Energy management and control platform based on carbon emission monitoring

Abstract

The invention belongs to the technical field of energy management and information technology crossing, and particularly discloses and provides an energy management and control platform based on carbon emission monitoring, which comprises a carbon emission monitoring module, a carbon emission calculation module, a carbon intensity analysis module, a task scheduling module, a task allocation optimization module and a task feedback terminal. According to the method, the indirect carbon emission, the total active power consumption and the energy carbon intensity of the data center are converted into the unit calculation power carbon intensity, so that the carbon emission efficiency of the data centers in different areas can be quantified, and the carbon efficiency priority sequence is generated accordingly, so that the task is intelligently and dynamically allocated to the data center with the highest carbon efficiency at present; and meanwhile, the distribution result is optimized by introducing the execution success rate and the completion time of the historical task, so that the scheduling deviation possibly caused by purely depending on the real-time carbon efficiency is effectively avoided, and the reliability and the service quality of task execution are ensured while the low carbon emission level is maintained.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of energy management and information technology, and relates to an energy management platform based on carbon emission monitoring. Background Technology

[0002] Driven by the digital economy, the energy consumption and carbon emissions of data centers are becoming increasingly prominent. To achieve dual carbon targets, energy management technologies integrating carbon emission monitoring have received widespread attention. These technologies typically collect data center energy consumption data and combine it with carbon accounting models to assess carbon emissions, providing a basis for energy efficiency optimization and carbon management. Currently, some platforms have implemented basic energy consumption monitoring and carbon accounting, and a few systems are attempting to introduce carbon factors into task scheduling to initially achieve carbon optimization during the operation and maintenance phase.

[0003] For example, Chinese invention patent CN117217497A discloses an integrated energy management platform and integrated energy management method. This platform constructs a system architecture including an infrastructure layer, a data transmission layer, a platform service layer, and an auxiliary service layer. It can collect comprehensive energy data from various energy devices and respond to user commands at the platform service layer, executing corresponding operations. Its auxiliary service layer further provides peak-shaving assistance to help achieve energy supply and demand balance and uniformly manages the carbon emissions of related energy devices, thereby realizing centralized monitoring and basic carbon management of multiple types of energy devices.

[0004] The existing technologies mentioned above have the following shortcomings: 1. Existing technologies cannot provide dynamic real-time carbon intensity factors that reflect the real-time power generation structure for eastern data centers that are mainly powered by the power grid. At the same time, it is difficult to accurately assess the actual carbon emission reduction benefits generated by western data centers absorbing local renewable energy. This leads to biases in the assessment of carbon performance of data centers in different regions and makes it difficult to support resource scheduling decisions guided by precise carbon efficiency.

[0005] 2. Existing technologies mainly focus on energy supply and demand balance, without establishing a linkage mechanism between carbon emissions and task scheduling, and also lack a unified indicator to convert carbon intensity into carbon intensity per unit of computing power. Consequently, it is impossible to dynamically schedule tasks to the western low-carbon data center based on carbon efficiency. At the same time, there is a lack of a closed-loop optimization mechanism based on historical task execution success rate and completion time, making it difficult to achieve the comprehensive optimization of carbon efficiency and task performance. Summary of the Invention

[0006] In view of this, in order to solve the problems mentioned in the background technology, an energy management platform based on carbon emission monitoring is proposed.

[0007] The objective of this invention can be achieved through the following technical solution: This invention provides an energy management platform based on carbon emission monitoring, comprising: a carbon emission monitoring module, which monitors the total active power consumption of each eastern data center in the current monitoring period through deployed smart meters, and obtains the real-time carbon intensity factor of the power grid in its region through a data interface.

[0008] The carbon emission calculation module calculates the indirect carbon emissions of each eastern data center during the current monitoring period based on the power consumption and real-time carbon intensity factor.

[0009] The carbon intensity analysis module monitors renewable energy data for each western data center during the current monitoring period and analyzes the energy carbon intensity of each western data center during the current monitoring period.

