Method for converting distributed energy into distributed computing power

Abstract

The invention provides a method for converting distributed energy into distributed computing power, relates to the technical field of distributed energy and distributed computing power conversion, and is used for improving the utilization efficiency of the distributed energy, reducing the computing power acquisition cost and realizing collaborative optimization of the energy and the computing power. The method comprises the steps that energy information of a distributed energy system is obtained, distributed energy comprises one or more of solar energy, wind energy and water energy, and the energy information comprises energy output and energy reserve; state information of distributed computing power nodes is obtained, the distributed computing power nodes comprise electronic equipment with computing power, and the state information comprises the computing power and task execution conditions of the distributed computing power nodes; and matching a target node based on the energy information, the state information and a target task, wherein the target node is the distributed computing power node with the highest matching degree with the target task.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of distributed energy and distributed computing power, and in particular, to a method for converting distributed energy into distributed computing power. Background Art

[0002] With the acceleration of the digitalization process, the global demand for computing power has shown explosive growth. However, the traditional centralized computing power supply model, such as large data centers, faces many problems. These data centers not only consume huge amounts of energy, have high construction and operating costs, but also have stringent requirements on geographical location and infrastructure. At the same time, distributed energy, such as solar energy, wind energy, tidal energy and other renewable energy, although they have the advantages of wide distribution and environmental friendliness, face challenges in energy storage and consumption due to their intermittent and volatile characteristics. A large amount of distributed energy is wasted in the production process because it cannot be used in a timely and effective manner, and a large part of the energy consumption is also lost in the inversion process from DC to AC, which causes a huge loss of resources.

[0003] Traditional centralized computing centers also face problems such as high energy consumption (and the need to convert AC to DC), high construction costs, and geographical restrictions. At the same time, distributed energy such as solar energy, wind energy, and small hydropower are widely distributed in various places, but their energy utilization needs to be converted from DC to AC and connected to the power grid. Their utilization efficiency needs to be further improved, and there are problems such as difficulty in energy consumption. Therefore, how to effectively integrate distributed energy and convert it into usable computing resources has become a technical problem that needs to be solved urgently to make up for the lack of traditional centralized computing power.

[0004] In existing technologies, computing centers generally rely on centralized power grids for power supply, which leads to problems such as high energy consumption, large carbon emissions, and high deployment costs in remote areas. At the same time, distributed energy (such as wind power and photovoltaics) has intermittent power supply characteristics, which makes it difficult to directly match the demand for high-stability computing power.

[0005] Therefore, there is an urgent need for a technical solution that can deeply integrate distributed energy and computing resources and achieve dynamic matching of energy and computing power. Summary of the invention

[0006] The embodiment of the present application provides a method for converting distributed energy into distributed computing power, so as to improve the utilization efficiency of distributed energy, reduce the cost of obtaining computing power, and achieve coordinated optimization of energy and computing power.

[0007] To achieve the above objectives, the embodiments of the present application adopt the following technical solutions:

[0008] The present application provides a distributed energy conversion and distributed computing power method, and the method includes: obtaining energy information of a distributed energy system, where the distributed energy includes one or more of solar energy, wind energy, and water energy, and the energy information includes energy output and energy reserve; obtaining status information of distributed computing power nodes, where the distributed computing power nodes include electronic devices with computing capabilities, and the status information includes the computing capabilities and task execution status of the distributed computing power nodes; matching a target node based on the energy information, the status information, and a target task, where the target node is the distributed computing power node with the highest matching degree to the target task.

[0009] In a possible implementation manner, before obtaining the energy information of the distributed energy system, the method further includes: constructing the distributed energy system based on the distributed energy and a energy storage unit, where the energy storage unit includes at least one of a battery energy storage unit and a supercapacitor energy storage unit; constructing the distributed computing power nodes based on electronic devices with computing capabilities and a configuration policy, where the configuration policy at least includes allocating identifiers and permissions to each of the distributed computing power nodes.

[0010] In a possible implementation manner, before obtaining the energy information of the distributed energy system, the method further includes:

[0011] collecting the distributed energy through a preset energy collection device in the distributed energy system;

[0012] performing preprocessing on the collected distributed energy, where the preprocessing at least includes any one of voltage adjustment, frequency conversion, and rectification and filtering.

[0013] In a possible implementation manner, matching the target node based on the energy information, the status information, and the target task includes: extracting task information of the target task, where the task information at least includes energy consumption, computational complexity, and time constraint; dynamically allocating the target task to the target node according to the status information, the energy information, and the task information through a multi-objective scheduling algorithm.

