Energy scheduling method and system of data center and electronic equipment

Abstract

The invention discloses an energy scheduling method and system of a data center and electronic equipment, and relates to the technical field of data processing, and the method comprises the following steps: obtaining the operation data of a battery unit in an energy storage system of a solar power station, and extracting an energy accumulation node according to the operation data of the battery unit; identifying a current loop extension direction of the energy accumulation node based on a preset wire connection relationship to obtain an energy echo region; cutting off an echo channel corresponding to the energy echo region according to a preset cutting-off frequency to obtain a first cutting-off region; extracting residual current attenuation characteristics of the first cut-off region, and performing channel recovery on the first cut-off region based on the residual current attenuation characteristics to obtain a second cut-off region; the time sequence path of electric energy flowing in the battery unit is dynamically adjusted based on the second cut-off region, and the problem that local energy accumulation occurs in the charging and discharging conversion process of the energy storage system, so that current oscillation occurs in part of nodes is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and more particularly, to an energy scheduling method and system of a data center and an electronic device. BACKGROUND

[0002] With the widespread application of renewable energy, solar power stations, as an important form of green energy, have been widely deployed worldwide. However, solar power generation has the characteristics of intermittency and instability, and its power generation is affected by factors such as weather, season and time, making it difficult to achieve stable power supply. Therefore, how to reasonably schedule energy in a solar power station to achieve a balance between power generation, energy storage and load has become an important issue in the operation of a solar power station.

[0003] However, the topology of existing solar energy storage devices is usually fixed at the design stage, and its wire connection mode, switch matrix layout and current flow path are static structures, lacking the ability to dynamically adapt during operation. When local energy accumulation occurs in the charging and discharging conversion process of the energy storage unit, the current will often flow along the established fixed path repeatedly, and cannot be discharged in time through structural reconstruction or path shunting to achieve energy dissipation, thereby forming a local loop echo phenomenon. When this echo effect persists, local nodes will experience current oscillation, even if the energy accumulation has been partially eliminated, the oscillation will continue due to factors such as wire impedance, parasitic capacitance and phase delay, resulting in incomplete energy release. Long-term energy echo not only reduces the energy storage conversion efficiency, but also causes node temperature rise, increased local loss and disturbance to the overall current stability. To address the above problems, the present application provides a solution. SUMMARY

[0004] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present application provide an energy scheduling method and system of a data center and an electronic device, which regulate the energy echo area of the battery unit to solve the problem of current oscillation at some nodes caused by local energy accumulation in the charging and discharging conversion process of the energy storage system.

[0005] To achieve the above-mentioned purposes, the present application provides the following technical solutions: The application discloses an energy scheduling method of a data center, and comprises the following steps: obtaining battery unit operation data in a solar power station energy storage system, and extracting an energy accumulation node according to the battery unit operation data; identifying a current loop extension direction of the energy accumulation node based on a preset wire connection relationship, and obtaining an energy echo area; cutting an echo channel corresponding to the energy echo area according to a preset cutting frequency, and obtaining a first cutting area; extracting a residual current decay characteristic of the first cutting area, and recovering a channel of the first cutting area based on the residual current decay characteristic, and obtaining a second cutting area; and dynamically adjusting a time sequence path of internal electric energy flow of the battery unit based on the second cutting area.

[0006] In a preferred embodiment, the extracting the energy accumulation node according to the battery unit operation data specifically comprises: extracting a multi-channel voltage sequence and a current sequence of the battery unit based on the battery unit operation data, and obtaining a battery operation state matrix; extracting a charge accumulation rate value and a discharge fluctuation amplitude of each battery unit based on the battery operation state matrix; calculating an energy difference value in a unit time based on the charge accumulation rate value and the discharge fluctuation amplitude, and identifying an initial energy accumulation point based on the energy difference value; obtaining a physical connection topology among the battery units, and performing cluster analysis on the initial energy accumulation point based on the physical connection topology, and obtaining an energy accumulation area; performing energy density evaluation on the energy accumulation area, and screening the initial energy accumulation point according to an evaluation result, and obtaining the energy accumulation node.

[0007] In a preferred embodiment, the identifying the current loop extension direction of the energy accumulation node based on the preset wire connection relationship, and obtaining the energy echo area specifically comprises: obtaining a wire connection topology graph of the solar power station energy storage system, and starting from the energy accumulation node, traversing along a forward direction and a reverse direction of the wire connection topology graph respectively, and obtaining a plurality of current flow paths; calculating an impedance value and a current phase delay value of each current flow path, and identifying a closed-loop current loop based on the impedance value and the current phase delay value; calculating an active power density vector among nodes of the closed-loop current loop, and obtaining a main loop and a branch loop of the closed-loop current loop based on the active power density vector; superimposing the energy accumulation nodes of the main loop, and obtaining an energy echo core area; and expanding a boundary of the energy echo core area based on the branch loop, and obtaining the energy echo area.