[0010] The task scheduling module monitors the operational status data of each data center based on the resource requirements of the computing tasks to be executed, and dynamically allocates computing tasks in conjunction with the indirect carbon emissions and the energy carbon intensity.

[0011] The task allocation optimization module optimizes task allocation based on the allocation results of computational tasks and combined with historical task execution data.

[0012] The task feedback terminal feeds back the task allocation optimization results to the energy management platform.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention monitors the real-time carbon intensity factor of the eastern data center and calculates the indirect carbon emissions by combining its energy consumption. At the same time, it analyzes the energy carbon intensity of the western data center by monitoring the renewable energy data, so as to truly reflect the actual carbon emission level of data centers in different regions, and thus provide an accurate and reliable data basis for subsequent cross-regional carbon efficiency scheduling.

[0014] (2) This invention converts the indirect carbon emissions, total active power consumption and energy carbon intensity of data centers into carbon intensity per unit computing power, making the carbon emission efficiency of data centers in different regions quantifiable. Based on this, a carbon efficiency priority sequence is generated, thereby intelligently and dynamically allocating tasks to the data center with the highest current carbon efficiency, thus realizing the low-carbon upgrade of scheduling logic.

[0015] (3) By introducing historical task execution success rate and completion time to optimize the allocation results, this invention effectively avoids scheduling deviations that may be caused by relying solely on real-time carbon efficiency. Thus, while maintaining a low carbon emission level, it ensures the reliability of task execution and service quality, achieving a balance between environmental protection goals and business stability.

[0016] (4) By sorting the tasks to be executed according to their submission time and dynamically matching and allocating them with the carbon efficiency priority sequence of available data centers, the present invention achieves global optimization of the overall carbon efficiency of the system and effectively avoids the situation where high-carbon tasks occupy low-carbon data center computing resources too early in local scheduling. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram showing the connections of the various modules in the system of the present invention.

[0019] Figure 2 This is a schematic diagram showing the connection steps of the energy carbon intensity analysis of the present invention.

[0020] Figure 3 This is a schematic diagram showing the connection of the dynamic allocation steps for computing tasks in this invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see Figure 1 As shown, the present invention provides an energy management platform based on carbon emission monitoring, including: a carbon emission monitoring module, a carbon emission calculation module, a carbon intensity analysis module, a task scheduling module, a task allocation optimization module, and a task feedback terminal.

[0023] In the above, the carbon emission calculation module is connected to the carbon emission monitoring module and the task scheduling module, the task scheduling module is connected to the carbon intensity analysis module and the task allocation optimization module, and the task allocation optimization module is also connected to the task feedback terminal.

[0024] The carbon emission monitoring module monitors the total active power consumption of each eastern data center during the current monitoring period through deployed smart meters, and obtains the real-time carbon intensity factor of the power grid in its region through a data interface.

[0025] For example, the monitoring of total active power consumption includes: deploying smart meters at the main power input circuit of each eastern data center, and periodically collecting the instantaneous active power of the main power input circuit through the smart meters to form active power time-series data.

[0026] The active power time-series data is integrated within the time window of the current monitoring period to calculate the total active power consumption of each eastern data center within the current monitoring period.

[0027] It should be added that the calculation of the total active power consumption is specifically as follows: the instantaneous active power value is periodically collected at a fixed sampling frequency by smart meters deployed in the main power input circuit of each data center. This forms a series of discrete data points arranged in chronological order, wherein the active power time-series data is... ,in Representative monitoring point.

[0028] Within the time window of the current monitoring period, the active power time-series data is integrated over the current monitoring period to obtain the electrical energy consumed in that period.

[0029] Calculate the total active power consumption during the current monitoring period. , , and The first The and the first Instantaneous active power values ​​at each monitoring point It is the interval between adjacent monitoring points.