[0014] In a possible implementation manner, the method further includes: obtaining a characteristic factor of the growing computing power demand; constructing a prediction model based on the characteristic factor, where the prediction model is used to predict the computing power demand; obtaining the matching degree between the current energy parameters and the historical energy parameters; generating first predicted computing power data according to the matching degree and the historical computing power data, where the first predicted computing power data includes the computing power demands of multiple time periods; obtaining the initial value of the characteristic factor according to the first predicted computing power data, the current actual computing power data, and the prediction model; predicting the computing power demand according to the prediction model corresponding to the initial value of the characteristic factor.

[0015] In a possible implementation, the characteristic factors include a first characteristic factor and a second characteristic factor. The first characteristic factor is used to characterize the driving effect of energy supply capacity on computing power demand, and the second characteristic factor is used to characterize the collaborative growth of computing power demand caused by network effects.

[0016] In a possible implementation, the expression of the prediction model is:

[0017]

[0018] where C(t) is the computing power demand at time t, E(t) is the energy supply capacity at time t, E max is the maximum available energy capacity of the system, p is the first characteristic factor, and q is the second characteristic factor.

[0019] In a possible implementation, obtaining the initial values of the characteristic factors according to the first predicted computing power data, the current actual computing power data, and the prediction model includes: constructing an intermediate model based on historical energy data; generating second predicted computing power data through the intermediate model; correcting the predicted value according to the deviation between the current actual computing power data and the second predicted computing power data to obtain third predicted computing power data; and obtaining the initial values of the characteristic factors according to the data of the trend nodes in the third predicted computing power data and the prediction model.

[0020] In a possible implementation, correcting the predicted value according to the deviation between the current actual computing power data and the second predicted computing power data to obtain third predicted computing power data includes: obtaining the average deviation between the first data in the second predicted computing power data and the current actual computing power data, where the first data is the data in the second predicted computing power data corresponding to the current actual computing power data; and correcting the second predicted computing power data according to the average deviation to obtain the third predicted computing power data.

[0021] In a possible implementation, the trend nodes include the energy supply saturation time point, the time point when the computing power demand changes from linear growth to exponential growth, and the critical point where network effects trigger the collaborative growth of computing power demand.

[0022] In a possible implementation, the method further includes: when it is detected that the energy supply volatility exceeds the threshold, triggering a dynamic scheduling protocol, where the dynamic scheduling protocol includes shutting down unnecessary high-power computing nodes, migrating tasks to nodes with stable energy supply, and correcting one or more of the second characteristic factors according to the migration results. Description of the Drawings

[0023] Figure 1 is a schematic flowchart of the method provided by some embodiments of the present application;

[0024] Figure 2 For Figure 1 the flowchart of S400 in some embodiments;

[0025] Figure 3 For Figure 1 the flowchart of S450 in some embodiments. Detailed implementation manners

[0026] Next, the technical solutions in the embodiments of the present application will be described with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments.

[0027] Hereinafter, terms such as "first" and "second" are only used for descriptive purposes, and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first", "second", etc. may explicitly or implicitly include one or more of such features. In the description of the present application, unless otherwise stated, the meaning of "a plurality" is two or more.

[0028] In addition, in the present application, orientation terms such as "upper", "lower", "left", and "right" may include, but are not limited to, being defined relative to the schematic placement of components in the accompanying drawings. It should be understood that these directional terms may be relative concepts, and they are used for relative description and clarification, and they may change accordingly with the change of the orientation of the components in the accompanying drawings.

[0029] In the present application, unless otherwise clearly defined and limited, the term "connection" should be understood in a broad sense. For example, "connection" may be a fixed connection, a detachable connection, or integrated; it may be directly connected, or indirectly connected through an intermediate medium. In addition, the term "electrical connection" may be a way of realizing electrical connection for signal transmission.

[0030] As used herein, "about", "substantially" or "approximately" includes the stated value and reference values within an acceptable deviation range of the specific value, where the acceptable deviation range is determined by those of ordinary skill in the art considering the measurements being discussed and the errors associated with the measurement of a particular quantity (i.e., the limitations of the measurement method).

[0031] In the embodiments of the present application, words such as "exemplarily" or "for example" are used to represent examples, illustrations or explanations. Any embodiment or design described as "exemplarily" or "for example" in the embodiments of the present application should not be construed as being more preferred or having more advantages than other embodiments or designs. Exactly, words such as "exemplarily" or "for example" are used.

[0032] With the acceleration of the digitalization process, the global demand for computing power has shown explosive growth. However, the traditional centralized computing power supply model, such as large data centers, faces many problems. These data centers not only consume huge amounts of energy, have high construction and operating costs, but also have stringent requirements on geographical location and infrastructure. At the same time, distributed energy, such as solar energy, wind energy, tidal energy and other renewable energy, although they have the advantages of wide distribution and environmental friendliness, face challenges in energy storage and consumption due to their intermittent and volatile characteristics. A large amount of distributed energy is wasted in the production process because it cannot be used in a timely and effective manner, and a large part of the energy consumption is also lost in the inversion process from DC to AC, which causes a huge loss of resources.