[0008] In a preferred embodiment, the expanding the boundary of the energy echo core area based on the branch extension direction, and obtaining the energy echo area specifically comprises: extracting a current oscillation characteristic of the branch loop, and performing cluster analysis on the branch loop according to the current oscillation characteristic, and obtaining an oscillation cluster; expanding the boundary along a direction of the oscillation cluster with the energy echo core area as the center, and obtaining a first extension zone; calculating an avoidance radius based on the first extension zone, and adjusting the boundary of the energy echo core area according to the avoidance radius, and obtaining the energy echo area.

[0009] In a preferred embodiment, the echo channel corresponding to the energy echo area is cut off according to a preset cut-off frequency to obtain a first cut-off area, specifically: the echo channel in the energy echo area is obtained, and the channel is classified according to the preset harmonic component according to the current amplitude and frequency characteristics of the echo channel to obtain the classified echo channel; the harmonic resonance characteristics of the classified echo channel are obtained, and a plurality of cut-off frequency points are preset according to the harmonic resonance characteristics; for each classified echo channel, the channel is cut off in turn according to the preset cut-off frequency point to obtain the first cut-off area.

[0010] In a preferred embodiment, the residual current decay characteristics of the first cut-off area are extracted, and the channel of the first cut-off area is recovered based on the residual current decay characteristics to obtain a second cut-off area, specifically: the residual current waveform and current response curve of the first cut-off area are collected, and the residual current waveform is exponentially fitted to obtain a residual decay time; the residual energy dissipation rate is calculated according to the residual decay time, and the recoverable channel is identified based on the dissipation rate distribution map; the rising rate of the voltage excitation is adjusted according to the slope change of the current response curve, and the voltage excitation is applied to the recoverable channel to obtain the second cut-off area.

[0011] In a preferred embodiment, the timing path of the internal electric energy flow of the battery unit is dynamically adjusted based on the second cut-off area, specifically: the three-dimensional coordinates of the second cut-off area are obtained to obtain the spatial distribution matrix of the second cut-off area; the preset energy distribution matrix of the matrix battery unit is reconstructed based on the spatial distribution of the second cut-off area to obtain the reconstructed energy distribution matrix; based on the reconstructed energy distribution matrix, a variable length time window is allocated to each battery unit, and voltage ripple and current total harmonic distortion data are collected in real time; the start delay and duration of the timing path are dynamically calibrated according to the voltage ripple and harmonic data to obtain an optimized set of electric energy flow timing paths.

[0012] The technical effects and advantages of the energy scheduling method, system and electronic equipment of the data center of the application are as follows: The application extracts energy accumulation nodes, accurately identifies the concentrated area of energy in space, then determines the extension direction of the current loop based on the preset wire connection relationship, constructs an energy echo area, thereby revealing the circulating echo effect of the current in the complex wire network, then cuts off the channel in the energy echo area according to the preset frequency to form a first cut-off area to interrupt the high-frequency energy reflux, on this basis, extracts the attenuation characteristics of the residual current, implements voltage excitation on the recoverable channel to obtain a second cut-off area, realizes the step-by-step recovery and optimization of the energy loop, and finally, based on the second cut-off area, dynamically adjusts the timing path of the internal electric energy flow of the battery unit, so that the energy is orderly distributed and timing balanced among different units. Through the above steps, not only the energy echo and high-frequency oscillation problem caused by local accumulation is effectively eliminated, but also the efficiency and stability of energy transmission are improved, the collaborative optimization of the energy storage network in the space and time dimensions is realized, and the intelligent scheduling and adaptive operation ability of the data center in the complex electric energy environment is significantly enhanced. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 It is a flowchart of the energy scheduling method of the data center of the application.

[0014] Figure 2 It is a structural schematic diagram of the energy scheduling system of the data center of the application.

[0015] Figure 3 It is an electronic device structure schematic diagram of the energy scheduling method of the data center of the application. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0017] Embodiment 1, Figure 1 The energy scheduling method of the data center of the application is given, including the following steps: S1, obtaining battery unit operation data in a solar power station energy storage system, and extracting energy accumulation nodes according to the battery unit operation data; In this example, the energy accumulation nodes are extracted according to the battery unit operation data, specifically: Based on the battery unit operation data, the multi-channel voltage sequence and the current sequence of the battery unit are extracted, and the battery operation state matrix is obtained; Based on the battery operation state matrix, the charge accumulation rate value and the discharge fluctuation amplitude of each battery unit are extracted; calculate the energy difference value in a unit of time based on the charge accumulation rate value and the discharge fluctuation amplitude, and identify the initial energy accumulation point based on the energy difference value; obtain a physical connection topology between the battery units, and perform clustering analysis on the initial energy accumulation point based on the physical connection topology to obtain an energy accumulation area; perform energy density evaluation on the energy accumulation area, and perform screening on the initial energy accumulation point according to the evaluation result to obtain an energy accumulation node.