[0030] The trapezoidal integral method described above significantly improves computational accuracy compared to the simple averaging method, which assumes constant power. The simple averaging method takes the arithmetic mean of the power at all sampling points and multiplies it by the total time. This can introduce significant errors when power fluctuates wildly, as it smooths out instantaneous peaks and troughs. In contrast, the method of this invention performs piecewise linear approximation of the instantaneous active power time-series data, calculates the area of ​​the trapezoidal region between each adjacent sampling point, and sums the results. This allows for a more accurate reconstruction of the non-constant actual power curve and a better fit to its true trajectory.

[0031] For example, the acquisition of the real-time carbon intensity factor includes: obtaining the active power of various generator sets at each monitoring time point within the current monitoring period from the power grid of the region to which each eastern data center belongs through a data interface, wherein the generator set types include, but are not limited to, coal-fired units, gas-fired units, hydroelectric units, wind turbine units, and photovoltaic units.

[0032] It should be noted that the data interface may be, for example, calling the public API provided by the regional power grid operator, or interacting with the power grid control system through the power dispatch data network.

[0033] By matching various types of generator sets with carbon emission coefficients per unit of power generation categorized by generator set type, carbon emission coefficients per unit of power generation for each type of generator set are obtained. The rules for setting these carbon emission coefficients are as follows: for fossil fuel units such as coal-fired and gas-fired units, a fixed emission coefficient based on fuel type is used. For renewable energy units such as hydropower units, wind power units, and photovoltaic units, a zero or near-zero emission coefficient is used.

[0034] The active power of various generator sets is multiplied by their carbon emission coefficient per unit of power generation, and the results are summed to obtain the total carbon emission rate of the power grid.

[0035] The total active power of the power grid is obtained by summing the active power of various generator sets.

[0036] The ratio of the total carbon emission rate of the power grid to the total active power of the power grid is used as the real-time carbon intensity factor at each monitoring time point.

[0037] The average value of the real-time carbon intensity factor at each monitoring time point is calculated to obtain the real-time carbon intensity factor of the power grid in the area where each eastern data center is located during the current monitoring period.

[0038] The carbon emission calculation module calculates the indirect carbon emissions of each eastern data center during the current monitoring period based on the power consumption and real-time carbon intensity factor.

[0039] For example, the calculation of the indirect carbon emissions includes: obtaining the total active power consumption of each eastern data center during the current monitoring period and the real-time carbon intensity factor of its region during the current monitoring period.

[0040] The total active power consumption during the current monitoring period is multiplied by the real-time carbon intensity factor to obtain the indirect carbon emissions of each eastern data center during the current monitoring period.

[0041] The carbon intensity analysis module monitors the renewable energy data of each western data center during the current monitoring period and analyzes the energy carbon intensity of each western data center during the current monitoring period.

[0042] For example, the monitoring of renewable energy data includes: real-time monitoring of the instantaneous active power data of the supporting renewable energy power generation facilities through monitoring systems deployed in various western data centers. The monitoring system reads data in real time using industrial communication protocols such as Modbus and OPCUA through data acquisition units deployed in photovoltaic inverters, wind turbine controllers, and hydro turbine control systems. The data includes, but is not limited to, instantaneous active power of photovoltaic power generation, instantaneous active power of wind power generation, and instantaneous active power of hydropower generation.

[0043] The instantaneous active power data is integrated within the time window of the current monitoring period to obtain the total renewable energy power generation of each western data center within the current monitoring period.

[0044] It should be added that the calculation process of the total renewable energy power generation is as follows: First, the instantaneous active power monitored by photovoltaic, wind power, hydropower and other power generation facilities is summed in real time to obtain the total instantaneous active power of renewable energy of the western data center at the current moment.

[0045] The resulting time-series data of total instantaneous active power of renewable energy is then integrated within the time window of the current monitoring period to obtain the total renewable energy power generation within that period.

[0046] The total active power consumption of each western data center is monitored by smart meters deployed in the main power input circuit of each western data center during the current monitoring period. It should be understood that the method for obtaining the total active power consumption of the western data center is the same as that for the eastern data center, and will not be repeated here.