[0033] Traditional centralized computing centers also face problems such as high energy consumption (and the need to convert AC to DC), high construction costs, and geographical restrictions. At the same time, distributed energy such as solar energy, wind energy, and small hydropower are widely distributed in various places, but their energy utilization needs to be converted from DC to AC and connected to the power grid. Their utilization efficiency needs to be further improved, and there are problems such as difficulty in energy consumption. Therefore, how to effectively integrate distributed energy and convert it into usable computing resources has become a technical problem that needs to be solved urgently to make up for the lack of traditional centralized computing power.

[0034] In existing technologies, computing centers generally rely on centralized power grids for power supply, which leads to problems such as high energy consumption, large carbon emissions, and high deployment costs in remote areas. At the same time, distributed energy (such as wind power and photovoltaics) has intermittent power supply characteristics, which makes it difficult to directly match the demand for high-stability computing power.

[0035] The embodiment of the present application provides a method for converting distributed energy into distributed computing power, so as to improve the utilization efficiency of distributed energy, reduce the cost of acquiring computing power, and achieve the coordinated optimization of energy and computing power. Figure 1 , Figure 2 as well as Figure 3 As shown, the method includes:

[0036] S010. Constructing the distributed energy system based on the distributed energy and the energy storage unit, wherein the energy storage unit includes at least one of a battery energy storage unit and a supercapacitor energy storage unit:

[0037] Exemplarily, in this embodiment, the distributed energy system includes solar energy and wind energy, and the energy storage unit includes a battery energy storage unit and a supercapacitor energy storage unit. Solar energy is converted into direct current electrical energy through photovoltaic panels, and wind energy is converted into alternating current electrical energy through wind turbines. After both are converted into direct current electrical energy through a rectifier, a part of it is directly supplied to the computing nodes for use, and the other part is stored in the battery energy storage unit and the supercapacitor energy storage unit. The battery energy storage unit is used to store a large amount of electrical energy to meet long-term energy demands, and the supercapacitor energy storage unit is used to store a small amount of electrical energy but has the characteristics of fast charge and discharge to meet high-power demands in a short period of time.

[0038] S020. Construct the distributed computing nodes based on an electronic device with computing capabilities and a configuration strategy, where the configuration strategy at least includes assigning an identifier and permissions to each of the distributed computing nodes.

[0039] Exemplarily, the distributed computing nodes are composed of electronic devices with computing capabilities, such as servers, personal computers, industrial control computers, etc. Through the configuration strategy, a unique identifier and permissions are assigned to each computing node. Among them, the identifier can be used to distinguish different computing nodes, and the permissions can be used to control the access and operation permissions of the computing nodes. For example, the computing node A is assigned an identifier of "Node001" and permissions of "high-level permissions", allowing it to access and process high-level computing tasks; the computing node B is assigned an identifier of "Node002" and permissions of "intermediate-level permissions", allowing it to access and process intermediate-level computing tasks.

[0040] S100. Obtain the energy information of the distributed energy system. The distributed energy includes one or more of solar energy, wind energy, and water energy, and the energy information includes energy output and energy reserve.

[0041] Exemplarily, in this embodiment, distributed energy includes solar energy, wind energy, and water energy. For different distributed energy sources, different energy collection devices preset in the distributed energy system can be used to collect the distributed energy. Specifically, for solar energy, photovoltaic panels are installed on the roofs of buildings, open spaces, etc. to convert light energy into direct current electrical energy. The output power of the photovoltaic panels is monitored in real time by a power sensor to obtain the energy output information of solar energy. At the same time, the electrical energy of solar energy is stored through a energy storage device (such as a battery energy storage unit), and the remaining capacity of the energy storage device is monitored in real time by a battery management system to obtain the energy reserve information. For wind energy, a wind turbine is used to convert wind energy into alternating current electrical energy, and the output power of the wind turbine is monitored in real time by an electricity meter to obtain the energy output information of wind energy. The energy storage device is also used to store the electrical energy of wind energy, and its remaining capacity is also monitored in real time by a battery management system. For water energy, a water turbine drives a generator to generate electrical energy, and the output power of the generator is monitored in real time by a power sensor to obtain the energy output information of water energy. The energy storage device is used to store the electrical energy of water energy, and its remaining capacity is also monitored in real time by a battery management system.

[0042] It can be understood that after the relevant energy is collected, in order to ensure normal use, the collected distributed energy also needs to be preprocessed. The preprocessing includes at least one of voltage adjustment, frequency conversion, and rectification and filtering, so as to meet the requirements of subsequent energy storage and use. At the same time, it is also convenient to use sensors to monitor the operating status of the energy collection device and the energy output data in real time.

