Intelligent decision system based on office equipment carbon footprint data and low-carbon design

Abstract

The application relates to the field of carbon footprint data processing, and in particular to an intelligent decision-making system based on office equipment carbon footprint data and low-carbon design. The system comprises: obtaining a multi-source data set of office equipment, based on which, carbon behavior analysis is performed on the office equipment, including time and space elasticity, digital carbon flow coupling degree and modal transition characteristics, to generate a device carbon footprint fluctuation information set; obtaining a specific office business process set, based on the device carbon footprint fluctuation information set and the specific office business process set, low-carbon decision-making is performed for process carbon chain optimization and multi-objective collaboration to generate a personalized low-carbon management strategy set; based on this, a suggestion for the collaborative scheduling and control of an associated office equipment cluster is constructed to generate a low-carbon process optimization proposal. In the low-carbon design decision-making process, the application can support the global collaborative optimization of the equipment cluster in deep integration with the business process, taking into account carbon efficiency and task performance.

Description

Technical Field

[0001] This application relates to the field of carbon footprint data processing, and in particular to an intelligent decision-making system based on carbon footprint data of office equipment and low-carbon design. Background Technology

[0002] As a large number of continuously operating terminal carriers in the daily operations of enterprises and organizations, the precise monitoring, analysis and optimization of carbon emissions from office equipment has become a key link in building low-carbon office buildings, fulfilling corporate social responsibility and controlling operating costs. It is also a cutting-edge field for the deep integration of the digital economy and the green economy.

[0003] However, existing energy-saving or carbon management systems for office equipment mostly focus on static energy consumption monitoring of individual devices or simple start-stop control based on fixed strategies. They lack collaborative perception and quantitative analysis of the multi-dimensional and dynamic carbon behavior characteristics of devices in real office scenarios. This makes it difficult for existing systems to support the deep integration of device clusters with business processes and the global collaborative optimization that balances carbon efficiency and task performance. As a result, the potential for carbon emission reduction has not been fully explored, and the overall carbon management efficiency is limited. Summary of the Invention

[0004] This application provides an intelligent decision-making system based on office equipment carbon footprint data and low-carbon design to solve the above-mentioned technical problems. The system includes: A multi-source dataset of office equipment is acquired. Based on this dataset, carbon behavior analysis is performed on the office equipment, including spatiotemporal elasticity, digital carbon flow coupling degree, and modal transition characteristics, generating a set of equipment carbon footprint fluctuation information. A set of specific office business processes is acquired. Based on the set of equipment carbon footprint fluctuation information and the set of specific office business processes, low-carbon decisions are made for process carbon chain optimization and multi-objective collaboration, generating a set of personalized low-carbon management strategies. Based on the set of personalized low-carbon management strategies, suggestions for the coordinated scheduling and control of the cluster of related office equipment are constructed, generating a low-carbon process optimization proposal.

[0005] The above technical solutions enable a leap from static accounting to dynamic perception of office carbon footprint, accurately identify the main causes of emissions, and avoid the indiscriminate and homogenous management of office equipment by deeply embedding carbon management into business processes. While ensuring business efficiency, the solutions tap into the maximum emission reduction potential and ultimately output understandable and executable process-level optimization solutions, effectively promoting the implementation of low-carbon measures and helping enterprises achieve refined and normalized sustainable operations.

[0006] Optionally, the process of generating the device carbon footprint fluctuation information set includes: the office equipment multi-source dataset contains device task time-series logs, network service call records, and device power state sequences; based on the office equipment multi-source dataset, the spatiotemporal elastic entropy, digital carbon flow coupling degree, and modal transition carbon steps of the device are quantitatively analyzed; the spatiotemporal elastic entropy, the digital carbon flow coupling degree, and the modal transition carbon steps are integrated to construct a carbon behavior feature profile of the office equipment, and encapsulated to form the device carbon footprint fluctuation information set.

[0007] Through the above technical solutions, this application achieves the following: Since the generation process of the equipment carbon footprint fluctuation information set introduces spatiotemporal elastic entropy, it can identify the schedulable uncertainty of tasks in the time and logical location dimensions, thereby supporting task orchestration oriented towards carbon opportunity windows; Since the introduction of digital carbon flow coupling degree, it can reveal the cloud-based indirect carbon emissions induced by local operations and the implicit carbon loss throughout the battery's life cycle, thereby supporting cloud area optimization and service path reconstruction oriented towards digital carbon chains; Since the introduction of mode transition carbon steps, it can quantify the additional carbon costs during equipment state switching, thereby supporting mode stability control oriented towards carbon inertial balance; The three indicators work together to form a multi-dimensional dynamic characterization of the carbon behavior of office equipment, making subsequent low-carbon decisions scientific, interpretable, and feasible.

[0008] Optionally, the spatiotemporal elastic entropy analysis process includes: parsing the start time, end time, resource requirements, and task priority tag of the current task in the device task timing log; analyzing the time window length of the current task from the start time to the end time, and combining it with the task priority tag to determine whether the current task is allowed to be executed in segments or suspended midway within the time window, thus obtaining the segment suspension capability; simultaneously, based on the resource requirements, analyzing the degree of dependence of the current task on specific network nodes or computing resources during execution, and evaluating its execution feasibility at different logical locations; combining the segment suspension capability and the execution feasibility, quantifying the schedulable uncertainty of the current task in the time and logical location dimensions; and generating a quantitative value characterizing the degree to which the device task can be delayed, migrated, or interrupted, based on the schedulable uncertainty, as the spatiotemporal elastic entropy.

[0009] Through the above technical solutions, this application achieves the following: Since the generation process of spatiotemporal elastic entropy is based on the length of the time window formed by the task's start time and end time, and combines the task priority label to correct the scheduling freedom, it can avoid erroneously including high-priority rigid tasks in delayed scheduling, thus improving policy security; Since the analysis process simultaneously evaluates the degree of dependence of tasks on specific network nodes or computing resources, and judges their execution feasibility in different logical locations accordingly, it can prevent strongly bound tasks from being migrated to infeasible nodes, thus enhancing policy implementability; Since schedulable uncertainty is a comprehensive quantitative result of time dimension elasticity and logical location dimension elasticity, the generated spatiotemporal elastic entropy can accurately support the subsequent identification of carbon flexible load devices and forward-looking task orchestration, making low-carbon decisions truly adapt to the actual operational constraints of office business.

[0010] Optionally, the analysis process of the digital carbon flow coupling degree includes: parsing the network service request logs initiated by the device and the data flow destination information in the network service call record; associating real-time carbon intensity data of different destination data centers or cloud service areas, and combining the power usage efficiency values ​​of data centers in each area to calculate the cloud indirect carbon emission intensity induced by device activities after efficiency conversion; simultaneously, analyzing the charging mode and health status data of the battery associated with the device, assessing the battery life decay rate based on the number of charging cycles and charging rate, and then quantifying the degree of impact of the implicit carbon loss of the battery throughout its entire life cycle caused by device usage behavior; and combining the cloud indirect carbon emission intensity after efficiency conversion with the degree of impact of the implicit carbon loss of the battery throughout its entire life cycle to generate a quantitative value characterizing the intensity of the device's digital ecological carbon footprint, which serves as the digital carbon flow coupling degree.

[0011] Through the above technical solutions, this application achieves the following: By introducing real-time carbon intensity data from different destination data centers or cloud service areas and combining it with power usage efficiency values ​​for performance conversion, it can accurately distinguish the differentiated cloud-based indirect carbon emissions generated by the execution of the same network service in different geographical areas; By analyzing the charging mode and health status data of the battery associated with the device and assessing the battery life degradation rate based on the number of charging cycles and charging rate, it can incorporate the impact of device usage behavior on the implicit carbon loss of the battery throughout its entire life cycle into a quantitative assessment; By generating a digital carbon flow coupling degree by combining the above two influences, the system has the ability to perceive and characterize the digital ecological carbon footprint of the entire link from local office equipment to network to cloud to battery, thereby supporting differentiated low-carbon decisions oriented towards real carbon costs.

[0012] Optionally, the analysis process of the carbon step of the mode transition includes: parsing the steady-state power value sequence of the equipment under different operating modes and the power rise or fall curve data during the transient process of mode switching in the equipment power state sequence; based on the power rise or fall curve data, identifying and extracting the absolute value of power change and the transition time required to reach the new steady state during the process of switching the equipment from the first steady-state operating mode to the second steady-state operating mode; based on the absolute value of power change and the transition time, integrating and calculating the additional carbon consumption generated during the transition period of a single mode switch, and combining the mode switching frequency obtained from the statistical analysis of the equipment power state sequence within a preset period to generate a quantitative value characterizing the carbon cost of equipment state switching, as the carbon step of the mode transition.