[0047] The total renewable energy generation and total active energy consumption are used as renewable energy data for each western data center.

[0048] Please see Figure 2 As shown, for example, the analysis of the energy carbon intensity of each western data center includes comparing the total active power consumption of each western data center with its total renewable energy generation during the current monitoring period.

[0049] When the total renewable energy generation is greater than or equal to the total active power consumption, it indicates that the power consumed by the data center during the current monitoring period is entirely generated by its supporting renewable energy facilities and does not consume any electricity that may be generated by fossil fuels in the regional power grid. Therefore, the energy carbon intensity of the western data center is determined to be 0.

[0050] When the total renewable energy generation is less than the total active energy consumption, it indicates that the self-generated renewable energy of the data center is insufficient to meet its total energy consumption. The power gap needs to be supplemented by the external regional power grid. The difference between the total active energy consumption and the total renewable energy generation is calculated to obtain the power grid supplement. The power grid supplement is then multiplied by the real-time carbon intensity factor to obtain the carbon emissions. The real-time carbon intensity factor is the real-time carbon intensity factor of the power grid in the region where the western data center is located, and its acquisition method is the same as that of the eastern data center.

[0051] The energy carbon intensity of each western data center is calculated by dividing the carbon emissions by the total active power consumption.

[0052] Understandably, this solution establishes a set of accounting rules for the energy carbon intensity of western data centers: prioritizing the full utilization of local renewable energy, with the shortfall only counted as grid-supplied electricity when renewable energy generation is insufficient to meet its own energy consumption. Under this rule, if renewable energy generation is greater than or equal to total active power consumption, it is equivalent to the data center not generating net grid electricity during the current period, thus its energy carbon intensity is counted as 0. This accounting boundary is designed to establish a unified and concise benchmark for the carbon intensity per unit computing power of data centers in the east and west. In actual deployments, for complex scenarios involving specific grid purchase agreements or green certificate traceability management, this method can be adapted and refined.

[0053] The task scheduling module monitors the operating status data of each data center based on the resource requirements of the computing tasks to be executed, and dynamically allocates computing tasks in conjunction with the indirect carbon emissions and the energy carbon intensity.

[0054] Please see Figure 3 As shown, exemplarily, the dynamic allocation of computing tasks includes: Q1, determining the set of available data centers for each computing task to be executed based on the operating status data of each data center and the resource requirements of the computing tasks to be executed.

[0055] Furthermore, determining the set of available data centers for each computing task to be executed includes: Q1-1, obtaining the required computing resources and required memory capacity for execution from the resource requirements of the computing tasks to be executed.

[0056] Q1-2. Obtain the current available computing resources and memory capacity of each data center from the running status data, and compare them with the computing resources and memory capacity required by the computing tasks to be executed.

[0057] Q1-3. Select data centers whose available computing resources are greater than or equal to the required computing resources and whose memory capacity is greater than or equal to the required memory capacity as available data centers for the computing tasks to be executed, and then generate a set of available data centers for each computing task to be executed.

[0058] Q2. Based on the indirect carbon emissions and total active power consumption of the eastern data centers and the energy carbon intensity of the western data centers, calculate the carbon intensity per unit computing power of each data center in the available data center set.

[0059] Furthermore, the calculation of the carbon intensity per unit computing power includes: Q2-1, if the data center is located in the eastern region, then the indirect carbon emissions are divided by its total active power consumption to obtain the carbon intensity per unit computing power of the data center.

[0060] The principle is as follows: Eastern data centers use indirect carbon emissions divided by total active power consumption as the carbon intensity per unit of computing power. This is because the carbon emissions of Eastern data centers come entirely from purchased electricity, and their carbon footprint is entirely composed of indirect carbon emissions. This formula calculates the carbon intensity per unit of energy consumption, that is, the amount of carbon emissions corresponding to each kilowatt-hour of electricity consumed. Since the computing power output and energy consumption of Eastern data centers are strongly linearly positively correlated, the carbon intensity per unit of energy consumption can be directly and equivalently used as the carbon intensity per unit of computing power. The lower this value, the lower the carbon emissions of the Eastern data center under the same computing power output.