[0043] In some embodiments, when monitoring the state of an energy harvesting device, various monitoring methods can be adopted. For example, first, the state data of the energy harvesting device, such as temperature, impedance, voltage, current, etc., are normalized. Subsequently, the normalized values are used as their grayscale values, and the state data are filled into a preset image area as pixel points according to the grayscale values to obtain a state image. Then, the state image is input into a pre-trained neural network, and a state evaluation value is obtained through its softmax function. When the state evaluation value is within a preset range, it is determined that the state of the device is normal; otherwise, it is judged as abnormal. Or a state matrix is constructed based on the normalized state data, and then the correlation value between the state matrix at the current moment and the state matrix at the same historical moment is calculated. When the correlation value is within a preset range, it is judged as normal; otherwise, it is judged as abnormal. The information interaction efficiency between the energy storage device and the energy harvesting device can also be calculated to judge the states of the energy harvesting device and the energy storage device. For example, the information exchange frequency, the amount of information in the communication packet, and the envelope composition of the recorded information packet of the information exchange interface corresponding to the energy storage device and the energy harvesting device are calculated. Subsequently, a communication state matrix is constructed based on the information exchange frequency, the amount of information, and the envelope composition. Then, the correlation value or similarity value between the communication state matrix and the standard state matrix is calculated. When the correlation value or similarity value is within a preset range, it is determined that the energy storage device and the energy harvesting device are normal; otherwise, it is judged as abnormal. Of course, to meet the monitoring requirements, those skilled in the art can also select other feasible monitoring methods other than the examples described in this application.

[0044] S200. Obtain the state information of the distributed computing power nodes, where the distributed computing power nodes include electronic devices with computing capabilities, and the state information includes the computing capabilities and task execution status of the distributed computing power nodes.

[0045] Exemplarily, the distributed computing power nodes include electronic devices with computing capabilities such as servers, personal computers, and industrial control computers. By installing performance monitoring software on the computing power nodes, the computing capability information of the computing power nodes, such as CPU usage rate, memory occupancy rate, and graphics card performance, can be obtained in real time. At the same time, the task execution status of the computing power nodes is obtained in real time, including the number of tasks being run, task types, task progress, etc. For example, for a server computing power node, the performance monitoring software shows that its CPU usage rate is 60%, its memory occupancy rate is 70%, and the task management module shows that it is running 5 tasks, including 2 data processing tasks and 3 model training tasks, where 1 data processing task has been completed 50%, 2 model training tasks have been completed 30%, and the other 2 tasks have just started running.

[0046] Understandably, when performing computing power distribution, the electronic device can be first upgraded in hardware and adapted in software to enable it to operate stably in a distributed energy supply environment. Then, the adapted device can be used as a distributed computing power node, which is connected to the distributed computing power network through a wired or wireless communication network, and a unique identifier and permission are assigned to each node to ensure secure communication and smooth data interaction between nodes; or through a software-defined method, its computing resources can be integrated into a computing power network that can be uniformly scheduled to achieve efficient application of computing power.

[0047] S300. Match a target node based on the energy information, the status information, and the target task, where the target node is the distributed computing power node with the highest matching degree to the target task. The S300 includes:

[0048] S310. Extract the task information of the target task, where the task information at least includes energy consumption, computational complexity, and time constraint.

[0049] Exemplarily, when a user submits a target task, the key information of the task is extracted through a task parsing module. For example, for a deep learning model training task, its estimated energy consumption is extracted as 100 kWh, the computational complexity is 10^12 floating-point operations, and the time constraint is to be completed within 48 hours.

[0050] S320. Dynamically allocate the target task to the target node according to the status information, the energy information, and the task information through a multi-objective scheduling algorithm.

[0051] Exemplarily, according to the status information of the computing power node (such as computing power, task execution situation), the energy information of the energy system (such as energy output, energy reserve), and the task information of the target task (such as energy consumption, computational complexity, time constraint), the matching degree between each computing power node and the target task is calculated through a multi-objective scheduling algorithm. The computing power node with the highest matching degree is the target node, and the target task is dynamically allocated to the target node for execution.

[0052] For example, the computing power of computing power node A is 80 TFLOPS, the current task execution situation is relatively easy, and the energy reserve is sufficient; the computing power of computing power node B is 120 TFLOPS, but the current task execution situation is relatively busy, and the energy reserve is less. The target task requires a computing power of 100 TFLOPS, an estimated energy consumption of 50 kWh, and a time constraint of being completed within 24 hours. Through calculation by the multi-objective scheduling algorithm, the matching degree between computing power node A and the target task is higher than that between computing power node B and the target task. Therefore, computing power node A is selected as the target node, and the target task is allocated to computing power node A for execution.

