the system sums up the net carbon emission values ​​of all grids in the new net carbon spatiotemporal distribution.

[0136] Finally, an energy management report for the target region is generated based on the comprehensive net carbon emissions and synergistic optimization strategies.

[0137] The lower the overall net carbon emissions, the better the environmental benefits. Based on simulation data and overall net carbon emissions, the system automatically generates a structured energy management report. This report includes at least a strategy description, a comparative analysis before and after strategy implementation, and decision-making recommendations, providing managers with comprehensive and intuitive decision-making support.

[0138] In one example, a zero-carbon integrated energy digital management platform is constructed, which mainly includes a data acquisition layer, a data processing and analysis layer (platform layer), an application function layer (business layer), and a user interaction layer.

[0139] The data acquisition layer includes energy data, equipment operation data, carbon source data, carbon sink data, and environmental data.

[0140] Energy data: Deploy smart meters, water meters, gas meters, heat meters, and cooling meters to collect real-time consumption data for various energy sources such as water, electricity, gas, heat, and cooling.

[0141] Carbon source data: Direct emissions: monitoring of fuel consumption of various production and chemical equipment, and energy consumption / emissions of internal transportation (forklifts, commuter vehicles). Indirect emissions: monitoring of the consumption of purchased electricity, heat, cooling, and steam. Other indirect emissions: carbon emissions from employee commuting, business trips, waste disposal, and purchased materials (to be gradually included as needed).

[0142] Carbon sink data: Monitoring of green space area and vegetation type in the park (satellite / drone remote sensing or GIS data). Data on potential carbon sink projects (such as the Carbon Capture, Utilization and Storage (CCUS) project).

[0143] Environmental data: meteorological station (temperature, humidity, light intensity, wind speed and direction), air quality monitoring station ( (Concentration, etc.).

[0144] Equipment operation data: operating status and efficiency parameters of key energy demand equipment, renewable energy equipment (photovoltaics, wind turbines, energy storage), and HVAC systems.

[0145] The aforementioned data acquisition layer can be integrated with third-party systems: connecting with the park's existing building automation system, production management system, ERP system, security system, etc., to obtain relevant energy and operational data.

[0146] The data processing and analysis layer (platform hub) includes the Internet of Things platform, big data platform, digital twin engine, and AI engine.

[0147] IoT platform: responsible for the access, protocol parsing, data cleaning, storage and management of massive heterogeneous devices.

[0148] Big data platform: Provides high-performance data storage, computing, and analysis capabilities.

[0149] Digital Twin Engine: Constructs a virtual mapping of physical entities (buildings, equipment, pipelines, microgrids) in the park to achieve real-time visualization and simulation of their status.

[0150] AI Engine: Integrates machine learning and deep learning algorithms for prediction, optimization, and intelligent decision-making.

[0151] The application functionality layer, with its core business led by the zero-carbon implementation plan research module, includes the following: Park Diagnosis and Baseline Accounting: Based on the collected full data, a standardized carbon accounting model is used to automatically and accurately calculate the park's historical and real-time total carbon emissions, structure, and quantity, draw carbon emission heat maps, and identify key emission sources and high energy-consuming processes.

[0152] Carbon neutrality pathway planning: strictly follow the five-dimensional strategy framework of reducing demand, increasing efficiency, increasing production capacity, carbon sinks, and control.

[0153] Demand Reduction: Analyze energy-saving potential (building energy-saving retrofits, production process optimization, demand-side response potential). Efficiency Improvement: Assess the potential for improving energy system efficiency (equipment energy efficiency upgrades, energy cascade utilization, system optimization). Capacity Building: Assess the development potential and economic viability of local renewable energy sources (distributed photovoltaic, wind power, ground source heat pumps, etc.); evaluate green electricity / green certificate procurement strategies. Carbon Sequestration: Assess the potential for increasing green space and ecological carbon sequestration; evaluate the necessity and strategies for purchasing compliant carbon credits. Control: Strengthen the construction of carbon emission monitoring, verification, and management systems.