[0013] Through the above technical solution, this application achieves the following: since the carbon step quantification of modal transitions quantifies the additional carbon cost of the equipment state switching process itself, it can identify equipment that is highly sensitive to switching behavior; since this indicator is strongly coupled with the equipment power state sequence, grid carbon intensity and statistical period, it can dynamically adapt to the actual operating characteristics of different regions, different time periods and different types of equipment; thus, in low-carbon decision-making for specific office business processes, it avoids implementing generalized energy-saving instructions for equipment with high switching carbon costs, and improves the physical rationality and engineering feasibility of the strategy.

[0014] Optionally, the process of generating the personalized low-carbon management strategy set includes: classifying the equipment into personalized decision-making devices with different carbon behavior characteristics based on the equipment carbon footprint fluctuation information set; constructing the personalized low-carbon management strategy set with the goal of minimizing the global carbon footprint based on the specific office business process set and the personalized decision-making devices; configuring a forward-looking task orchestration strategy based on carbon opportunity windows, a real-time modal stabilization strategy based on carbon inertia balance, and an end-to-end process reengineering strategy based on digital carbon chain decomposition for different personalized decision-making devices in the personalized low-carbon management strategy set.

[0015] Through the above technical solutions, this application achieves the following: By classifying equipment according to spatiotemporal elastic entropy, digital carbon flow coupling degree, and modal transition carbon steps, it is possible to distinguish between equipment types suitable for flexible scheduling and those suitable for stable operation; By configuring a forward-looking task orchestration strategy based on carbon opportunity windows for carbon flexible load equipment, high-carbon cost operations can be migrated to low-carbon periods and low-carbon areas without violating task deadlines; By configuring a real-time modal stabilization strategy based on carbon inertial balance for carbon inertial load equipment, it is possible to balance the reduction of total system carbon emissions and the minimization of equipment control costs when the grid carbon intensity changes abruptly, through the rapid response of flexible equipment and the steady-state maintenance of inertial equipment; By introducing an end-to-end process reengineering strategy based on digital carbon chain decomposition, high-carbon bottleneck nodes can be identified and optimized at the process structure level, realizing the leap from single-point optimization to system-level optimization.

[0016] Optionally, the classification process of the personalized decision-making device includes: analyzing the spatiotemporal elastic entropy and the digital carbon flow coupling degree, identifying devices that have adaptive potential to the fluctuations of carbon intensity signals in the power grid and cloud at the time and digital service levels, and defining them as carbon flexible load devices; analyzing the modal transition carbon step, identifying devices whose carbon footprint is more sensitive to its own working state switching behavior than a preset sensitivity threshold, and whose state maintenance is more carbon efficient than state switching, and defining them as carbon inertial load devices.

[0017] Through the above technical solutions, this application achieves the following: by simultaneously analyzing the spatiotemporal elastic entropy and digital carbon flow coupling degree of office equipment, it can identify equipment with carbon signal adaptive potential at the time and digital service levels, thereby providing suitable targets for forward-looking task orchestration strategies; by separately analyzing the modal transition carbon steps of office equipment and setting sensitive thresholds, it can identify equipment whose state switching carbon cost is significantly higher than the steady-state operation benefit, thereby providing suitable targets for real-time modal stability maintenance strategies; by establishing the definitions of the two types of equipment on the basis of a unified carbon behavior feature profile, it ensures that the classification results are consistent with the actual carbon behavior of the equipment, avoiding additional carbon emissions caused by strategy mismatch; and finally, it supports the accurate triggering and coordinated execution of the three types of differentiated low-carbon strategies in Examples 8 to 10.

[0018] Optionally, the forward-looking task orchestration strategy based on carbon opportunity windows includes: for the carbon flexible load device, obtaining its tasks to be executed and their associated spatiotemporal elastic entropy, and constructing a dynamic carbon price surface that integrates future power grid carbon intensity prediction data and target cloud service area carbon intensity data; analyzing the resource requirements and digital service call paths of the tasks to be executed by the carbon flexible load device, and mapping them to several carbon cost requirement packages on the dynamic carbon price surface; for each carbon cost requirement package, under the premise of satisfying its task deadline and resource constraints, finding the future execution time point and digital resource call path with the minimum carbon cost integral on the dynamic carbon price surface, and generating a task orchestration strategy item containing specific time offset instructions and cloud area selection instructions.

[0019] Through the above technical solutions, this application achieves the following: by constructing a dynamic carbon price surface that integrates the carbon intensity of the power grid and the cloud for the carbon flexible load device, discrete carbon intensity data can be transformed into a continuous and optimizable spatiotemporal decision space; by mapping task resource requirements and digital service call paths to carbon cost requirement packages, complex business actions can be decomposed into quantifiable, comparable, and schedulable basic units; by performing a carbon cost integral minimization search under the premise of meeting deadlines and resource constraints, the execution scheme with the lowest global carbon cost can be accurately locked while ensuring business determinism; thus, without changing the original office process structure and terminal hardware configuration, the low-carbon operation efficiency of the office equipment cluster in real business scenarios can be significantly improved.

[0020] Optionally, the real-time modal stabilization strategy based on carbon inertia balance includes: real-time monitoring of fluctuations in the grid carbon intensity signal; when a short-term sudden change in the grid carbon intensity signal is detected, prioritizing the adjustment of the operating power of the carbon flexible load device or immediately starting or stopping its interruptible tasks to respond to changes in the carbon signal and offset overall carbon cost fluctuations; simultaneously, maintaining the current operating mode of the carbon inertia load device to avoid significant carbon steps in the mode transition caused by its state switching; through the active adjustment of the carbon flexible load device and the state maintenance of the carbon inertia load device, achieving overall carbon inertia balance of the system under the premise of meeting business process requirements, and generating a corresponding real-time power adjustment and state locking instruction set.

[0021] Through the above technical solutions, this application achieves the following: by monitoring the grid carbon intensity signal in real time and setting a sudden change identification mechanism, it can accurately capture the timing of carbon signal disturbances; by implementing power regulation or task start-up and shutdown only for carbon flexible load equipment, it can quickly respond to carbon fluctuations without disrupting business continuity; by implementing mode locking rather than forced load reduction for carbon inertial load equipment, it avoids the carbon step amplification effect caused by frequent state switching; by unifying the response actions of the two types of equipment into the carbon inertial balance optimization framework and generating a structured instruction set, it achieves a paradigm upgrade from single-point energy saving to system-level carbon response coordination, effectively supporting the functional definition and protection boundary of carbon inertial load equipment as described in Example 7.

[0022] Optionally, the end-to-end process reengineering strategy based on digital carbon chain decomposition includes: parsing the specific office business process set, modeling all device activities and digital service calls triggered sequentially or in parallel by the completed business processes as a digital carbon chain with node and edge relationships; assigning corresponding carbon behavior characteristic data to each node in the digital carbon chain based on the device carbon footprint fluctuation information set; simulating the total carbon footprint of the digital carbon chain under different node execution sequences and different service call paths, identifying the key bottleneck nodes with the largest carbon emission contribution; for the key bottleneck nodes, if they belong to the carbon flexible load device, optimizing their task orchestration and digital service call path; if they belong to the carbon inertial load device, optimizing their upstream task sorting to reduce unnecessary mode switching.

[0023] Through the above technical solutions, this application achieves the following: by modeling office business processes as digital carbon chains that are assignable, simulable, and comparable, it is possible to penetrate the surface power consumption of the equipment itself and identify structural high-carbon links caused by the coupling of service call paths and execution timing; by binding device-level carbon behavior characteristics to each node and classifying and implementing policies accordingly, it is possible to avoid one-size-fits-all process transformation, implement path migration optimization for carbon flexible load devices, and implement upstream cycle control for carbon inertial load devices, thereby minimizing the process-level carbon footprint while ensuring business continuity. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram illustrating an application scenario provided in one embodiment of this application; Figure 2A flowchart of an intelligent decision-making system based on office equipment carbon footprint data and low-carbon design, provided as an embodiment of this application. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0027] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0028] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0029] Existing energy-saving or carbon management systems for office equipment mostly focus on static energy consumption monitoring of individual devices or simple start-stop control based on fixed strategies. They lack collaborative perception and quantitative analysis of the multi-dimensional and dynamic carbon behavior characteristics of devices in real office scenarios. This makes it difficult for existing systems to support the deep integration of device clusters with business processes and the global collaborative optimization that balances carbon efficiency and task performance. As a result, the potential for carbon emission reduction has not been fully explored, and the overall carbon management efficiency is limited.