[0061] Q2-2. If the data center is located in the western region, then energy carbon intensity shall be used as its unit computing power carbon intensity.

[0062] Considering the characteristics of western data centers, their energy carbon intensity is a normalized indicator reflecting their carbon footprint per unit of energy consumption. It comprehensively considers the contribution of local renewable energy and the impact of grid-supplied power. Therefore, the energy carbon intensity of western data centers is completely consistent and comparable in physical meaning and dimensions to the carbon emission intensity per unit of energy consumption obtained by division from eastern data centers. Directly using it as the carbon intensity per unit of computing power avoids double-counting of the same concept, ensuring system efficiency and guaranteeing a fair carbon efficiency ranking of eastern and western data centers under the same standard.

[0063] Q3. Based on the carbon intensity per unit computing power, sort the data centers in the available data center set from low to high according to the carbon intensity per unit computing power, and generate a carbon efficiency priority sequence for each computing task to be executed.

[0064] Q4. Obtain the task submission time of each pending computing task from the resource requirements, and sort each pending computing task in ascending order from earliest to latest according to its task submission time to generate a sequence of tasks to be scheduled.

[0065] Q5. Obtain the first unassigned computing task to be executed from the task sequence to be scheduled, assign it to the first available data center in the corresponding carbon efficiency priority sequence, update the available resource status of the data center according to this assignment, and synchronously update the carbon efficiency priority sequence of all affected tasks.

[0066] Q6. Repeat the task allocation operation until all pending computing tasks have been allocated to the data center.

[0067] As an example, this embodiment is illustrated with a simplified scenario. Suppose that at a certain scheduling time, there are 3 computing tasks T1, T2, and T3 to be executed in the system, and 3 available data centers, where D1 is in the east, D2 is in the west, and D3 is in the west.

[0068] S1. Generate a scheduling sequence. First, obtain the submission times of all tasks and sort them in ascending order from earliest to latest. The generated task sequence to be scheduled is T1→T2→T3. The system generates an independent carbon efficiency priority sequence for each task. Assuming that by calculating the carbon intensity per unit computing power of each data center and sorting them from low to high, the initial sequence obtained is D2→D3→D1, this indicates that D2 is the data center with the lowest carbon intensity per unit computing power, and D1 is the data center with the highest carbon intensity per unit computing power.

[0069] S2. Assign and update the sequence for task T1. The system takes the first task T1 from the sequence of tasks to be scheduled and assigns it to the preferred data center D2 in its carbon efficiency priority sequence. Subsequently, the system updates the carbon efficiency priority sequences of all unassigned tasks T2 and T3 based on the remaining available resources after D2 is assigned a task. If the remaining resources of D2 cannot meet the needs of T2 or T3, then T2 or T3 is removed from the sequence of T2 and T3, and the updated sequence becomes D3→D1.

[0070] S3. Assign and update the sequence for task T2. The system then processes task T2, assigning it to the preferred data center D3 in its current carbon efficiency priority sequence. Similarly, the system updates the carbon efficiency priority sequence of the remaining unassigned task T3 based on the remaining resources after D3's allocation. If the remaining resources in D3 are insufficient to satisfy T3, then T3 is removed from T3's sequence.

[0071] S4. Allocate a data center for task T3. Finally, the system processes task T3. At this point, only data center D1 remains available in its carbon efficiency priority sequence, so the system allocates T3 to D1. The scheduling result is: T1→D2, T2→D3, T3→D1.

[0072] As can be seen, this embodiment combines task submission time sorting with a dynamically updated carbon efficiency priority sequence, and synchronizes resource status in real time during the allocation process. This achieves efficient and stable cross-regional carbon-aware scheduling by prioritizing the allocation of tasks to the available data center with the highest carbon efficiency while ensuring scheduling fairness.