[0018] It should be noted that in the solar energy storage scenario, each battery unit outputs real-time multi-dimensional operation data, such as terminal voltage, current, temperature, charging and discharging state identifier, SOC (state of charge), etc. In order to extract the multi-channel voltage and current sequence, first, the original operation data in a certain sampling period (for example, once per second or once per millisecond) needs to be obtained from the battery management device. Then, the voltage and current data of each battery unit at different sampling times are recorded by channel, for example, in a 10-minute monitoring period, a voltage sequence and a current sequence varying with time can be formed. After time alignment, noise filtering and normalization processing, a multi-dimensional matrix composed of time, channel and power parameters can be constructed, which is called a battery operation state matrix. It directly reflects the input and output behavior of the power of each battery unit in the entire sampling period. For example, in an array of 100 battery units, each unit corresponds to a row of time sequence data, and the columns of the matrix correspond to the voltage, current, temperature and other signal channels.

[0019] After obtaining the battery operation state matrix, the trends of voltage and current varying with time can be extracted. The charge accumulation rate value can be understood as the power accumulation speed reflected by the growth trend of current with time in the charging stage. The specific method is as follows: the current sequence is segmented by time window, the rising trend or stability of current in each time period is calculated, and the average level of voltage change is integrated, so as to reflect the charge accumulation speed of the battery unit in this stage. The discharge fluctuation amplitude corresponds to the fluctuation of current output in the discharging stage, which can be obtained by analyzing the fluctuation range of current curve and the voltage drop rate in the discharging period. For example, if the voltage of a certain battery unit drops smoothly and the current fluctuation amplitude is small within 10 seconds, it indicates that the energy output of the unit is stable; if the current fluctuation is severe, it means that the discharging characteristics are uneven and there may be local energy accumulation. Finally, each battery unit will obtain two characteristic parameters: one represents the power accumulation ability (charge accumulation rate), and the other represents the energy release stability (discharge fluctuation amplitude).

[0020] Further, when the charge accumulation rate and discharge fluctuation amplitude of each battery cell are known, the energy change of each battery cell in a unit time can be compared. In simple terms, the difference between the "energy absorbed" and the "energy released" of the battery in the same time period is compared. If a certain battery cell exhibits a high accumulation rate, a low discharge stability, i.e. fast charging but slow discharging, and a significant energy retention in multiple consecutive time periods, this cell can form a local energy accumulation phenomenon. Such a cell is identified as an initial energy accumulation point. For example, in a 100-node battery energy storage module, if the batteries numbered 45 and 46 have a larger energy input than output and a larger discharge curve fluctuation in multiple time periods, they will be identified as initial energy accumulation points. These points are often the source of subsequent energy imbalance.

[0021] Secondly, battery cells are usually connected in series and parallel to form a whole energy storage array, so the connection relationship of each cell can be determined in the design stage. In order to understand the spatial distribution of energy in the array, the electrical connection relationship between the cells needs to be extracted from the design drawing or monitoring platform, such as which batteries are adjacent, which are in the same branch, and which are in the same voltage level. Then, combined with the initial energy accumulation points identified above, a clustering algorithm (such as hierarchical clustering based on adjacency relationship and energy feature similarity) is used to group multiple accumulation points that are spatially related and have similar energy difference trends. These aggregated areas are defined as energy accumulation zones. In simple terms, an energy accumulation zone refers to a local area in the energy storage array where multiple battery cells simultaneously exhibit the common characteristics of energy accumulation, insufficient discharge, or over-saturation of charging.

[0022] Finally, after obtaining the energy accumulation zones, it is necessary to further determine which regions have an energy concentration degree that reaches the threshold of dispatch intervention. This can be done by calculating the energy density in the region, i.e. the energy accumulation degree or the total amount of energy remaining per unit number of batteries in the same time period. The comprehensive indicators of voltage, current, and temperature are considered to reflect the compactness of energy distribution. Regions with high energy density and fast change rate indicate severe local energy retention. Subsequently, based on the evaluation results, the initial accumulation points in the region are screened, and cells with slight fluctuations or small energy differences are removed, leaving only those with the highest energy concentration and the greatest impact on overall energy flow. These points are the energy accumulation nodes.

[0023] S2, based on the preset wire connection relationship, identifying the current loop extension direction of the energy accumulation node to obtain an energy echo zone; In this example, based on the preset wire connection relationship, the current loop extension direction of the energy accumulation node is identified to obtain an energy echo zone, specifically: A lead connection topology graph of a solar power plant energy storage system is acquired, and starting from an energy accumulation node, traversing along the lead connection topology graph in a forward direction and a reverse direction respectively, a plurality of current flow paths are obtained; Impedance values and current phase delay values of each current flow path are calculated, and a closed-loop current loop is identified based on the impedance values and the current phase delay values; Active power density vectors between nodes of the closed-loop current loop are calculated, and a main loop and a branch loop of the closed-loop current loop are obtained based on the active power density vectors; The energy accumulation nodes of the main loop are superimposed to obtain an energy echo core area; Based on the branch loop, the boundary of the energy echo core area is expanded to obtain an energy echo area.