[0030] Based on this, this application provides an intelligent decision-making system based on office equipment carbon footprint data and low-carbon design. First, it collects multi-source data on office equipment tasks, network status, and power status. Through innovative carbon behavior analysis, it quantifies the equipment's spatiotemporal elasticity, digital carbon flow coupling degree, and modal transition characteristics, generating a dynamic profile of equipment carbon footprint fluctuations. Then, the system combines these carbon behavior characteristics with specific office business processes, modeling the processes as a "digital carbon chain." Under the premise of meeting business constraints, it performs global collaborative optimization, generating personalized low-carbon strategies that integrate forward-looking orchestration, real-time stabilization, and process reengineering. Finally, the system integrates these strategies into actionable collaborative scheduling suggestions, outputting a structured low-carbon process optimization proposal, which is then distributed to company staff.

[0031] Figure 1 This is a schematic diagram illustrating an application scenario provided by this application. In the low-carbon design decision-making process, the application provided by this application can support the global collaborative optimization of equipment clusters, deeply integrating them with business processes while balancing carbon efficiency and task performance.

[0032] Specifically, this application is applied to any server that communicates with the office equipment operating system and the company project management tool, and obtains the multi-source dataset of office equipment provided by the office equipment operating system and the specific set of office business processes provided by the company project management tool through the server.

[0033] The specific implementation method can be referred to in the following embodiments, wherein the data mentioned in the embodiments are only for reference and examples, so that relevant personnel can better understand them.

[0034] Figure 2 This is a flowchart illustrating an embodiment of an intelligent decision-making system based on office equipment carbon footprint data and low-carbon design, which can be applied to servers in the above scenarios. Figure 2 As shown, the system includes: S201. Obtain multi-source datasets of office equipment. Based on the multi-source datasets of office equipment, perform carbon behavior analysis on office equipment, including spatiotemporal elasticity, digital carbon flow coupling degree and modal transition characteristics, and generate a set of equipment carbon footprint fluctuation information.

[0035] The multi-source dataset for office equipment can be a heterogeneous data set comprehensively describing the operational activities of office equipment, consisting of equipment task time-series logs, network service call records, and equipment power state sequences, originating from the office equipment operating system. Spatiotemporal elastic carbon behavior analysis can be an analytical method that quantifies the degree of delayability, portability, or interruptibility of equipment tasks in the temporal and logical location dimensions by analyzing task time windows, resource dependencies, and interruption probabilities. Digital carbon flow coupling degree carbon behavior analysis can be an analytical method that comprehensively evaluates the intensity of indirect carbon footprint induced by equipment activities in its digital service ecosystem by correlating cloud service carbon emissions with the implicit carbon loss of equipment batteries. Modal transition characteristic carbon behavior analysis can be an analytical method that quantifies the additional carbon consumption caused by state switching by calculating the transient power changes and frequency when equipment switches between different operating modes. The equipment carbon footprint fluctuation information set can be a structured data set that encapsulates dynamic carbon behavior characteristic parameters such as spatiotemporal elastic entropy, digital carbon flow coupling degree, and modal transition carbon steps obtained after quantitative analysis for each device.

[0036] Specifically, carbon emissions from modern office equipment are no longer merely the direct energy consumption of its operation, but rather a complex behavior deeply integrated into digital workflows, exhibiting dynamic and interconnected characteristics. Traditional management solutions typically rely solely on simple power meter readings, statically linking carbon emissions to the grid's average carbon intensity. This approach has significant blind spots: it completely ignores the substantial indirect carbon emissions (i.e., "digital carbon flows") induced by equipment tasks via the internet and occurring in remote cloud data centers, fails to identify the schedulable capabilities of equipment tasks themselves in time and space (i.e., "spatiotemporal elasticity"), and fails to quantify the additional carbon costs arising from power transients when equipment switches between different operating modes (i.e., "modal transitions"). This results in existing carbon footprint assessments being both one-sided and static, unable to provide an effective basis for refined management. This solution addresses this issue by aggregating heterogeneous data from multiple sources, including equipment task logs, network call records, and power status sequences, to construct a novel "carbon behavior analysis" framework. It deconstructs the carbon emission patterns of equipment from three dynamic dimensions: spatiotemporal elasticity, digital carbon flow coupling degree, and modal transition characteristics. This generates an "equipment carbon footprint fluctuation information set" that accurately reflects the fluctuation patterns of its carbon footprint, laying a core data foundation for subsequent differentiated and precise low-carbon regulation.

[0037] S202. Obtain a set of specific office business processes. Based on the set of equipment carbon footprint fluctuation information and the set of specific office business processes, make low-carbon decisions oriented towards process carbon chain optimization and multi-objective collaboration, and generate a set of personalized low-carbon management strategies.

[0038] Specific office business process sets can be standardized sets of office activities that define business objectives, step sequences, equipment dependencies, and constraints (such as deadlines), derived from company project management tools. Process-oriented carbon chain optimization can be an optimization approach that models all equipment and digital service calls involved in completing a specific business process as a "carbon chain," and minimizes the overall carbon footprint of the process by adjusting the execution strategies of nodes on the chain. Multi-objective collaborative low-carbon decision-making can be a collaborative decision-making process that seeks the optimal solution for the system's global carbon footprint while satisfying multiple constraints such as time, resources, and reliability in business processes. Personalized low-carbon management strategy sets can be sets of instructions customized based on the carbon behavior characteristics of equipment and its role in business processes, including differentiated strategies such as forward-looking task orchestration, real-time modal stabilization, and end-to-end process reengineering.

[0039] Specifically, office carbon emissions are embedded in daily business processes. Any effective low-carbon management must be deeply integrated with business logic; otherwise, it will become a mere indiscriminate power restriction measure that interferes with business efficiency. The fundamental flaw of existing extensive energy-saving strategies (such as uniformly setting equipment hibernation times) lies in the failure to implement differentiated management based on business context. Their uniform execution rules are prone to conflict with critical tasks, or inadvertently increase carbon emissions in other areas while pursuing energy savings in specific equipment, failing to achieve overall optimization. For example, forcing a device about to begin critical computing to hibernate may delay the entire project process, resulting in more wasted resources. This solution addresses this problem by introducing a "specific office business process set," reweaving the discrete carbon behavior characteristics of equipment (equipment carbon footprint fluctuation information set) into the business chain that generates value. The system models each business process as a "digital carbon chain" and implements "low-carbon decision-making oriented towards process carbon chain optimization and multi-objective collaboration." This means that the decision-making objective is to comprehensively optimize the execution strategies of all device nodes on the chain under the premise of meeting multiple constraints such as business process time and resources, and to seek to minimize the global carbon footprint, thereby generating a "personalized low-carbon management strategy set" that both fits the physical characteristics of the equipment and ensures smooth business operations.

[0040] S203. Based on a set of personalized low-carbon management strategies, construct suggestions for the coordinated scheduling and control of clusters of related office equipment, and generate a low-carbon process optimization proposal.

[0041] An office equipment cluster can be a group of interconnected office devices that work collaboratively to achieve a common business goal in a specific office setting or business process. A low-carbon process optimization proposal can be a structured document or set of digital instructions that integrates collaborative scheduling instructions, carbon emission reduction projections, business impact assessments, and actionable guidelines.

[0042] Specifically, even if an optimization strategy possesses significant emission reduction potential and technical rationale, it will be difficult to achieve actual carbon emission reduction results if a standardized and implementable execution plan cannot be formed. Currently, many intelligent management solutions output scattered instructions or complex data models, lacking integration and interpretation at the execution layer. This leads to difficulties for IT administrators or employees in understanding and collaborating on the strategy, or concerns about policy conflicts impacting business operations, ultimately resulting in the solution being shelved. This solution addresses this challenge in the strategy implementation phase by designing an engineering encapsulation and delivery process for the strategy. Based on the potentially scattered and time-dependent "personalized low-carbon management strategy set" generated in the previous stage, the system resolves conflicts and performs global orchestration, constructing a set of "coordinated scheduling and control recommendations" that are consistent in terms of time, resources, and objectives. Finally, the system integrates these recommendations to generate a structured "low-carbon process optimization recommendation document." The proposal not only includes specific operational instructions for each device (such as "postponing the rendering task of device A to 14:00 and specifying the use of cloud region B"), but also clarifies the overall carbon benefits after optimization, the impact assessment on the business timeline, and the key points of implementation from a process perspective, thereby transforming the intelligent decision-making in the background into an understandable, trustworthy, and executable operational guide or automated work order in the front end.