[0073] The task allocation optimization module optimizes task allocation based on the allocation results of the calculated tasks and combined with historical task execution data.

[0074] For example, the task allocation optimization includes: assigning the data center to each computing task to be executed as a data center to be verified, and obtaining the historical task execution success rate and the average task completion time of the same task type from the historical task execution data.

[0075] The historical task execution success rate and the average task completion time are compared with their preset thresholds.

[0076] When the historical task execution success rate is higher than or equal to the preset task execution success rate threshold, and the average task completion time is lower than or equal to the preset task completion time threshold, it indicates that within the current monitoring period, the data center allocated to the task to be executed not only performs well in terms of carbon intensity per unit of computing power, but also its actual operating status is sufficient to stably and efficiently support the execution of this type of task. Therefore, it is determined that the allocation of the task to be executed does not need to be optimized.

[0077] It should be added that the preset task execution success rate threshold refers to the minimum acceptable percentage of task execution success to ensure the reliability of the computing service. It can be obtained based on the Service Level Agreement (SLA), specifically by setting it according to the task execution success rate target promised in the SLA signed with the user. For example, if the agreement promises 99.9% availability for a specific service, then this threshold can be set to 99.9%.

[0078] It should be added that the preset task completion time threshold refers to the maximum allowable completion time set for a specific type of computational task to meet service quality requirements. Its acquisition method can be determined based on the task-level protocol, specifically: parsing the task type from the description information of the computational task to be executed, and setting the threshold according to the target completion time set for that type of task. Experimental verification shows that for interactive tasks, this threshold can be set to 2 seconds, while for batch processing tasks, it can be set to 2 hours.

[0079] When the historical task execution success rate is lower than the preset task execution success rate threshold, or the average task completion time exceeds the preset task completion time threshold, it indicates that although the allocated data center has an advantage in theoretical carbon intensity, its actual operating state may have problems such as intense resource competition, local performance bottlenecks, or network instability, which makes it unable to provide reliable service quality assurance for the current task. Therefore, it is determined that the computing task to be executed needs to be optimized in terms of task allocation.

[0080] Remove the allocated data centers from the carbon efficiency priority sequence of the task to be executed, and then calculate the comprehensive optimization index of the task to be executed on the remaining data centers based on the carbon intensity per unit computing power of the remaining data centers, as well as the historical task execution success rate and average task completion time. Based on this, generate an optimized scheduling sequence, and extract the data center with the first ranking as the data center to be allocated to the task to be executed.

[0081] It should be added that the generation of the optimized scheduling sequence includes: normalizing the carbon intensity per unit computing power of all remaining data centers so that the data center with the lowest carbon intensity scores the highest in this aspect, for example, using a linear normalization method: In the formula For the first Normalized carbon intensity per unit computing power for the remaining data centers For the first Carbon intensity per unit computing power of the remaining data centers and These represent the maximum and minimum carbon intensity per unit of computing power for the remaining data centers.

[0082] According to the above Similarly, the normalization process is applied to the historical average task completion time to obtain the normalized historical average task completion time for each remaining data center, and denoted as [missing information]. .

[0083] Calculate the overall optimization index of each remaining data center , In the formula For the first Historical task execution success rate of the remaining data centers , and These are the weights for normalized unit computing power carbon intensity, historical task execution success rate, and normalized historical average task completion time, respectively.

[0084] It should be added that by using weighted fusion to calculate the comprehensive optimization index for each data center, on the one hand, the weight allocation can reflect the actual weight of the impact of carbon intensity per unit of computing power, historical task execution success rate, and historical average task completion time on different dimensions of task scheduling decisions, reflecting the differences in the contribution of different performance indicators to the overall performance of the data center. On the other hand, it can directly integrate information from the three dimensions of carbon emissions, reliability, and efficiency, comprehensively considering the impact of the three on task scheduling effectiveness.