[0154] Phased carbon reduction plan development: Based on diagnosis and pathway planning, customized implementation strategies are generated for each phase (e.g., short-term 1-3 years, medium-term 3-5 years, long-term 5-10 years), clearly defining the goals (total amount, quantity), key tasks, technology roadmap, investment estimates, expected emission reductions, and KPIs for each phase. Multi-scenario simulations are supported to evaluate the cost-effectiveness and emission reduction effects of different strategy combinations.

[0155] Carbon neutrality technology database and solution selection: Establish a dynamically updated zero-carbon technology knowledge base (covering energy-saving technology, energy efficiency technology, renewable energy technology, energy storage technology, carbon capture, utilization and storage technology, carbon sink technology, digital management technology, etc.), and combine it with park characteristics (industry type, resource endowment, economic affordability), technology maturity, cost-benefit analysis, intelligently select and recommend the most suitable technology combination to form a refined carbon emission control solution.

[0156] Dynamic optimization and adjustment: Based on real-time monitoring data from the platform and changes in external policies, technologies, and markets, the path and plan are reassessed periodically (or triggered) to provide dynamic optimization suggestions.

[0157] The supporting core functional modules include panoramic energy monitoring and energy efficiency analysis, real-time carbon emission monitoring and tracking, and intelligent optimized operation and control.

[0158] Panoramic energy monitoring and energy efficiency analysis monitors the energy flow of the park and key energy-consuming units in real time, conducts multi-dimensional (time, space, equipment, department) energy efficiency analysis and benchmarking, and identifies anomalies and energy-saving opportunities.

[0159] Real-time carbon emission monitoring and tracking: Based on the accounting model, it enables near real-time tracking and visualization of carbon emissions, and generates carbon emission reports (daily, monthly, and annual reports) to meet compliance and disclosure requirements.

[0160] Intelligent Optimized Operation and Control: Microgrid Coordinated Control: Intelligent coordinated and optimized scheduling of distributed photovoltaic, wind power, energy storage, charging piles, and flexible loads within the park to maximize local clean energy consumption and reduce energy costs and carbon emissions. Energy System Optimization: Based on AI algorithms, operational strategies for HVAC, lighting, and production processes are optimized to achieve demand management, load forecasting, and demand response.

[0161] Carbon sink management and trading support: managing carbon sink resources within the park (green space data, potential projects).

[0162] Connect with external carbon market information to assist in the account management, trading strategy analysis, and decision support of carbon assets (carbon allowances, carbon credits).

[0163] Integrated Management and Decision Support, Visualized Dashboard: Provides managers with a global view and key KPI dashboards for core indicators such as energy, carbon emissions, economy, and environment within the park. Alarm and Incident Management: Sets threshold alarms for energy anomalies, excessive carbon emissions, and equipment malfunctions, and supports work order processing. Project Management and Performance Evaluation: Tracks the progress of tasks, investment completion, and emission reduction effects in the zero-carbon implementation plan, supporting performance evaluation.

[0164] The user interaction layer includes a web-based management platform, a mobile app, and API interfaces.

[0165] Web-based management platform: Designed for park managers, energy engineers, and operations and maintenance personnel, providing comprehensive data viewing, analysis, reporting, configuration, and management functions.

[0166] Mobile App: Provides convenient functions such as viewing key indicators, receiving alarms, processing work orders, and remote control.

[0167] API Interface: An open data interface that supports data interaction and business collaboration with government regulatory platforms, corporate headquarters systems, and third-party service provider systems.

[0168] Based on the same idea, some embodiments of this application also provide devices and non-volatile computer storage media corresponding to the above methods.

[0169] Figure 2 A schematic diagram of a zero-carbon integrated energy digital management device provided in this application embodiment includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform any of the aforementioned zero-carbon integrated energy digital management methods.

[0170] Some embodiments of this application provide a zero-carbon integrated energy digital management non-volatile computer storage medium storing computer-executable instructions, which are capable of executing any of the zero-carbon integrated energy digital management methods described above.