[0043] The method provided in this embodiment first collects multi-source data such as office equipment task, network, and power status. Through innovative carbon behavior analysis, it quantifies the spatiotemporal elasticity, digital carbon flow coupling degree, and modal transition characteristics to generate a dynamic profile of equipment carbon footprint fluctuations. Subsequently, the system combines these carbon behavior characteristics with specific office business processes, modeling the processes as a "digital carbon chain." Under the premise of meeting business constraints, it performs global collaborative optimization to generate personalized low-carbon strategies that integrate forward-looking orchestration, real-time stabilization, and process reengineering. Finally, the system integrates the strategies into actionable collaborative scheduling suggestions, outputs a structured low-carbon process optimization proposal, and distributes it to company staff. This solution achieves a leap from static accounting to dynamic perception of office carbon footprint, accurately identifies the main causes of emissions, and avoids "one-size-fits-all" control by deeply embedding carbon management into business processes. It maximizes emission reduction potential while ensuring business efficiency, ultimately outputting understandable and executable process-level optimization solutions, effectively promoting the implementation of low-carbon measures, and helping enterprises achieve refined and normalized sustainable operations.

[0044] In some embodiments, the multi-source dataset of office equipment includes equipment task time-series logs, network service call records, and equipment power state sequences. Based on the multi-source dataset of office equipment, the spatiotemporal elastic entropy, digital carbon flow coupling degree, and modal transition carbon steps of the equipment are quantitatively analyzed. The spatiotemporal elastic entropy, digital carbon flow coupling degree, and modal transition carbon steps are integrated to construct a carbon behavior feature profile of the office equipment and encapsulate it into a set of equipment carbon footprint fluctuation information.

[0045] Device task timing logs can be datasets that record precise timestamps and task tags for task creation, start, suspension, resumption, and completion. Network service call records can be datasets that record the target domain name / IP, port, packet size, and timestamp of HTTP / HTTPS requests initiated by the device. Device power state sequences can be real-time power consumption values ​​of the device recorded at fixed sampling intervals (e.g., 1 second) and their corresponding power state tags (e.g., sleep, idle, high load). Spatiotemporal elasticity entropy can be an indicator that quantifies the degree of uncertainty in the scheduling (e.g., delay, migration, interruption) of device tasks in the temporal and logical spatial dimensions by analyzing task timing logs; the higher the value, the greater the flexibility in the execution time and location of the task, provided the deadline is met. Digital carbon flow coupling degree can be an indicator that quantifies the intensity of indirect and implicit carbon footprint induced by device digital activities by correlating network service calls with cloud carbon emission data and device battery loss models; the higher the value, the greater the contribution of device activities to cloud carbon emissions and implicit battery carbon loss. Modal transition carbon step can be an indicator that quantifies the additional carbon cost of a single operating state switch (such as from sleep to high load) by analyzing the transient process of mode switching in the power state sequence of a device; the higher the value, the greater the carbon penalty brought about by the device's state switching behavior itself. Carbon behavior profile can be a structured data model that encapsulates three scalar values ​​of a device—"spatiotemporal elastic entropy," "digital carbon flow coupling degree," and "modal transition carbon step"—along with some metadata used as the basis for its calculation (such as the maximum window of allowable task latency and major cloud service providers) in a unified JSON or Protobuf format data object, forming a digital description of the device's dynamic carbon footprint characteristics.

[0046] As an optional embodiment, the specific implementation of this application's solution is as follows: A unit deploys office computers, all-in-one printers, scanning terminals, conference tablets, and several mobile office devices. The system periodically collects task timing logs, cloud call records, and power status sequences from each device, and obtains real-time carbon intensity data of the power grid and carbon intensity data of different cloud service regions from the carbon intensity service interface. For mobile terminals, battery cycle count, charging rate, and health status are further collected. Within a certain workday, the system identifies three typical tasks: end-of-day batch archiving and uploading tasks (high spatiotemporal elastic entropy, high digital carbon flow coupling), pre-meeting batch printing tasks (high modal transition carbon order), and real-time screen projection and remote collaboration tasks during meetings (low spatiotemporal elastic entropy). Based on this, the system calculates three indicators: for archive upload tasks, a high spatiotemporal elastic entropy value is assigned based on their characteristics of allowing a two-hour delay and switching cloud regions; for their frequent calls to OCR and cloud storage, a high digital carbon flow coupling value is assigned based on the fact that the carbon intensity of cloud services in South China is lower than that in East China; for all-in-one printers, based on actual measured data of a preheating power of 800W and a transition time of 12s, the carbon step of a single switch is calculated to be 0.042kgCO2e, and the high-modal transition carbon step value is obtained by superimposing the average switching frequency of the day; finally, the three indicators are integrated with fields such as equipment identification, process, and typical low-carbon window, encapsulated into a set of equipment carbon footprint fluctuation information, and output to the personalized low-carbon management strategy generation module.

[0047] In some embodiments, the start time, end time, resource requirements, and task priority tag of the current task in the device task timing log are parsed; the length of the time window from the start time to the end time of the current task is analyzed, and combined with the task priority tag, it is determined whether the current task is allowed to be executed in segments or suspended midway within the time window, thus obtaining the segment suspension capability; at the same time, based on the resource requirements, the degree of dependence of the current task on specific network nodes or computing resources during execution is analyzed, and its execution feasibility at different logical locations is evaluated; combining the segment suspension capability and execution feasibility, the schedulable uncertainty of the current task in the time and logical location dimensions is quantitatively calculated; based on the schedulable uncertainty, a quantitative value characterizing the degree to which the device task can be delayed, migrated, or interrupted is generated as the spatiotemporal elastic entropy.

[0048] Resource requirements can be a clearly defined list of specific hardware devices, software licenses, dedicated data, or network services that the task must rely on for execution. Task priority tags can be discrete classification identifiers assigned to tasks by the system or users to indicate their relative urgency and importance, typically "Critical," "High," "Medium," or "Low." Time window length can be the total time span during which the task is allowed to be executed, calculated by the difference between the task's "deadline" and "start time," usually in seconds or hours. Segmented suspension capability can be a Boolean attribute (yes / no) determining whether the task is allowed to be temporarily interrupted (suspended) by the operating system or management system during execution, and resumed at a later time, based on the task's "task priority tag" and a pre-defined business continuity rule base. Different logical locations can be alternative locations at the physical or virtual resource level where the task can be scheduled for execution. This includes not only different physical devices (such as workstation A or workstation B), but also different virtual machine instances, different cloud service availability zones, or different computing cluster nodes. Scheduling uncertainty can be obtained by combining the "segmentation suspension capability" (temporal dimension flexibility) and "execution feasibility" (spatial dimension flexibility) of a comprehensive task, and then fusion calculation through a preset algorithm (such as weighted scoring, fuzzy inference) to obtain an overall potential metric for the task to be adjusted in time and space.

[0049] As an optional embodiment, the specific implementation of the solution in this application is as follows: Taking the daily batch archiving and uploading task of electronic vouchers by the finance department of a certain unit as an example, the task is initiated by an office computer. The task log records that its start time is 17:00 on the same day (the system is idle before the end of the workday), and the end time is 8:00 the next day (before the start of the workday), with a time window length of 15 hours; the task priority label is medium, with no mandatory real-time requirement; the resource requirement is to call the local PDF generation library + upload to the East China Cloud regional object storage, without relying on dedicated peripherals; after system parsing, it is determined that its segmented suspension capability is strong (supports breakpoint resume, and the checkpoint mechanism has been enabled), and the execution feasibility is high (the East China Cloud regional service interface is stable, the uploaded data has been anonymized, and it complies with the enterprise cloud strategy); after weighted fusion calculation: the segmented suspension capability is assigned a value of 0.9, the execution feasibility is assigned a value of 0.85, and the weighted fusion calculation is performed with a weight of 0.6:0.4, i.e., the spatiotemporal elastic entropy = 0.9×0.6+0.85×0.4=0.54+0.34=0.87, resulting in a spatiotemporal elastic entropy value of 0.87 (weight explanation: in office equipment task scheduling, the ability to suspend tasks in a segmented time dimension has a more significant impact on the flexibility of carbon emission reduction scheduling, and is the core determining factor for whether tasks can be delayed or interrupted, therefore it is assigned a weight of 0.6; the feasibility of execution at the logical location is an auxiliary constraint condition, which only affects the selection of the scheduling range, therefore it is assigned a weight of 0.4.). This high entropy value triggers the system to include it in the carbon opportunity window matching process, and combined with the prediction of the grid carbon intensity for the next 24 hours and the comparison of the carbon intensity in the East China / South China cloud region, it is finally recommended to postpone the upload action to 2:00 am the next day, and to route the OCR recognition subtask to the South China cloud region with lower carbon intensity, thereby reducing the overall digital carbon flow coupling degree while ensuring the completion of business.