[0085] The weights can be set based on platform operation strategies and actual scheduling experience, or they can be obtained through historical operating data. For example, historical data on carbon intensity per unit computing power, task execution success rate, average task completion time, and corresponding scheduling effects of the data center can be collected first. The correlation coefficients between the three and the platform's overall performance can be calculated. The contribution of each performance indicator to the overall performance can be determined through regression analysis. After normalization, the contribution is converted into the weights of carbon intensity per unit computing power, historical task execution success rate, and historical average task completion time, and the total weight is 1. In this way, the comprehensive optimization index of the data center can be accurately quantified.

[0086] Finally, all remaining data centers are sorted from high to low according to their comprehensive optimization index to generate an optimized scheduling sequence.

[0087] The task feedback terminal feeds back the task allocation optimization results to the energy management platform.

[0088] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0089] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0090] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0091] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0092] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An energy management platform based on carbon emission monitoring, characterized in that: include: The carbon emission monitoring module monitors the total active power consumption of each eastern data center during the current monitoring period through deployed smart meters, and obtains the real-time carbon intensity factor of the power grid in its region through a data interface. The carbon emission calculation module calculates the indirect carbon emissions of each eastern data center during the current monitoring period based on the power consumption and real-time carbon intensity factor. The carbon intensity analysis module monitors renewable energy data for each western data center during the current monitoring period and analyzes the energy carbon intensity of each western data center during the current monitoring period. The task scheduling module monitors the operating status data of each data center based on the resource requirements of the computing tasks to be executed, and dynamically allocates computing tasks in combination with the indirect carbon emissions and the energy carbon intensity. The task allocation optimization module optimizes task allocation based on the allocation results of computational tasks and combined with historical task execution data. The task feedback terminal feeds back the task allocation optimization results to the energy management platform.

2. The energy management platform based on carbon emission monitoring according to claim 1, characterized in that: The monitoring of total active power consumption includes: Smart meters are deployed at the main power input circuit of each eastern data center, and the instantaneous active power of the main power input circuit is periodically collected through the smart meters to form active power time series data. The active power time-series data is integrated within the time window of the current monitoring period to calculate the total active power consumption of each eastern data center within the current monitoring period.

3. The energy management platform based on carbon emission monitoring according to claim 1, characterized in that: The acquisition of the real-time carbon intensity factor includes: Through the data interface, the active power of various generator sets at each monitoring time point within the current monitoring cycle is obtained from the power grid of the region to which each eastern data center belongs; By matching various types of generator sets with the carbon emission coefficients per unit of power generation classified by generator set type, the carbon emission coefficients per unit of power generation of various types of generator sets are obtained. The active power of various generator sets is multiplied by their carbon emission coefficient per unit of power generation, and the results are summed to obtain the total carbon emission rate of the power grid. The total active power of the power grid is obtained by summing the active power of various generator sets. The ratio of the total carbon emission rate of the power grid to the total active power of the power grid is used as the real-time carbon intensity factor at each monitoring time point. The average value of the real-time carbon intensity factor at each monitoring time point is calculated to obtain the real-time carbon intensity factor of the power grid in the area where each eastern data center is located during the current monitoring period.

4. The energy management platform based on carbon emission monitoring according to claim 1, characterized in that: The calculation of the indirect carbon emissions includes: Obtain the total active power consumption of each eastern data center during the current monitoring period and the real-time carbon intensity factor of its region during the current monitoring period; The total active power consumption during the current monitoring period is multiplied by the real-time carbon intensity factor to obtain the indirect carbon emissions of each eastern data center during the current monitoring period.

5. An energy management platform based on carbon emission monitoring according to claim 1, characterized in that: The monitoring of the renewable energy data includes: The monitoring system deployed in various western data centers monitors the instantaneous active power data of their supporting renewable energy power generation facilities in real time. The instantaneous active power data is integrated within the time window of the current monitoring period to obtain the total renewable energy power generation of each western data center within the current monitoring period; The total active power consumption of each data center in the western region is monitored by smart meters deployed in the main power input circuit. The total renewable energy generation and total active energy consumption are used as renewable energy data for each western data center.