[0053] In some examples, the matching degrees of various parameters are calculated as follows:

[0054] Computing power matching degree where C node is the currently available computing power (TFLOPS) of the node, and C task is the computing power requirement (TFLOPS) of the task.

[0055] Energy reserve matching degree where E node is the remaining available energy (kWh) of the node, and E task is the estimated energy consumption (kWh) of the task.

[0056] Task load matching degree where T running is the number of tasks being executed by the node, and T max is the maximum concurrent task capacity of the node.

[0057] Time constraint matching degree where t predicted is the estimated task completion time (hours) of the node, and t task is the task time constraint (hours).

[0058] Comprehensive matching degree S z can be obtained by weighted summation:

[0059] S z = w 1 × S compute + w 2 × S energy + w 3 × S load + w 4 × time

[0060] where w1 - w4 are weight coefficients, and their sum is 1.

[0061] By quantifying the matching degrees of computing power, energy reserve, task load, and time constraint, and combining weight allocation, this method can accurately select the optimal node to ensure efficient task execution and reasonable energy utilization.

[0062] It can be understood that calculating the matching degree of each performance parameter and task parameter and obtaining the comprehensive matching degree by weighting is a conventional technical means for those skilled in the art. This application does not limit it here. Those skilled in the art can, under the inspiration of this application, independently select the scheduling algorithm for allocating target nodes.

[0063] The computing power demand prediction technology is an important part of distributed energy management. Its purpose is to predict the development trend of computing power demand by analyzing historical data and energy factors, so as to provide a scientific basis for the reasonable allocation of computing power resources. The existing computing power demand prediction methods mainly include experience-based prediction, statistical model prediction, machine learning prediction, etc.

[0064] Although the existing computing power demand prediction technologies can provide references for distributed energy management to a certain extent, they generally have deficiencies, resulting in poor computing power demand prediction effects. For example, existing prediction models often simply extrapolate based on historical data, resulting in a large deviation between the prediction results and the actual situation; or many prediction models highly rely on the accumulation of a large amount of historical data. For regions with incomplete data collection or lack of historical data, the applicability and accuracy of the models are limited; or the growth of computing power demand is affected by various factors such as energy supply capacity and network effects. Existing models often have difficulty accurately simulating the dynamic effects of these factors on computing power demand.

[0065] Therefore, the prediction methods in the existing technology have poor prediction effects on computing power demand. The method in the embodiment of this application further includes the following steps to improve the problem of poor prediction effects of the prediction methods in the existing technology on computing power demand.

[0066] S410. Obtain the characteristic factors for the growth of computing power demand. The characteristic factors include a first characteristic factor and a second characteristic factor. The first characteristic factor is used to characterize the driving effect of energy supply capacity on computing power demand, and the second characteristic factor is used to characterize the collaborative growth of computing power demand caused by network effects.

[0067] The growth of computing power demand is affected by various factors, among which the most crucial ones are energy supply capacity and network effects. Energy supply capacity determines the upper limit of computing power resources, while network effects affect the collaborative growth of computing power demand.

[0068] Energy supply capacity is the basis for the growth of computing power demand. The improvement of energy supply capacity can support the operation of more computing power devices, thus directly promoting the growth of computing power demand. For example, the increase in the power generation of distributed energy, the increase in the remaining capacity of energy storage systems, etc., will enhance the energy supply capacity and further promote the growth of computing power demand.

[0069] Network effect refers to the mutual influence and collaborative growth of computing power demand among users. As more and more users access the computing power network, the computing power resources and service quality that each user can obtain will be improved, thus attracting more users to join, forming a collaborative growth of computing power demand. For example, the increase in the number of users of a cloud computing platform will drive the overall increase in the computing power demand of the platform. The improvement of the task collaboration efficiency between computing power nodes can also lead to a positive feedback effect on the growth of the overall computing power demand.

[0070] Therefore, in this application, the driving effect of energy supply capacity on computing power demand is taken as the first characteristic factor, and the collaborative growth of computing power demand caused by network effects is taken as the second characteristic factor. The first characteristic factor and the second characteristic factor jointly determine the growth trend of computing power demand.

[0071] S420. Construct a prediction model based on the characteristic factors, where the prediction model is used to predict the computing power demand:

[0072] Exemplarily, a prediction model is constructed according to the extracted characteristic factors. The expression of the prediction model is:

[0073]

[0074] This prediction model comprehensively considers the influence of energy supply capacity and network effects on computing power demand. Among them, C(t) is the computing power demand at time t (TFLOPS / hour), E(t) is the energy supply capacity at time t (kW), E max is the maximum available energy capacity of the system, p is the first characteristic factor, reflecting the driving effect of energy supply capacity on computing power demand, and q is the second characteristic factor, reflecting the collaborative growth of computing power demand caused by network effects.