[0050] In some embodiments, the network service request logs initiated by the device and the destination information of the data stream in the network service call record are parsed; real-time carbon intensity data of different destination data centers or cloud service regions are correlated, and combined with the power usage efficiency values ​​of data centers in each region, the cloud indirect carbon emission intensity induced by device activities and converted to performance is calculated; at the same time, the charging mode and health status data of the battery associated with the device are analyzed, and the battery life decay rate is evaluated based on the number of charging cycles and the charging rate, thereby quantifying the degree of impact of the implicit carbon loss of the battery throughout its entire life cycle caused by device usage behavior; by combining the cloud indirect carbon emission intensity converted to performance and the degree of impact of the implicit carbon loss of the battery throughout its entire life cycle, a quantitative value characterizing the intensity of the device's digital ecosystem carbon footprint is generated as the digital carbon flow coupling degree.

[0051] Network service request logs can be time-series data entries recording network communication activities initiated by devices, originating from log files of enterprise firewalls, network proxy servers, or operating system network stacks. Data flow destination information can be the physical location or logical service area ultimately reached by the network request, determined by domain name resolution, IP geolocation lookup, or cloud service provider metadata APIs, based on the target address in the network service request logs. Real-time carbon intensity data can be a dynamic value reflecting the carbon dioxide equivalent (gCO2eq / kWh) generated per unit of power generation in a specific power grid area. Power utilization efficiency can be an indicator of data center energy efficiency, calculated as the ratio of total data center energy consumption (IT equipment energy consumption + cooling, lighting, etc.) to IT equipment energy consumption. Cloud indirect carbon emission intensity can be a quantified value of carbon emissions generated by cloud data centers due to network services initiated by devices, derived through a calculation model. Battery life degradation rate can be a percentage or absolute value predicted by analyzing historical battery usage data (such as charging cycle count, charge / discharge rate, and operating temperature) based on a battery chemical aging model, representing the percentage or absolute value of the decrease in maximum usable battery capacity per unit of usage (e.g., per cycle). The impact of implicit carbon loss throughout the battery's life cycle can be quantified by allocating the responsibility for the amount of lifespan degradation caused by battery usage behavior to the total carbon emissions (i.e., implicit carbon) generated during the battery's manufacturing stage according to the lifespan loss ratio, thereby determining the implicit carbon responsibility value (e.g., kgCO2) attributable to the current equipment usage behavior.

[0052] As an optional embodiment, the specific implementation of this application is as follows: A mobile office laptop deployed by an enterprise centrally calls cloud document collaboration services (including real-time saving, version comparison, and AI summary generation) from 9:00 AM to 11:00 AM daily. Its network service call records show that the target area is the East China cloud node, with a corresponding real-time carbon intensity of 560 gCO2e / kWh and a PUE of 1.38. After conversion, the cloud-based indirect carbon emission intensity is 773 gCO2e / kWh. At the same time, the device adopts a 2C fast charging strategy, with an average cycle count of 2.4 times / day in the past 30 days, and its health has dropped to 82%. Based on this, the impact of its battery's implicit carbon loss over its entire life cycle is assessed as equivalent to 45 kWh of grid carbon emissions. The system integrates the above two indicators with a weight of 0.7:0.3 (weight explanation: cloud-based indirect carbon emissions are the main component of the digital ecosystem carbon footprint of office equipment, with a high proportion and large fluctuations, playing a dominant role in the overall carbon emission results, therefore assigned a weight of 0.7). Weighting; the implicit carbon loss throughout the battery's entire lifecycle is a hidden emission throughout the device's entire lifecycle, with a relatively small impact from a single use, and is therefore a secondary accounting item, thus assigned a weight of 0.3. That is, the digital carbon flow coupling degree = 773 × 0.7 + 45 × 0.3 = 541.1 + 13.5 = 554.6. Combining this with the regional carbon emission correction factor of 1.22, 554.6 × 1.22 ≈ 678, generating a digital carbon flow coupling degree of 678 gCO2e / kWh. Given that this value is higher than the average of devices in the same batch (412 gCO2e / kWh), the system identifies it as a high digital carbon flow coupling device and prioritizes matching it to low-carbon cloud regions (such as the Southwest Node, with a carbon intensity of 110 gCO2e / kWh and a PUE of 1.21, which is equivalent to 133 gCO2e / kWh) in subsequent strategy generation. Simultaneously, it recommends that users adjust their charging strategy to below 1C to slow down battery degradation.

[0053] In some embodiments, the steady-state power value sequence of the device under different operating modes and the power ramp-up or ramp-down curve data during the transient process of mode switching are analyzed in the device power state sequence. Based on the power ramp-up or ramp-down curve data, the absolute value of power change and the transition time required to reach the new steady state during the process of switching the device from the first steady-state operating mode to the second steady-state operating mode are identified and extracted. Based on the absolute value of power change and the transition time, the additional carbon consumption generated during the transition period of a single mode switch is calculated by integration. Combined with the mode switching frequency obtained by statistical analysis of the device power state sequence within a preset period, a quantitative value characterizing the carbon cost of device state switching is generated as a carbon step of mode transition.

[0054] A steady-state power value sequence can be a subsequence of continuous data points whose power readings fluctuate within a small range (e.g., less than 5% of the nominal value) during a period when the equipment is identified as being in a certain stable operating mode (e.g., high load). A mode-switching transient process can be the brief period during which the equipment transitions from one steady-state operating mode to another, during which the power readings exhibit a non-steady-state "climb" or "decline" curve. Power climb or decline curve data can refer to continuous sampling point data describing the power change over time, extracted from the "mode-switching transient process." The absolute value of the power change can be the absolute value of the change in equipment power from the initial steady-state value to the target steady-state value during a complete mode-switching transient process, calculated using the formula: |Target steady-state power - Initial steady-state power|. Additional carbon consumption can be the carbon dioxide equivalent emission value obtained by calculating the excess energy consumption by numerically integrating the deviation curve over the transition time during the transient process of equipment operating mode switching, due to the inability of physical systems (such as circuits, motors, hard drives) to achieve instantaneous response.

[0055] As an optional embodiment, the specific implementation of this application is as follows: Parsing the power state sequence: For example, the module reads the power sequence [5.1, 5.0, 120, 410, 488, 490, 489, 5.2, ...] and corresponding mode labels [sleep, sleep, wake up, ready, printing, printing, printing, sleep, ...] sampled at a frequency of 1Hz from a laser printer over the past hour. Next, the transient curve is identified and extracted: The algorithm detects the time point when the label changes from sleep to wake up, and expands the window forward and backward until the power change rate is below a threshold, thereby extracting the complete power ramp-up curve data from sleep (e.g., 5W) to printing (e.g., 490W), containing approximately 15 sampling points over a period of 15 seconds. The algorithm extracts the absolute value of the power change during this switch as 485W, with a transition time of 15 seconds. Subsequently, the system calculates the additional carbon consumption for a single switch: it numerically integrates the rising curve, calculates the total area under the curve, and then subtracts the area corresponding to an ideal instantaneous switch (such as a horizontal line that jumps instantaneously from 5W to 490W), resulting in an additional energy consumption of 1250 joules (approximately 0.000347 kWh). With the grid carbon intensity at this time being 0.4 kgCO2 / kWh, the additional carbon consumption for this switch is 0.000139 kgCO2. Next, the system statistically analyzes the mode switching frequency: the module scans the sequence of the past 24 hours and counts 40 instances of the same "sleep-printing-sleep" complete cycle. Finally, the system generates a mode transition carbon step: multiplying the additional carbon consumption for a single switch (e.g., 0.000139 kgCO2) by the frequency (e.g., 40) yields the total additional carbon cost for the daily switch, e.g., 0.00556 kgCO2. Comparing this value with the equipment's total carbon emissions over 24 hours (e.g., 0.5 kg CO2), and normalizing it (normalized value = 0.00556 ÷ 0.5 = 0.01112), rounded to three decimal places, the final modal transition carbon step quantization value for the printer is 0.011 (indicating that approximately 1.1% of its carbon footprint is contributed by state switching). The equipment's total carbon emissions over 24 hours are derived by multiplying the 24-hour steady-state total energy consumption by the grid's average carbon intensity. The steady-state total energy consumption is obtained by collecting the equipment's steady-state power from power sensors, combining this with the cumulative duration of each state monitored and calculated, and then summing the products. The grid's average carbon intensity is obtained by acquiring real-time and predicted carbon emission intensity data from the regional power grid through a grid carbon intensity monitoring interface, and then calculating the average over a periodic period. For example, the equipment's steady-state total energy consumption is about 2.06 kWh in 24 hours. Based on the average carbon intensity of the power grid of 0.243 kg CO2 / kWh, 2.06 × 0.243 ≈ 0.5 kg CO2.