[0171] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device and medium embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the description of the method embodiments.

[0172] The devices and media provided in this application are one-to-one with the methods. Therefore, the devices and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0173] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0174] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0175] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0176] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0177] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0178] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0179] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0180] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0181] The above description is merely an embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the technical principles of this application should fall within the protection scope of this application.

Claims

1. A zero-carbon integrated energy digital management method, characterized in that, The method includes: Carbon calculations were performed on the multi-dimensional data of the park during the scheduling cycle to obtain a dynamic carbon flow model of the spatiotemporal distribution of carbon emissions and a dynamic carbon sink model of the spatiotemporal changes of carbon absorption. The dynamic carbon flow model and the dynamic carbon sink model are coupled and calculated on a unified spatiotemporal scale to fit and generate the net carbon spatiotemporal tensor of the park during the scheduling cycle; the net carbon spatiotemporal tensor is used to quantify the net carbon emissions of each area in the park at each time slice. Tensor decomposition is performed on the net carbon spatiotemporal tensor to select target regions that are spatially adjacent but functionally opposed. Based on the carbon balance level and time dynamics of the target region, energy supply and load demand among the target regions are regulated to generate a collaborative optimization strategy.

2. The method according to claim 1, characterized in that, The process of performing carbon calculations on multi-dimensional data of the park during the scheduling cycle to obtain the park's dynamic carbon flow model and dynamic carbon sink model specifically includes: The park's geographical area is divided into grid units, and a mapping relationship is established between each grid and the carbon source equipment and carbon sink resources within the park; Based on the operating parameters of energy demand equipment and the carbon emission factor labels of energy supply equipment, the carbon emissions of each grid in each time slice are calculated to construct the dynamic spatial distribution of carbon flow in the park. Based on the carbon sequestration efficiency factor of carbon sink resources and the adjustment factor corresponding to environmental data, the carbon absorption of each grid in each time slice is calculated to construct the dynamic spatial distribution of carbon sink in the park.

3. The method according to claim 2, characterized in that, Based on the operating parameters of energy demand equipment and the carbon emission factor labels of energy supply equipment, the carbon emissions of each grid in each time slice are calculated, specifically including: Based on the carbon emission factor of energy supply equipment, carbon accounting is performed on the operating parameters of energy demand equipment within the grid to obtain the direct carbon emissions of energy demand equipment in the current grid. Based on the distance-dependent diffusion coefficient and exponential decay function, the carbon emissions of adjacent grids in the previous time slice are compensated to obtain the carbon diffusion influence of adjacent grids. The carbon diffusion influence decreases exponentially with increasing distance and is related to the modulation effect of wind direction and topography on carbon diffusion. The carbon emissions of the current grid in the current time slice are obtained by summing the direct carbon emissions with the carbon diffusion effects of all adjacent grids.

4. The method according to claim 2, characterized in that, The dynamic carbon flow model and the dynamic carbon sink model are coupled and calculated on a unified spatiotemporal scale to fit and generate the net carbon spatiotemporal tensor of the park during the scheduling cycle, specifically including: When determining that the dynamic carbon flow model and the dynamic carbon sink model are based on the same spatial grid coordinate system and time series, net carbon emissions are calculated for each time slice and each grid cell. All net carbon emission values ​​are constructed into a three-dimensional data structure according to spatial and temporal coordinates to obtain the net carbon spatiotemporal tensor; the dimension of the net carbon spatiotemporal tensor is related to spatial latitude and longitude and time series.

5. The method according to claim 1, characterized in that, Tensor decomposition is performed on the net carbon spatiotemporal tensor, selecting target regions that are spatially adjacent but functionally opposed, specifically including: Based on the spatial distribution of dynamic carbon flow and the spatial distribution of dynamic carbon sink, a net carbon spatiotemporal tensor is constructed within the scheduling period; the net carbon spatiotemporal tensor includes a spatial grid dimension and a time series dimension. Tensor decomposition is performed on the net carbon spatiotemporal tensor to extract the target spatial emission pattern vector representing the typical spatial distribution of carbon emissions, the target spatial absorption pattern vector representing the spatial distribution of carbon absorption, and the target temporal dynamic pattern vector representing the dynamic changes of carbon. Based on the spatial relationship between the target space emission pattern vector and the target space absorption pattern vector, regions that are spatially adjacent but functionally opposed are identified to obtain the target region.