[0075] This prediction model describes the dynamic change of computing power demand over time, as well as the contributions of energy supply capacity and network effects to the growth of computing power demand. Specifically, the first term p·(E max -E(t)) describes the driving effect of energy supply capacity on computing power demand. As the energy supply capacity increases, the unused energy capacity (E max -E(t)) decreases, and the growth rate of computing power demand also changes accordingly. The second term describes the collaborative growth of computing power demand caused by network effects. As the energy supply capacity increases, the network effects gradually emerge, and the growth rate of computing power demand is affected by network effects. At the same time, this prediction model regards the growth of computing power demand as a dynamic system jointly affected by energy supply capacity and network effects, which can accurately capture the non-linear characteristics of computing power demand and reflect the complex relationship of computing power demand growth.

[0076] S430. Obtain the matching degree between the current energy parameters and the historical energy parameters. The energy parameters include distributed energy generation power, remaining capacity of the energy storage system, energy volatility, etc. The changes of these parameters directly affect the energy supply capacity and thus affect the growth of computing power demand.

[0077] Exemplarily, by deploying various sensors and monitoring devices in a distributed energy system, parameters such as the power generation of distributed energy, the remaining capacity of the energy storage system, and the energy volatility are obtained in real time. For example, the power generation of solar energy is obtained through a power sensor on a photovoltaic panel, the remaining capacity of the energy storage system is obtained through the management system of the energy storage battery, and the energy volatility is calculated through data analysis software. The currently obtained energy parameter data is compared with the historical energy parameter data, and the similarity between the two is quantified through the Pearson correlation coefficient to generate a matching degree score. For example, calculate the Pearson correlation coefficient between the current solar power generation curve and the solar power generation curves of each day in the past week to obtain a matching degree score to evaluate the similarity between the current energy situation and the historical energy situation. The higher the score, the higher the similarity between the current energy parameters and the historical energy parameters.

[0078] Exemplarily, in some embodiments, the calculation method is as follows:

[0079]

[0080] where x i and y i are respectively the i-th data points of the current energy parameter and the historical energy parameter, and are respectively the average values of the current energy parameter and the historical energy parameter, and ρ is the similarity.

[0081] S440. Generate first predicted computing power data according to the matching degree and the historical computing power data, where the first predicted computing power data includes the computing power requirements for multiple time periods.

[0082] Exemplarily, according to the matching degree between the current energy parameter and the historical energy parameter, combined with the historical computing power data, the first predicted computing power data is generated through weighted average or other suitable prediction methods to predict the computing power requirements for multiple future time periods. For example, according to the matching degree between the current energy parameter and the historical energy parameter being 80%, combined with the computing power requirement data for the corresponding time periods in the past week, the first predicted computing power data is generated through the weighted average method to predict the computing power requirements for each hour within the next 24 hours. Or directly according to the matching degree score, select the computing power data corresponding to the historical energy parameter with the highest matching degree with the current energy parameter as the reference data. Use the computing power requirements in the reference data as the first predicted computing power data. It can be understood that this data includes the computing power requirements for multiple time periods and is used to preliminarily predict the trend of computing power requirements in the current situation.

[0083] S450. Obtain the initial value of the characteristic factor according to the first predicted computing power data, the current actual computing power data, and the prediction model. The S450 includes:

[0084] S451. Build an intermediate model based on historical energy data.

[0085] Exemplarily, using historical energy data, construct an intermediate model for predicting computing power demand through time series analysis or other suitable modeling methods. For example, based on the energy data of the past year, use the ARIMA model to construct an intermediate model to predict the trend of computing power demand in the future for a period of time. In some other examples, the model can also be composed of multiple predictive models weighted, such as BP neural network model, fuzzy neural network model, and adaptive probabilistic neural network model, etc. The weights of each model are determined according to the accuracy of their prediction results. The higher the accuracy, the greater the weight.

[0086] S452. Generate second predicted computing power data through the intermediate model.

[0087] Exemplarily, input the current energy data into the intermediate model to generate second predicted computing power data. For example, input the current solar power generation data, etc. into the intermediate model to obtain the predicted computing power demand data for each hour within the next 24 hours.

[0088] S453. Correct the predicted value according to the deviation between the current actual computing power data and the second predicted computing power data to obtain third predicted computing power data. The S453 includes:

[0089] S4531. Obtain the average deviation between the first data in the second predicted computing power data and the current actual computing power data, where the first data is the data corresponding to the current actual computing power data in the second predicted computing power data.

[0090] Exemplarily, select the data corresponding to the current actual computing power data from the second predicted computing power data and calculate the average deviation between the two. For example, select the data for each hour within the past 24 hours in the second predicted computing power data, compare it with the current actual computing power data, and calculate the average deviation to be 8 TFLOPS / hour.