[0056] In some embodiments, based on the device carbon footprint fluctuation information set, devices are classified into personalized decision-making devices with different carbon behavior characteristics; based on the specific office business process set and personalized decision-making devices, a personalized low-carbon management strategy set aimed at minimizing the global carbon footprint is constructed: in the personalized low-carbon management strategy set, a forward-looking task orchestration strategy based on carbon opportunity window, a real-time modal stabilization strategy based on carbon inertia balance, and an end-to-end process reengineering strategy based on digital carbon chain decomposition are configured for different personalized decision-making devices respectively.

[0057] Personalized decision-making devices can be instances of device categories classified into different combinations of carbon behavior characteristics according to preset classification rules (such as threshold comparison and clustering algorithms). For example, they can be classified as "carbon flexible load devices" or "carbon inertial load devices." A collaborative decision-making model can be a mathematical optimization model with "minimizing the global carbon footprint" as the objective function, and the task dependencies, completion deadlines, and carbon behavior characteristics (such as elasticity entropy and coupling degree) of the "specific office business process set" as constraints. A forward-looking task orchestration strategy can be a type of strategy designed for "carbon flexible load devices." Its core idea is to leverage the flexibility of these devices at the time and digital service levels to postpone or advance their tasks to future periods (carbon opportunity windows) when the carbon intensity of the grid or cloud is lower. A real-time modal stabilization strategy can be a type of real-time response strategy designed specifically for "carbon inertial load devices" and executed in coordination with "carbon flexible load devices." Its core is to prioritize scheduling flexible devices for power regulation to "hedge" carbon cost fluctuations when the grid carbon intensity fluctuates, while instructing inertial devices to maintain their current state to avoid switching carbon costs. End-to-end process reengineering strategies can be a type of structural optimization strategy that starts from the perspective of the entire office business process and identifies and reconstructs high-carbon links by modeling, simulating and optimizing the "digital carbon chain" consisting of all equipment activities and digital service calls in the process.

[0058] As an optional embodiment, the specific implementation of this application is as follows: A government agency deploys 20 office computers, 3 high-speed multifunction printers, 5 scanning terminals, 2 conference tablets, and several mobile office terminals, connected to a local document management system, East China regional cloud OCR service, and South China regional archiving cloud platform. The system collects daily task time-series logs, network service call records, and power status sequences from each device, and obtains real-time carbon intensity data from the State Grid carbon monitoring platform and cloud service provider APIs. Analysis identifies 8 office computers and 3 scanning terminals as carbon flexible load devices (spatiotemporal elastic entropy ≥ 0.75, digital carbon flow coupling degree ≥ 0.68), and 2 high-speed multifunction printers and 1 conference tablet as carbon inertial load devices (modal transition carbon step ≥ 1.2 kgCO2e / cycle). During the daily expense reimbursement process, the system binds the expense form scanning node to a proactive task scheduling strategy, recommending that scanning tasks be postponed to 22:00-23:00 and that OCR recognition requests be routed to the South China Low-Carbon Cloud Zone. The system also binds the voucher printing node to a real-time modal stabilization strategy, maintaining its continuous operation for at least 45 minutes to avoid multiple warm-ups. The upload-archiving process is modeled as a bottleneck node in the digital carbon chain, identifying redundant cloud calls caused by repeated uploads and recommending that original vouchers and approval opinions be merged into a single encrypted upload package, reducing cloud data transmission volume by 37%. The resulting personalized low-carbon management strategy set is then pushed to the IT operations platform, which automatically issues execution instructions according to the strategy items.

[0059] In some embodiments, the spatiotemporal elastic entropy and digital carbon flow coupling degree are analyzed to identify devices that have the potential to adapt to fluctuations in carbon intensity signals in the power grid and cloud at the time and digital service levels, and these devices are defined as carbon flexible load devices; the mode transition carbon step is analyzed to identify devices whose carbon footprint is more sensitive to its own working state switching behavior than a preset sensitivity threshold, and whose state maintenance is more carbon efficient than state switching, and these devices are defined as carbon inertial load devices.

[0060] Carbon flexible load devices are a core term coined in this solution, referring to a specific type of "personalized decision-making device." The definition is based on a comprehensive evaluation of two indicators: "spatiotemporal elastic entropy" and "digital carbon flow coupling degree." Specifically, high spatiotemporal elastic entropy (e.g., greater than a threshold of 0.6) indicates that the device's tasks can be flexibly scheduled in time and logically migrated; high digital carbon flow coupling degree (e.g., greater than a threshold of 0.5) indicates that the device's activities are closely related to cloud carbon intensity and can be optimized by selecting low-carbon regional services. Devices that simultaneously or primarily meet these two characteristics are considered to be able to adaptively adjust to fluctuations in carbon intensity in both the power grid (time dimension) and the cloud (service dimension), hence the definition of carbon flexible load devices. Carbon inertial load equipment can be another core term created in this solution, referring to another specific type of "personalized decision-making equipment". Its definition is mainly based on the evaluation of the "modal transition carbon step" index. The preset sensitivity threshold (such as 0.1) is a threshold value set through historical data analysis or expert experience. When the modal transition carbon step of the equipment is higher than this threshold, it indicates that the additional carbon cost generated by its state switching is significant. Then, by comparing its "carbon cost corresponding to state maintenance energy consumption" with "carbon cost of state switching + carbon cost of new state maintenance", it is determined that state maintenance is more carbon efficient than state switching. Equipment that meets this condition has a carbon footprint that is highly sensitive to state switching behavior and tends to maintain the current state to reduce the overall carbon cost. Therefore, it is defined as carbon inertial load equipment.

[0061] As an optional embodiment, the specific implementation of the solution in this application is as follows: An administrative service center is equipped with 20 office computers, 3 high-speed A3 all-in-one machines, 5 high-speed document scanners, 2 sets of conference collaboration tablets, and several mobile government affairs tablets. The system periodically collects task time sequence logs (including task name, allowed start time, end time, priority, and interruptible flag), network service call records (including OCR call frequency, cloud storage destination, and data volume), and power status sequences (including steady-state power and switching transient curves in standby / activation / printing / scanning / conference modes) from each device. Analysis revealed that a mobile government affairs Pad had a spatiotemporal elasticity entropy of 0.86 and a digital carbon flow coupling degree of 0.79, thus identifying it as a carbon flexible load device. A high-speed A3 all-in-one printer, however, had a modal transition carbon level of 142 Wh, exceeding the preset threshold of 120 Wh. Furthermore, its current steady-state carbon emission rate in printing mode was 180 gCO2 / h. Switching to sleep mode and then waking it up would incur an additional 135 gCO2 carbon cost, therefore it was identified as a carbon inertial load device. Based on this, when the grid carbon intensity suddenly increased to 680 gCO2 / kWh at 10:00 AM that day, the system immediately suspended the non-urgent cloud document synchronization task initiated by the Pad and locked the all-in-one printer in its current printing mode. This prevented unnecessary start-stop operations in response to control commands, thereby achieving minute-level carbon inertial balance while ensuring the continuity of the document stamping-scanning-uploading business process.