6. The energy management platform based on carbon emission monitoring according to claim 1, characterized in that: The analysis of the energy carbon intensity of each western data center includes: Compare the total active power consumption of each western data center with its total renewable energy generation during the current monitoring period; When the total renewable energy generation is greater than or equal to the total active power consumption, the energy carbon intensity of the western data center is determined to be 0. When the total renewable energy generation is less than the total active energy consumption, the difference between the total active energy consumption and the total renewable energy generation is calculated to obtain the grid supplementary power. The grid supplementary power is then multiplied by the real-time carbon intensity factor to obtain the carbon emissions. The energy carbon intensity of each western data center is calculated by dividing the carbon emissions by the total active power consumption.

7. An energy management platform based on carbon emission monitoring according to claim 1, characterized in that: The dynamically allocated computation tasks include: Q1. Based on the operational status data of each data center and the resource requirements of the computing tasks to be executed, determine the set of available data centers for each computing task to be executed. Q2. Based on the indirect carbon emissions and total active power consumption of the eastern data center and the energy carbon intensity of the western data center, calculate the carbon intensity per unit computing power of each data center in the available data center set. Q3. Based on the carbon intensity per unit computing power, sort the data centers in the available data center set from low to high according to the carbon intensity per unit computing power, and generate a carbon efficiency priority sequence for each computing task to be executed. Q4. Obtain the task submission time of each pending computing task from the resource requirements, and sort each pending computing task in ascending order from earliest to latest according to its task submission time to generate a sequence of tasks to be scheduled. Q5. Obtain the first unassigned computing task to be executed from the task sequence to be scheduled, assign it to the first available data center in the corresponding carbon efficiency priority sequence, update the available resource status of the data center according to this assignment, and synchronously update the carbon efficiency priority sequence of all affected tasks. Q6. Repeat the task allocation operation until all pending computing tasks have been allocated to the data center.

8. An energy management platform based on carbon emission monitoring according to claim 7, characterized in that: The set of available data centers for each computing task to be executed includes: Obtain the required computing resources and memory capacity for execution from the resource requirements of the computing task to be executed; The current available computing resources and memory capacity of each data center are obtained from the running status data, and then compared with the computing resources and memory capacity required by the computing tasks to be executed. Data centers with available computing resources greater than or equal to the required computing resources and memory capacity greater than or equal to the required memory capacity are selected as available data centers for the computing tasks to be executed, thereby generating a set of available data centers for each computing task to be executed.

9. An energy management platform based on carbon emission monitoring according to claim 7, characterized in that: The calculation of the carbon intensity per unit computing power includes: If the data center is located in the eastern region, the indirect carbon emissions are divided by its total active power consumption in the current monitoring period to obtain the carbon intensity per unit computing power of the data center. If the data center is located in the western region, then energy carbon intensity will be used as its unit computing power carbon intensity.

10. An energy management platform based on carbon emission monitoring according to claim 7, characterized in that: The task allocation optimization includes: The data centers assigned to each computing task to be executed are designated as data centers to be verified, and the historical task execution success rate and average task completion time of the same task type are obtained from the historical task execution data. The historical task execution success rate and the average task completion time are compared with their preset thresholds respectively; When the historical task execution success rate is higher than or equal to the preset task execution success rate threshold, and the average task completion time is lower than or equal to the preset task completion time threshold, it is determined that the allocation of the task to be executed does not need to be optimized. When the historical task execution success rate is lower than the preset task execution success rate threshold, or the average task completion time exceeds the preset task completion time threshold, it is determined that the computation task to be executed needs to be optimized in terms of task allocation. Remove the allocated data centers from the carbon efficiency priority sequence of the task to be executed, and then calculate the comprehensive optimization index of the task to be executed on the remaining data centers based on the carbon intensity per unit computing power of the remaining data centers, as well as the historical task execution success rate and average task completion time. Based on this, generate an optimized scheduling sequence, and extract the data center with the first ranking as the data center to be allocated to the task to be executed.

Images