[0091] S4532. Correct the second predicted computing power data according to the average deviation to obtain the third predicted computing power data.

[0092] Exemplarily, according to the calculated average deviation, use a weighted algorithm to perform dynamic compensation on it to generate third predicted computing power data. For example, according to the average deviation of 8 TFLOPS / hour, compensate the second predicted computing power data through the weighted algorithm to obtain the third predicted computing power data and improve the accuracy of the prediction. In some examples, the following algorithm can be used for calculation:

[0093] C 修正 (t) = C 预测 (t) + α · deviation

[0094] Among them, C 修正 (t) is the corrected computing power demand, C 预测 (t) is the second predicted computing power data, α is the weighting coefficient, which is determined by those skilled in the art according to requirements. In this embodiment, it can be 0-1, such as 1, and the deviation is the average deviation between the second predicted computing power data and the current actual computing power data.

[0095] S454. Obtain the initial value of the characteristic factor according to the data of the trend nodes in the third predicted computing power data and the prediction model.

[0096] Exemplarily, the trend nodes include the energy supply saturation time point, the time point when the computing power demand changes from linear growth to exponential growth, and the critical point where the network effect triggers the collaborative growth of the computing power demand.

[0097] The energy supply saturation time point refers to the time point when the energy supply reaches the maximum value. The time point when the computing power demand changes from linear growth to exponential growth refers to the time point when the growth rate of the computing power demand changes from linear growth to exponential growth. The critical point where the network effect triggers the collaborative growth of the computing power demand refers to the time point when the network effect begins to significantly affect the growth of the computing power demand.

[0098] Substitute the data of these nodes into the prediction model to solve the initial value of the characteristic factor. For example, extract the energy supply inflection point data and the computing power demand peak data in the third predicted computing power data, and substitute them into the prediction model to calculate the initial value of the characteristic factor.

[0099] S460. Predict the computing power demand according to the prediction model corresponding to the initial value of the characteristic factor.

[0100] Exemplarily, substitute the obtained initial value of the characteristic factor into the prediction model to predict the computing power demand in a future period. For example, substitute the initial value of the characteristic factor (such as p = 0.6, q = 0.4) into the prediction model to predict the daily computing power demand in the next 7 days, providing a basis for the scheduling and allocation of distributed computing power resources.

[0101] S500. When it is detected that the energy supply volatility exceeds the threshold, trigger the dynamic scheduling protocol.

[0102] Exemplarily, the energy supply volatility is monitored in real time through the monitoring system. When the volatility exceeds the preset threshold, the dynamic scheduling protocol is triggered. For example, when it is detected that the volatility of wind power generation exceeds 30%, the dynamic scheduling protocol is triggered.

[0103] Exemplarily, according to the task importance and energy consumption situation of the computing power nodes, unnecessary high-power computing power nodes are shut down to reduce the overall energy consumption and ensure the normal operation of critical tasks. For example, high-power computing power nodes performing non-urgent tasks are shut down, such as computing power nodes for low-priority data analysis tasks in progress.

[0104] Exemplarily, tasks originally executed on high-power computing power nodes are migrated to computing power nodes with stable energy supply according to the characteristics and requirements of the tasks to ensure the continuity and stability of the tasks. For example, an ongoing artificial intelligence training task is migrated from a computing power node affected by energy fluctuations to another computing power node with stable energy supply.

[0105] Exemplarily, according to the computing power node distribution and task execution situation after task migration, the collaborative growth of computing power requirements caused by network effects is re-evaluated, and the second characteristic factor (network effect coefficient) is corrected. For example, after task migration, the collaboration mode among computing power nodes changes. By analyzing the new task distribution and collaboration frequency, and according to the change in computing power requirements after task migration, a new network effect coefficient is calculated and substituted into the prediction model to improve the accuracy of the prediction model.

[0106] In summary, the present application can at least achieve one or more of the following beneficial effects:

[0107] 1. Improve energy utilization efficiency: Ineffectively utilized distributed green energies such as solar energy, wind energy, and tidal energy are efficiently converted into valuable computing power resources, reducing energy waste, achieving maximum benefit utilization, and improving the comprehensive utilization efficiency of distributed green energies.

[0108] 2. Reduce computing power costs: Low cost is a basic requirement for the popularization of artificial intelligence technology. Using distributed energy and ordinary devices to build a distributed computing power network avoids various high costs of large-scale construction and operation of centralized data centers. Such distributed computing power devices are easier to deploy, solving the problem of large upfront investment in centralized computing power construction with small capital investment (obtaining large benefits with small capital investment) and reducing the threshold for obtaining computing power.

[0109] 3. Enhance system flexibility and stability: The distributed energy and computing power architecture enables the system to be flexibly adjusted according to changes in energy and demand, enhancing the system's risk resistance and stability.