[0062] In some embodiments, for carbon flexible load devices, the tasks to be executed and their associated spatiotemporal elastic entropy are obtained, and a dynamic carbon price surface is constructed that integrates future power grid carbon intensity prediction data and target cloud service area carbon intensity data. The resource requirements and digital service call paths of the tasks to be executed by the carbon flexible load devices are analyzed, and the two are mapped to several carbon cost requirement packages on the dynamic carbon price surface. For each carbon cost requirement package, under the premise of satisfying its task deadline and resource constraints, the future execution time point and digital resource call path with the minimum carbon cost integral are found on the dynamic carbon price surface, and a task orchestration strategy item containing specific time offset instructions and cloud area selection instructions is generated.

[0063] The dynamic carbon price surface can be a three-dimensional mathematical model, where the two horizontal axes represent future time (e.g., the next 24 hours, at 15-minute intervals) and selectable target cloud service regions (e.g., region A, region B, region C), and the vertical axis represents the overall carbon intensity (unit: kgCO2 / kWh). This surface is constructed by fusing future grid carbon intensity prediction data (predictions from the grid API) with target cloud service region carbon intensity data (real-time or predicted data from various cloud service providers). A carbon cost requirement package can be an abstract data object formed after carbon cost modeling of a task to be executed by a "carbon flexible load device." When analyzing the task, the system extracts its resource requirements (e.g., required CPU cores, memory GB, network bandwidth) and digital service call paths (e.g., which cloud storage API needs to be called, its data source, and destination region). The carbon cost integral can be a mathematical integral or weighted summation operation used to calculate the total carbon emissions of a carbon cost requirement package under a specific execution path (i.e., a specific [time point, region] sequence) on the dynamic carbon price surface. Task orchestration policy items can be the specific output of policy generation. They are structured instruction objects that contain specific time offset instructions (e.g., "delay the task start time from 14:00 to 02:00") and cloud region selection instructions (e.g., "switch the computing service from region A to region B").

[0064] As an optional embodiment, the specific implementation of this application's solution is as follows: An administrative unit needs to complete the scanning, OCR recognition, and cloud archiving of all contract documents for the day before 17:00 daily. The system identifies that this task runs on an office computer connected to a local scanner and cloud OCR service. This device is classified as a carbon flexible load device, with a spatiotemporal elasticity entropy score of 0.82 (out of 1.0), indicating that it has strong time extensibility and regional migration capabilities. The dynamic carbon price surface constructed by the system shows that: during the period from 00:00 to 02:00 the next day, the carbon intensity of the power grid in East China is at its lowest point (28 gCO2 / kWh), but the carbon intensity of cloud services in Southwest China is even lower (19 gCO2 / kWh), and the PUE is better (1.18 vs 1.32); while during the peak period from 17:00 to 18:00, the carbon intensity in both regions is higher than 45 gCO2 / kWh. The system mapped the task into three carbon cost requirement packages and determined that: scanning should be scheduled during local off-peak hours (01:00), OCR calls should be directed to the Southwest region (01:05), and archiving uploads should be synchronized to the same regional object storage (01:08). The final generated strategy items were: Time offset instruction: overall start time postponed to 01:00; Cloud region selection instruction: OCR service call target set to 'CN-SW', archive storage bucket set to 'bucket-sw-contract-2025'. After receiving the data, the device agent automatically adjusted the task triggering logic, without affecting the timeliness of contract archiving. The measured carbon emission reduction rate for a single task was 36.2%.

[0065] In some embodiments, fluctuations in the grid carbon intensity signal are monitored in real time. When a short-term change in the grid carbon intensity signal is detected, the operating power of the carbon flexible load device is adjusted first, or its interruptible tasks are immediately started or stopped to respond to changes in the carbon signal and offset overall carbon cost fluctuations. At the same time, for carbon inertial load devices, their current operating mode is maintained to avoid significant mode transition carbon steps caused by their state switching. Through the active adjustment of the carbon flexible load device and the state maintenance of the carbon inertial load device, the overall carbon inertial balance of the system is achieved under the premise of meeting the business process requirements, and a corresponding real-time power adjustment and state locking instruction set is generated.

[0066] The grid carbon intensity signal can be a data stream obtained in real time from grid operators or third-party data service providers, reflecting the average CO2 emission equivalent per unit of electricity (typically 1 kWh) supplied by the grid. Short-term abrupt changes can be significant fluctuations in the grid carbon intensity signal exceeding a preset threshold (e.g., a change exceeding 0.1 kgCO2 / kWh) within a short period (e.g., within 5 minutes). Carbon inertial balance is a core concept created in this scheme, describing a system steady state achieved through differentiated collaborative control. Its implementation path is as follows: allowing carbon-flexible load devices to act as "regulators," actively absorbing (hedging) the carbon cost impact caused by grid carbon signal fluctuations; simultaneously, allowing carbon-inertial load devices to act as "stabilizers," maintaining their state to avoid generating new, high-cost switching carbon costs.

[0067] As an optional embodiment, the specific implementation of this application is as follows: The system is deployed in a government office building, connecting to the local power grid carbon intensity API, document management system, meeting reservation platform, and equipment energy gateway. At 10:17:23 AM on a certain weekday morning, the system captures a sudden change in carbon intensity (+0.092 kgCO2 / kWh / 2min) and immediately identifies three types of devices: Carbon flexible load devices include an office computer E (spatiotemporal elastic entropy = 0.83, digital carbon flow coupling degree = 0.76) performing background log archiving, and a high-speed scanner F (spatiotemporal elastic entropy = 0.91) in standby mode; Carbon inertial load devices include a meeting terminal G (modal transition carbon step = 0.12 gCO2) running a remote video reception system, and an A3-size all-in-one machine H (modal transition carbon step = 0.15 gCO2) in preheating mode. The system generated the following instruction set: CPU frequency was limited to 70% for E and archiving tasks were delayed by 15 minutes; energy-saving scanning mode was enabled for F (reducing LED brightness and motor speed); and full-time modal lock instructions were issued to G and H (prohibiting entry into sleep mode and preventing disconnection of HDMI and USB peripherals). After execution, the carbon emission increase of the entire building was reduced by 87.6% within 10 minutes, the conference system experienced zero interruptions, and printing tasks were completed on time as scheduled, verifying the feasibility and robustness of this strategy in a real office scenario.

[0068] In some embodiments, a specific set of office business processes is parsed, and all device activities and digital service calls triggered sequentially or in parallel to complete the business processes are modeled as a digital carbon chain with node and edge relationships. Based on the set of device carbon footprint fluctuation information, each node in the digital carbon chain is assigned its corresponding carbon behavior characteristic data. The total carbon footprint of the digital carbon chain under different node execution sequences and different service call paths is simulated to identify the key bottleneck nodes with the largest carbon emission contribution. For key bottleneck nodes, if they belong to carbon flexible load devices, their task orchestration and digital service call paths are optimized; if they belong to carbon inertial load devices, their upstream task sequencing is optimized to reduce unnecessary mode switching.

[0069] A digital carbon chain can be a complete sequence of all device activities (such as computer operations and printer startups) and digital service calls (such as cloud storage uploads and video conferencing connections) triggered sequentially or in parallel within a system to complete a specific office business process. Node and edge relationships can be a graph model used to describe the logical connections between elements in the digital carbon chain. Carbon behavior characteristic data can be quantitative characteristic values ​​assigned to the specific devices or services associated with each node in the chain during the construction of the digital carbon chain, derived from the device carbon footprint fluctuation information set. The total carbon footprint can be the sum of the carbon footprints of all nodes calculated in process reengineering analysis after simulating a complete digital carbon chain (including all its nodes and edges) according to a specific node execution order and service call path (i.e., a specific process instantiation scheme). A critical bottleneck node can be the node(s) whose individual carbon footprint contribution rate (or its induced impact on the carbon footprint of other nodes in the chain) is significantly higher than that of other nodes during the simulation of the total carbon footprint of the digital carbon chain under different node execution orders and different service call paths. A digital service call path can be the specific flow and destination sequence of digital services (such as data access, computing, and communication) involved in the execution of a device's active node.