6. The method according to claim 5, characterized in that, The tensor decomposition of the net carbon spatiotemporal tensor specifically includes: When using CP decomposition, the net carbon spatiotemporal tensor is decomposed into the sum of rank-number components. Each component includes two spatial mode vectors, one temporal dynamic mode vector, and a scalar representing the component weights. When using Tucker decomposition, the net carbon spatiotemporal tensor is decomposed into a core tensor and the product of two spatial factor matrices and a time factor matrix. From the vector group or factor matrix obtained after decomposition, extract at least one spatial emission pattern vector that represents the typical spatial distribution of carbon emissions, at least one spatial absorption pattern vector that represents the spatial distribution of carbon absorption, and at least one temporal dynamic pattern vector that represents the dynamic changes of carbon. Calculate the weight of each component in the CP decomposition or the sum of squares of each element in the core tensor in the Tucker decomposition to determine the contribution of each extracted mode vector to the net carbon spatiotemporal tensor. The pattern vectors corresponding to the top-ranked number of components or core tensor elements in terms of contribution are determined as the target pattern vectors.

7. The method according to claim 5, characterized in that, The step of identifying spatially adjacent but functionally opposing regions based on the spatial relationship between the target space emission pattern vector and the target space absorption pattern vector specifically includes: By mapping the target space emission pattern vector and the target space absorption pattern vector back to two-dimensional geographic space, typical carbon source distribution maps and typical carbon sink distribution maps are obtained. Find strong carbon source regions that have a weight higher than a preset emission threshold in the emission mode and a weight lower than a preset absorption threshold in the absorption mode, and find strong carbon sink regions that have a weight higher than a preset absorption threshold in the absorption mode and a weight lower than a preset emission threshold in the emission mode. Based on spatial topological relationships, strong carbon source regions and strong carbon sink regions that meet the condition of being physically adjacent to each other are selected; spatial proximity includes geometric adjacency determination and / or distance threshold determination. Regions that simultaneously meet the conditions of functional opposition and spatial proximity are paired to identify regions that are spatially adjacent but functionally opposed.

8. The method according to claim 1, characterized in that, Based on the carbon balance level and time dynamics of the target region, energy supply and load demand among the target regions are regulated to generate a collaborative optimization strategy, specifically including: Identify the transferable loads in the carbon source areas within the target region, and adjust the operating schedule of the transferable loads according to the peak carbon absorption period indicated by the time dynamic pattern vector corresponding to the target region and / or the low carbon period indicated by the grid carbon emission factor label, so as to achieve optimal time matching between carbon emissions and carbon absorption. And / or, Based on the carbon emissions and carbon absorption of the target area, the carbon balance level of the target area is determined; when the carbon balance level is lower than the preset level, local clean energy is adjusted to meet the load demand of the carbon source area within the target area.

9. The method according to claim 1, characterized in that, Based on the carbon balance level and time dynamic pattern of the target region, after regulating the energy supply and load demand among the target regions to generate a collaborative optimization strategy, the method further includes: The aforementioned collaborative optimization strategy is simulated and reasoned in the digital twin model of the park to obtain the spatiotemporal distribution of net carbon in the target area; For each time slice, the net carbon emissions of all grids in the spatiotemporal distribution of net carbon are aggregated to obtain the comprehensive net carbon emissions of the target region. An energy management report for the target region is generated based on the comprehensive net carbon emissions and synergistic optimization strategies.

10. A zero-carbon integrated energy digital management device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the zero-carbon integrated energy digital management method according to any one of claims 1-9.

Images