[0110] 4. Promote the integrated development of energy and information technology: The present application provides new computing power approaches and methods for the in-depth integration of the energy field and the information technology field, improving the utilization efficiency of distributed green energies and making up for the deficiencies of centralized computing power resources.

[0111] Through the description of the above embodiments, those skilled in the art can clearly understand that the diagnostic methods in the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases, the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence or the part that contributes to the related technology, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to enable a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0112] Through the description of the above embodiments, those skilled in the art can clearly understand that for the convenience and conciseness of description, only the above division of each functional module is used as an example. In actual applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above.

[0113] In several embodiments provided in the present application, it should be understood that the disclosed device and method can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in electrical, mechanical or other forms.

[0114] The units described as separate components may or may not be physically separated. The components displayed as units may be one physical unit or multiple physical units, that is, they can be located in one place or distributed to multiple different places. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0115] In addition, each functional unit in the various embodiments of the present application can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware.

[0116] The above content is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for converting distributed energy into distributed computing power, characterized in that: The method comprises: Acquiring energy information of a distributed energy system, wherein the distributed energy includes one or more of solar energy, wind energy and hydropower, and the energy information includes energy output and energy reserve; Acquire status information of a distributed computing power node, wherein the distributed computing power node includes an electronic device with computing capability, and the status information includes computing capability and task execution status of the distributed computing power node; Based on the energy information, the state information and the target task matching target node, the target node is the distributed computing power node with the highest matching degree with the target task.

2. The method according to claim 1, characterized in that Before acquiring the energy information of the distributed energy system, the method further includes: Constructing the distributed energy system based on the distributed energy and the energy storage unit, wherein the energy storage unit includes at least one of a battery energy storage unit and a supercapacitor energy storage unit; The distributed computing power nodes are constructed based on electronic devices with computing capabilities and configuration strategies, and the configuration strategies at least include assigning identification and permissions to each of the distributed computing power nodes.

3. The method according to claim 2, characterized in that Before acquiring the energy information of the distributed energy system, the method further includes: Collecting the distributed energy through a preset energy collection device in the distributed energy system; The collected distributed energy is preprocessed, and the preprocessing includes at least one of voltage adjustment, frequency conversion and rectification and filtering.

4. The method according to claim 2, characterized in that: The matching of the target node based on the energy information, the state information and the target task includes: Extracting task information of the target task, wherein the task information at least includes energy consumption, computational complexity and time constraint; According to the state information, the energy information and the task information, the target task is dynamically allocated to the target node through a multi-objective scheduling algorithm.

5. The method according to claim 1, characterized in that: The method further comprises: Obtain characteristic factors of computing power demand growth; Building a prediction model based on the characteristic factors, wherein the prediction model is used to predict computing power requirements; Obtain the matching degree between current energy parameters and historical energy parameters; Generate first predicted computing power data according to the matching degree and the historical computing power data, wherein the first predicted computing power data includes computing power requirements for multiple time periods; Obtaining an initial value of the characteristic factor according to the first predicted computing power data, the current actual computing power data, and the prediction model; The computing power demand is predicted according to the prediction model corresponding to the initial value of the characteristic factor.

6. The method according to claim 5, characterized in that The characteristic factors include a first characteristic factor and a second characteristic factor. The first characteristic factor is used to characterize the driving effect of energy supply capacity on computing power demand, and the second characteristic factor is used to characterize the coordinated growth of computing power demand caused by network effects.

7. The method according to claim 6, characterized in that The expression of the prediction model is: Where C(t) is the computing power demand at time t, E(t) is the energy supply capacity at time t, and E max is the maximum available energy capacity of the system, p is the first characteristic factor, and q is the second characteristic factor.

8. The method according to claim 5, characterized in that The obtaining the initial value of the characteristic factor according to the first predicted computing power data, the current actual computing power data and the prediction model includes: Build an intermediate model based on historical energy data; Generate second predicted computing power data through the intermediate model; According to the deviation between the current actual computing power data and the second predicted computing power data, the predicted value is corrected to obtain the third predicted computing power data; An initial value of the characteristic factor is obtained based on the data of the trend node in the third predicted computing power data and the prediction model.

9. The method according to claim 8, characterized in that The step of correcting the predicted value according to the deviation between the current actual computing power data and the second predicted computing power data to obtain the third predicted computing power data includes: Obtaining an average deviation between first data in the second predicted computing power data and the current actual computing power data, wherein the first data is data in the second predicted computing power data corresponding to the current actual computing power data; The second predicted computing power data is corrected according to the average deviation to obtain the third predicted computing power data.

10. The method according to claim 8, characterized in that The trend nodes include the time point when energy supply is saturated, the time point when computing power demand changes from linear growth to exponential growth, and the critical point when network effects trigger the coordinated growth of computing power demand.

Images