[0070] As an optional embodiment, the specific implementation of the solution in this application is as follows: Taking the contract stamping and archiving process of a certain unit as an example, the original execution sequence of this process is: printing paper contracts → manual signing → scanning into PDF → calling cloud OCR to recognize text → uploading OCR results to the document management system → initiating electronic approval → automatic archiving after approval. The system first models this process as a digital carbon chain, which includes 7 nodes (printing, signing, scanning, OCR recognition, uploading, approval, archiving) and 6 directed edges. After assignment, it was found that the OCR recognition node is bound to a conference terminal with high digital carbon flow coupling and medium spatiotemporal elastic entropy, and the scanning node is bound to a high-speed all-in-one machine with high modal transition carbon steps. After simulating multiple execution paths, it was identified that the OCR recognition node has an average contribution rate of 42% in all high carbon paths, making it a key bottleneck node; its associated equipment is classified as a carbon flexible load device as mentioned above. The system then generated an optimization strategy: offloading the OCR recognition action from the local conference terminal and having it handled by an OCR microservice deployed in the Southwest Low-Carbon Cloud region; postponing the recognition task to the period of lowest grid carbon intensity in the early morning of the following day for batch execution; simultaneously, to reduce the frequency of scanning node triggers, the system suggested merging multiple contracts for scanning and uploading in a single batch, reducing the number of calls in the scan-OCR-upload link. The final low-carbon process optimization proposal included the aforementioned path change instructions, the expected carbon emission reduction (estimated to be a decrease of 31.6%), an assessment of the impact on approval timeliness (delay not exceeding 2 hours), and a rollback contingency plan.

Claims

1. An intelligent decision-making system based on office equipment carbon footprint data and low-carbon design, characterized in that, include: A multi-source dataset of office equipment is obtained. Based on the multi-source dataset of office equipment, carbon behavior analysis of office equipment is performed, including spatiotemporal elasticity, digital carbon flow coupling degree and modal transition characteristics, to generate a set of equipment carbon footprint fluctuation information. Obtain a set of specific office business processes, and based on the set of equipment carbon footprint fluctuation information and the set of specific office business processes, make low-carbon decisions oriented towards process carbon chain optimization and multi-objective collaboration, and generate a set of personalized low-carbon management strategies. Based on the aforementioned personalized low-carbon management strategy set, suggestions are made for the coordinated scheduling and control of the cluster of related office equipment, and a low-carbon process optimization proposal is generated.

2. The system according to claim 1, characterized in that, The process of generating the device carbon footprint fluctuation information set includes: The multi-source dataset of office equipment includes equipment task time sequence logs, network service call records, and equipment power status sequences. Based on the multi-source dataset of the office equipment, the spatiotemporal elastic entropy, digital carbon flow coupling degree, and modal transition carbon step of the equipment are quantitatively analyzed. By integrating the spatiotemporal elastic entropy, the digital carbon flow coupling degree, and the modal transition carbon step, a carbon behavior feature profile of office equipment is constructed, and the information is encapsulated to form a set of carbon footprint fluctuation information of the equipment.

3. The system according to claim 2, characterized in that, The analysis process of the spatiotemporal elastic entropy includes: Parse the start time, end time, resource requirements, and task priority tag of the current task in the device task timing log; Analyze the time window length of the current task from the start time to the end time, and combine it with the task priority label to determine whether the current task is allowed to be executed in segments or suspended midway within the time window, thus obtaining the segment suspension capability. At the same time, based on the resource requirements, the degree of dependence of the current task on specific network nodes or computing resources is analyzed, and its execution feasibility in different logical locations is evaluated. Combining the segmented suspension capability and the execution feasibility, the schedulable uncertainty of the current task in terms of time and logical location is quantitatively calculated; Based on the aforementioned schedulable uncertainty, a quantitative value characterizing the degree to which a device task can be delayed, migrated, or interrupted is generated, which serves as the spatiotemporal elastic entropy.

4. The system according to claim 2, characterized in that, The analysis process of the digital carbon flow coupling degree includes: Parse the network service request logs and data stream destination information initiated by the device in the network service call record; By associating real-time carbon intensity data from different destination data centers or cloud service regions and combining this with the power usage efficiency values ​​of data centers in each region, the indirect carbon emission intensity in the cloud induced by equipment activities, after efficiency conversion, is calculated. At the same time, the charging mode and health status data of the battery associated with the device are analyzed, and the battery life decay rate is assessed based on the number of charging cycles and the charging rate, thereby quantifying the impact of device usage behavior on the implicit carbon loss of the battery throughout its entire life cycle. By combining the cloud-based indirect carbon emission intensity after efficiency conversion with the degree of impact of implicit carbon loss throughout the battery's life cycle, a quantitative value characterizing the intensity of the device's digital ecological carbon footprint is generated, which serves as the digital carbon flow coupling degree.

5. The system according to claim 2, characterized in that, The analysis process of the carbon step modal transition includes: Analyze the steady-state power value sequence of the device under different operating modes and the power rise or fall curve data during the transient process of mode switching in the device power state sequence; Based on the power ramp-up or ramp-down curve data, identify and extract the absolute value of power change and the transition time required to reach the new steady state during the process of switching the device from the first steady-state operating mode to the second steady-state operating mode; Based on the absolute value of the power change and the transition duration, the additional carbon consumption generated during a single mode switch during the transition period is calculated by integration. Combined with the mode switching frequency obtained from the statistical analysis of the device power state sequence within a preset period, a quantitative value characterizing the carbon cost of device state switching is generated as the carbon step of the mode transition.

6. The system according to claim 3, characterized in that, The process of generating the personalized low-carbon management strategy set includes: Based on the device carbon footprint fluctuation information set, the devices are classified into personalized decision-making devices with different carbon behavior characteristics; Based on the specific set of office business processes and the personalized decision-making device, a set of personalized low-carbon management strategies is constructed with the goal of minimizing the global carbon footprint: In the set of personalized low-carbon management strategies, different personalized decision-making devices are configured with a forward-looking task orchestration strategy based on carbon opportunity windows, a real-time modal stabilization strategy based on carbon inertia balance, and an end-to-end process reengineering strategy based on digital carbon chain decomposition.

7. The system according to claim 6, characterized in that, The classification process of the personalized decision-making device includes: By analyzing the spatiotemporal elastic entropy and the digital carbon flow coupling degree, we can identify devices that have adaptive potential to the fluctuations of carbon intensity signals in the power grid and cloud at the time and digital service levels, and define them as carbon flexible load devices. The modal transition carbon step is analyzed to identify devices whose carbon footprint is more sensitive to its own working state switching behavior than a preset sensitivity threshold, and whose state maintenance is more carbon efficient than state switching, and these devices are defined as carbon inertial load devices.

8. The system according to claim 7, characterized in that, The forward-looking task orchestration strategy based on carbon opportunity windows includes: For the aforementioned carbon flexible load device, obtain its task to be executed and its associated spatiotemporal elastic entropy, and construct a dynamic carbon price surface that integrates future power grid carbon intensity prediction data and target cloud service area carbon intensity data; The resource requirements and digital service call paths of the tasks to be executed by the carbon flexible load device are analyzed, and the two are mapped to several carbon cost requirement packages on the dynamic carbon price surface. For each carbon cost requirement package, under the premise of meeting its task deadline and resource constraints, the future execution time point and digital resource call path with the minimum carbon cost integral are found on the dynamic carbon price surface, and a task orchestration strategy item containing specific time offset instructions and cloud region selection instructions is generated.

9. The system according to claim 7, characterized in that, The real-time modal stabilization strategy based on carbon inertial equilibrium includes: Real-time monitoring of fluctuations in the power grid carbon intensity signal; when a short-term sudden change in the power grid carbon intensity signal is detected, priority is given to adjusting the operating power of the carbon flexible load equipment or immediately starting or stopping its interruptible tasks in response to changes in the carbon signal and offsetting overall carbon cost fluctuations. Meanwhile, the current operating mode of the carbon inertial load device is maintained to avoid significant carbon step transitions in the mode due to its state switching. By actively adjusting the carbon flexible load device and maintaining the state of the carbon inertial load device, the overall carbon inertial balance of the system is achieved while meeting the requirements of the business process, and a corresponding real-time power adjustment and state locking instruction set is generated.

10. The system according to claim 7, characterized in that, The end-to-end process reengineering strategy based on digital carbon chain decomposition includes: The specific set of office business processes is analyzed, and all device activities and digital service calls triggered sequentially or in parallel by the completed business processes are modeled as a digital carbon chain with node and edge relationships. Based on the device carbon footprint fluctuation information set, each node in the digital carbon chain is assigned its corresponding carbon behavior characteristic data. Simulate the total carbon footprint of the digital carbon chain under different node execution sequences and different service call paths to identify the key bottleneck nodes that contribute the most to carbon emissions. For the aforementioned critical bottleneck nodes, if they belong to the carbon flexible load device, their task orchestration and digital service call paths will be optimized. If it belongs to the carbon inertial load device, optimize its upstream task sequencing to reduce its unnecessary mode switching.

Images