[0024] It should be noted that in the energy storage device of a solar power plant, a large number of battery units are connected to each other in series and parallel through leads, and these connection relationships are usually recorded as electrical topology structure files or wiring diagrams in the design stage. The lead connection topology graph is a graphical representation of the network structure obtained by abstracting the mutual connection relationship of all battery units, busbars, connection leads and measurement nodes in the energy storage device, wherein each node represents an electrical element (such as a battery unit or a busbar), and each connection line represents the electrical connection of a section of lead. This topology graph can clearly describe the possible transmission paths of energy in the entire energy storage architecture. There are usually two ways to obtain this topology graph: one is to directly derive a digitized connection relationship table from engineering design drawings; the other is to identify the electrical connection path in reverse through online monitoring equipment (such as a voltage sampling module and a bus monitoring point). After obtaining the topology graph, starting from the determined energy accumulation node, the lead connection relationship is expanded layer by layer outward. On the one hand, the forward traversal is along the main direction of current flow (i.e. the energy output path), and on the other hand, the reverse traversal is against the flow direction (i.e. the energy return path). Through this forward and reverse traversal, a plurality of complete current flow paths can be obtained. For example, starting from the energy accumulation node numbered 45, the forward path is formed by sequentially passing through the units numbered 46, 47 and 48, and then reaching the busbar; and the reverse path may pass through the parallel branch to return to the units numbered 44 or 43. All these traversal paths are recorded to form the basis for subsequent current analysis.

[0025] After obtaining the multiple current flow paths, the electrical energy conduction characteristics of these paths need to be evaluated. Impedance values are used to reflect the overall degree of hindrance to current flow by the conductors and their connection points, which is usually dependent on the length, material and contact quality of the conductors; while current phase delay values are used to describe the time lag or phase difference when the current is transmitted in the path, which reflects the inductance, capacitance and other effects of the path. By collecting voltage and current signals at different nodes, the equivalent impedance and phase delay characteristics of each path can be calculated. Then, the impedance and phase delay of all paths are analyzed in association. If it is found that several paths are connected at both ends and the phase delay change shows a periodic closing trend, it means that there is energy circulation between these paths, forming a so-called closed-loop current loop. In simple terms, such a loop is like a channel for energy to repeatedly circulate in a certain area, for example, in three parallel branches, one branch forms a partial loop current due to low impedance, causing current to flow back and forth in that branch and adjacent conductors, which is identified as a closed-loop current loop in the topology graph. Such a loop is often a potential area for energy retention and heat generation.

[0026] Further, after identifying the closed-loop current loop, the energy distribution between the nodes in the loop needs to be further analyzed. By comprehensively analyzing the voltage, current and phase at each node, the direction and strength of energy transfer between each pair of adjacent nodes can be obtained, which is called active power density vector. It reflects the real flow trend of energy in the loop. For example, if the power density on several segments of a closed loop is significantly higher than other parts, it means that the segment is the main energy transmission channel. According to the distribution of these vectors, the closed-loop structure can be divided into two categories: one is the main loop, which is the main path of energy flow, usually carrying most of the active power in the loop, and the energy flows most stably and lasts the longest; the other is the branch loop, which is connected to the main loop but has lower power density, and often only participates in energy transmission under certain conditions (such as transient load changes). For example, if the path from node 45-46-47-48 in a closed loop has the strongest energy flow, while the branch from 47-49-50 has weaker flow, the former is called the main loop and the latter is the branch loop. The main loop is responsible for main energy transmission, while the branch mainly affects local energy distribution balance.

[0027] Finally, after identifying the main loop of the closed loop, the energy accumulation nodes distributed on the main loop can be traced back. These nodes are originally the positions of energy concentration or stagnation, and when they are located on the main loop of energy flow, they will have a feedback effect on the entire energy transmission. By superimposing the energy characteristics of all energy accumulation nodes on the main loop, the core area with the most concentrated energy and the most significant energy reflection and redistribution phenomenon can be determined, which is called the energy echo core area. The so-called energy echo core area can be understood as a spatial range in the energy transmission network where the coupling effect of energy accumulation nodes leads to repeated fluctuations and feedback of current and power signals in the local area. The echo here is the multiple reflection and repropagation of electrical energy in this area. For example, in a certain energy storage architecture, if batteries 45, 46, and 47 are all energy accumulation nodes and are located on the main loop, the energy interaction among them will lead to enhanced local current oscillation, forming a concentrated area of energy echo.

[0028] In this example, the boundary of the energy echo core area is expanded based on the extension direction of the branch to obtain the energy echo area, specifically: The current oscillation characteristics of the branch loop are extracted, and the branch loop is analyzed according to the current oscillation characteristics to obtain an oscillation cluster; The boundary is expanded along the oscillation cluster direction with the energy echo core area as the center to obtain a first extension zone; The avoidance radius is calculated based on the first extension zone, and the boundary of the energy echo core area is adjusted according to the avoidance radius to obtain the energy echo area.

[0029] It should be noted that after identifying the closed-loop current loop, there are still several branch circuits outside the main trunk, which often bear the role of local energy regulation and current distribution. In order to further analyze the influence of these branches on energy distribution, it is necessary to extract their current oscillation characteristics. The specific method is as follows: in multiple branch circuits, time series analysis is performed on the current signals at each node, and characteristic parameters such as fluctuation range of current amplitude, oscillation frequency, phase drift and energy attenuation trend are recorded. Through these characteristics, it can be judged whether a certain branch circuit exists periodic energy back-and-forth flow phenomenon. For example, if the current direction of a branch changes repeatedly and the amplitude fluctuates significantly in a short time, it means that the branch has strong energy oscillation behavior. Next, after vectorizing the oscillation characteristics of all branch circuits, similarity analysis or clustering algorithm is used to group branches with similar oscillation characteristics into the same group. These branches with the same oscillation mode, similar frequency and energy reflection characteristics are called oscillation clusters. The "oscillation cluster" actually represents a group of current oscillation channels that are consistent in energy flow behavior. They are often in spatially adjacent or strongly electrically coupled areas. For example, if the 47-49-50 and 46-48-51 branches in a certain energy storage unit produce synchronous fluctuations at similar frequencies, they will be grouped into the same oscillation cluster. The significance of this concept is that it reveals the dynamic region where energy is mutually coupled between branches and collectively influences the direction of energy propagation.

[0030] Further, after obtaining the oscillation cluster, it is necessary to evaluate the energy interaction relationship between the oscillation cluster and the energy echo core area. The energy echo core area can be regarded as the center of the strongest energy reflection, and the oscillation cluster is often the bridge for energy to propagate from the core to the periphery. Therefore, first, the energy echo core area is taken as the center point or center surface, and its position is marked in the three-dimensional space coordinates. Then, according to the average oscillation direction, energy propagation tendency and coupling strength of each oscillation cluster with the core area, the spatial boundary of the core area is gradually expanded along these directions. This expansion is not a simple geometric expansion, but a directional extension according to the oscillation energy propagation path. For example, if the current phase of a certain oscillation cluster is strongly coupled with the core area, it means that energy has a higher probability of propagation in that direction, so the coverage range of the boundary is expanded in that direction; on the contrary, if the oscillation amplitude is small or the phase is inconsistent with the core area, the expansion is less. After such gradual expansion, an energy influence zone is formed around the core area, which is naturally extended along the oscillation direction. This region is called the first extended zone, which represents the first envelope region of the core area energy extension propagation, like a transition circle of energy wavefront spreading outward. For example, in an energy storage array, if the core area covers the 45-47 units, and the branch 49-50 unit connected to the 47 unit belongs to the same oscillation cluster, then the first extended zone will cover the 49-50 area, forming a belt-shaped structure of energy conduction from the core to the outside.

[0031] Finally, after the first extension zone is determined, the overlap, interference and safety distance of energy propagation in different directions also need to be considered. For this purpose, a parameter called avoidance radius is calculated to determine the minimum distance between different regions during energy propagation to avoid the superposition of oscillations or excessive concentration of energy. Specifically, by analyzing the energy density distribution and oscillation phase difference in each direction within the first extension zone, regions with strong energy interference can be found, and then a reasonable spatial avoidance range can be determined according to the energy fluctuation amplitude. The size of the avoidance radius depends on the strength of energy coupling and the spatial distribution density. For example, in areas with intense energy fluctuations, the avoidance radius is larger; while in directions with faster energy decay, the avoidance radius is smaller. Then, according to this avoidance radius, the original boundary of the energy echo core area is fine-tuned or deformed so that it can cover the main energy propagation direction in space and maintain a moderate separation from high interference areas. The new boundary area formed in this way is called the energy echo area. This energy echo area is equivalent to a dynamic energy interaction spatial range, which not only includes the original energy echo core area, but also includes the transition zone of energy propagation and decay. It reflects the real spatial distribution of energy in the process of reflux, reflection, decay and rebalancing within the energy storage structure. From a practical perspective, for example, the core area is concentrated in units 45-47, and the first extension zone extends to units 48-50. After calculating the avoidance radius, the final energy echo area may cover the range of units 45-50. This ensures complete capture of the energy propagation direction and avoids interference and superposition of different oscillation directions.

[0032] S3, cutting the echo channel corresponding to the energy echo area according to the preset cut-off frequency to obtain a first cut-off area; In this example, the echo channel corresponding to the energy echo area is cut off according to the preset cut-off frequency to obtain a first cut-off area, specifically: Obtain the echo channel in the energy echo area, and classify the channel according to the preset harmonic component according to the current amplitude and frequency characteristics of the echo channel to obtain the classified echo channel; Obtain the harmonic resonance characteristics of the classified echo channel, and preset a plurality of cut-off frequency points according to the harmonic resonance characteristics; For each classified echo channel, the channel is cut off in turn according to the preset cut-off frequency point to obtain a first cut-off area.

[0033] It should be noted that after the identification of the energy echo area is completed, further analysis of the details of the energy flow within the region is required. The energy echo area usually contains multiple electrical energy propagation paths, in which current reflection, wave motion and energy back-and-forth propagation phenomena occur. In order to depict the specific trajectories of these energy flows, echo channels need to be extracted. Echo channel refers to a specific conduction path formed by current reflection and harmonic interference within the echo area, which is manifested as a periodic back-and-forth energy transmission trajectory between conductors, nodes or battery cells. In other words, echo channel is the actual path of energy repeated propagation in the echo area. The process of obtaining these echo channels includes two steps: first, by monitoring the current waveform and phase change of each node in the echo area, identify the current path that has obvious reflection characteristics or periodic fluctuations in time; second, abstract these paths into traceable channel data structures for subsequent analysis. After obtaining the echo channels, they need to be classified according to the current amplitude and frequency characteristics. The specific method is to analyze the current amplitude change and corresponding frequency distribution of each channel within a period of time, extract its main harmonic components, such as fundamental wave, second harmonic, third harmonic, etc. Then according to the pre-set harmonic classification standard, the channels are divided into several groups, such as "low frequency high amplitude", "medium frequency medium amplitude", "high frequency low amplitude" and other categories. For example, in an energy echo area, channels numbered 45-46-48 may exhibit low frequency, larger current fluctuations, while channels 47-49-50 exhibit high frequency, low amplitude oscillation, so they will be divided into different categories. The results of classification provide the basis for subsequent targeted cutting and energy reconstruction.

[0034] Further, when the echo channels are classified, the next step is to analyze the harmonic resonance characteristics of different categories of channels. The so-called harmonic resonance characteristics refer to the phenomenon that energy is amplified, superimposed or delayed dissipation in the channel at a certain frequency. This resonance is usually due to the matching relationship between the length of the wire, inductance and capacitance characteristics. In order to obtain the resonance characteristics, frequency analysis needs to be performed on the current waveform of each classified echo channel to observe the energy response strength and phase stability of each frequency band. If the energy amplitude of a certain frequency band is significantly higher than that of the surrounding frequency band and persists, it can be determined that it is the resonance point of the channel. Through the analysis of multiple channels, the resonance rules of different frequency bands of each type of channel can be summarized. According to these rules, a set of cut-off frequency points will be preset in engineering, which represent the frequency bands that are most likely to cause harmonic accumulation or energy backflow in energy transmission. The principle of selecting these points is: both to cover the high-risk harmonic concentration frequency band and to avoid affecting normal energy flow. For example, in an energy storage architecture, it may be found that the energy transmission is normal in the low frequency band (such as 50Hz), while the energy resonance is obvious in the higher frequency band (such as 250Hz, 400Hz), so 250Hz and 400Hz will be selected as the cut-off frequency points. The frequency points will be used as a control reference for subsequent cut-off operations to orderly interrupt or isolate the energy reflection of the corresponding channel, thereby weakening the energy accumulation effect in the echo area.

[0035] Finally, after obtaining the cut-off frequency points, the different types of echo channels need to be intervened in an orderly manner. Channel cutting is not physically disconnecting the wire, but adjusting the current transmission path or control signal in the channel to inhibit or interrupt energy propagation at a specific frequency. The specific steps are: for each type of echo channel, perform cut-off operation from low frequency to high frequency according to the preset cut-off frequency points. When cutting, use electronic switches, filter modules or control signals to adjust so that the energy flow of the channel is blocked at the corresponding frequency. For example, in the aforementioned example, if a channel has harmonic resonance at 250Hz, trigger the cut-off control at this frequency to prevent energy from being transmitted back along this frequency path; if the same channel also has resonance at 400Hz, perform a second cut-off. After cutting, the area where the energy propagation is limited in the cut-off frequency range but still retains some normal low-frequency energy flow is called the first cut-off area. The first cut-off area can be understood as an area in the energy echo area where the preliminary energy channel suppression has been completed, which is equivalent to a frequency-isolated energy band layer. Here, high-frequency energy reflection is blocked, and low-frequency energy can still be transmitted normally, thereby achieving preliminary weakening of energy oscillation. For example, if the 45-48 path in the energy echo area has a significant decrease in high-frequency harmonic energy after cutting at 250Hz and 400Hz, while the dominant current is still transmitted smoothly, then this area is designated as the first cut-off area.

[0036] S4, extracting a residual current decay characteristic of the first cut-off zone, and performing channel recovery on the first cut-off zone based on the residual current decay characteristic to obtain a second cut-off zone; In this example, a residual current decay characteristic of the first cut-off zone is extracted, and channel recovery is performed on the first cut-off zone based on the residual current decay characteristic to obtain a second cut-off zone, specifically: The residual current waveform and the current response curve of the first cut-off zone are collected, and the residual current waveform is subjected to exponential fitting to obtain a residual decay time; The residual energy dissipation rate is calculated according to the residual decay time, and the recoverable channel is identified based on the dissipation rate distribution diagram; The rising rate of the voltage excitation is adjusted according to the slope change of the current response curve, and the voltage excitation is applied to the recoverable channel to obtain a second cut-off zone.

[0037] It should be noted that although the high-frequency harmonic channel in the first cut-off zone has been isolated, there will still be a part of the current signal that has not completely dissipated, which is referred to as residual current. It usually manifests as a slow decay or slight oscillation of the current waveform within a short period of time, indicating that the local energy has not been completely released. To analyze these residual phenomena, high-precision sampling devices are needed to collect current waveform data of each channel in real time within the first cut-off zone, and to record the current response curve when a small excitation is applied after cutting. The current waveform reflects the decay trend of the instantaneous energy, while the current response curve reflects the reaction speed and stability of the channel to external excitation. Next, the collected residual current waveform is subjected to time series analysis to observe its gradual decay pattern over time. Usually, this decay is not linear, but presents a fast-slow energy release process. In order to quantitatively describe this feature, the decay process is approximated as an exponential decay model through data fitting, thereby extracting the key time parameter representing the decay rate, which is referred to as the residual decay time. The residual decay time refers to the time length from the moment of cutting to the time when the amplitude of the residual current drops to a stable range, which reflects the speed of energy dissipation in the cut-off channel. For example, if the current of channels 45-46 quickly tends to zero within 3 seconds after cutting, it indicates that the residual decay time is short and the energy release is sufficient; if the current of channels 47-8 has not completely stabilized after 15 seconds, it indicates that the residual decay time is long and the energy retention in the channel is serious. This time index is an important basis for subsequent judgment of which channels can be recovered and which channels still need to be isolated.

[0038] Further, when the residual decay time of each channel is determined, the dissipation rate of residual energy can be calculated accordingly. The dissipation rate represents the speed of residual energy being released or converted per unit time, which reflects the energy recovery potential of the channel in the cut-off state. The shorter the residual decay time, the faster the energy is released, and the easier the channel is to recover normal conduction; on the contrary, the longer the decay time, the more energy is retained, and the channel is temporarily not suitable for recovery. To present these features more intuitively, the dissipation rates of all channels are mapped to the spatial coordinates to draw a dissipation rate distribution map. This map shows the distribution of energy decay speed of channels at different positions. For example, if there is a high dissipation rate area between battery units 45-47 and a low speed area between 48-50 in the map, it means that the former has sufficient energy release, and the latter still has local accumulation. Then, through clustering or hierarchical analysis of the distribution map, a batch of channels with sufficient energy dissipation, stable oscillation and low recovery risk can be identified, which are called recoverable channels. In simple terms, recoverable channels are those that have safely released energy after being cut off and have the ability to re-establish electrical energy conduction. For example, among the 50 channels, if 20 of them decay smoothly within a short time and have no abnormal fluctuations in the response curve, these 20 channels can be marked as recoverable channels to provide target objects for subsequent recovery operations.

[0039] Finally, after identifying the recoverable channels, the next step is to gradually restore their energy flow state. During the recovery process, the application method of voltage excitation needs to be controlled to avoid sudden changes causing new energy shocks. The specific method is as follows: first, analyze the slope of the current response curve of each recoverable channel. The slope represents the response sensitivity of the channel to voltage excitation. The larger the slope, the faster the channel responds during energy transmission and the higher the risk; a smaller slope represents a more stable channel. According to these slope characteristics, adjust the rising rate of the voltage excitation, i.e. for sensitive channels, use a slow-rising voltage excitation to gradually restore conduction; for channels with a slow response, the voltage rising speed can be appropriately accelerated. Through this differentiated control method, new echoes or overshoots of energy during the recovery phase can be effectively prevented. After completing the voltage excitation of the recoverable channels, these channels gradually transition from a completely cut-off state to a controlled conduction state, and energy re-forms a stable flow path between them. At this time, the first cut-off area evolves into a new working area after the channel recovery operation, which is called the second cut-off area. The main difference between the second cut-off area and the first cut-off area is as follows: The first cut-off zone belongs to the energy isolation stage, and its main goal is to block high-frequency harmonics and energy backflow, and temporarily retain energy for dissipation. The second cut-off zone belongs to the energy recovery stage, and its main goal is to re-establish a balanced energy path through precise voltage excitation, so that local energy participates in overall energy scheduling again. For example, assuming that channels 45-47 are determined to be recoverable after dissipation analysis in the first cut-off zone, and after slow voltage excitation is applied to them, the current gradually recovers to a stable waveform, and the energy flow in this region is re-established, then the 45-47 region becomes part of the second cut-off zone.

[0040] S5, dynamically adjusting the timing path of the internal electric energy flow of the battery unit based on the second cut-off zone.

[0041] In this example, the timing path of the internal electric energy flow of the battery unit is dynamically adjusted based on the second cut-off zone, specifically: Obtain the three-dimensional coordinates of the second cut-off zone to obtain the spatial distribution matrix of the second cut-off zone; Reconstruct the preset energy distribution matrix of the matrix battery unit based on the spatial distribution of the second cut-off zone to obtain the reconstructed energy distribution matrix; Based on the reconstructed energy distribution matrix, allocate a variable length time window to each battery unit and collect voltage ripple and current total harmonic distortion data in real time; According to the voltage ripple and harmonic data, dynamically calibrate the start delay and duration of the timing path to obtain an optimized set of electric energy flow timing paths.

[0042] It should be noted that after completing the voltage excitation of the recoverable channels, a local energy flow path has been re-established in the second cut-off zone. At this time, in order to further optimize the overall energy scheduling, it is necessary to determine the distribution position and mutual relationship of these channels in space. Obtaining the three-dimensional coordinates of the second cut-off zone is to collect and map the position data (such as X, Y, Z coordinates) of each recovered channel, node, and battery unit in the actual physical layout. The specific method is to identify which channels belong to the second cut-off zone through the installation coordinate points of each battery unit and wire in the energy storage architecture, and combine their coordinates to form a three-dimensional space grid. This process is equivalent to converting the channels originally defined from an electrical perspective into a geometric structure that can be expressed in space.

[0043] On this basis, a spatial distribution matrix can be constructed, which is a data structure used to describe the relative distribution and connection relationship of each energy path in the second cutoff zone in three-dimensional space. Each element of it represents the position of a specific channel or node in space, energy flow direction and adjacency relationship. For example, in a 10x10x3 energy storage layout, if the second cutoff zone is mainly concentrated in the 4x4 area of the bottom layer, the spatial distribution matrix can clearly show this local energy flow concentration area and present the three-dimensional form of the energy recovery hot spot. In colloquial terms, the spatial distribution matrix is a kind of three-dimensional mapping table, which is used to reflect which positions of the channel have been recovered, how they are connected, and how the energy flow extends in space.

[0044] Further, in energy scheduling, each battery unit originally has a preset energy allocation matrix, which defines the proportion, priority order and time allocation rules of energy inflow and outflow between different units. However, when the second cutoff zone is formed, the energy flow structure has changed locally - some channels are recovered and some are still in isolation. In order to adapt to this new change, the original energy allocation matrix needs to be reconstructed. The specific process is: first, according to the spatial distribution matrix obtained in the previous step, identify the energy flow concentration area where the second cutoff zone is located, and extract the spatial adjacency relationship and energy channel weight of the battery units in this area. Then, compare these new connection relationships with the original energy allocation matrix, and update those units whose channel state has changed. For example, assume that in the original allocation matrix, A unit only has energy conduction relationship with B and C, but after the formation of the second cutoff zone, A and D units recover energy flow, then the new matrix needs to increase the allocation proportion of A-D, and rebalance the weight of A, B and C. The matrix obtained after this update is the reconstructed energy allocation matrix. It can better reflect the actual state of energy flow than the original matrix, and is the basis for real-time scheduling and time sequence control.

[0045] Secondly, after the energy flow is redistributed, the load state and energy exchange rate of different battery units will be different. In order to ensure smooth overall operation, a variable length time window needs to be set for each unit to control the time range of its electric energy input or output. The length of the time window is dynamically adjusted according to the energy distribution matrix after reconstruction: units with frequent energy flow and large load changes are allocated shorter time windows to calibrate more frequently; units with stable energy flow and lighter load are allocated longer windows to reduce the adjustment frequency. For example, in a 100-unit energy storage architecture, units 1-20 may have a time window of only 2 seconds if they are in an energy interaction intensive area, while units 80-100 in the low interaction area may have a time window of up to 10 seconds. Within each time window, two key indicators are collected in real time: voltage ripple and total harmonic distortion (THD). Voltage ripple reflects the degree of fluctuation of internal energy balance of the unit, and THD reflects the stability of energy flow and whether there is harmonic interference. By continuously collecting these data, it can be monitored in real time whether the energy conduction process is smooth, whether there is oscillation or backflow phenomenon.

[0046] Finally, after collecting real-time data, the energy flow timing of each battery unit needs to be dynamically calibrated. The timing path refers to the start order and duration arrangement of energy conduction between different units. The start delay represents the time a unit needs to wait at the beginning of energy flow, and the duration represents the length of time the unit remains in energy conduction state. The calibration process is dynamically completed according to the change trend of voltage ripple and harmonic distortion data: if the voltage ripple of a unit is too large, it means that the energy input is changing too fast, so the start delay should be extended to make it participate in energy conduction later; if the harmonic distortion is high, it means that there is interference or resonance in energy flow, so the duration should be shortened to avoid accumulation. For example, in an energy conduction path A-B-C, if it is detected that the ripple at node B increases significantly, the start delay of A-B may be extended by 0.5 seconds, while the duration of B-C is shortened to balance the flow rhythm. After several rounds of real-time calibration, the timing paths of all units will tend to be coordinated, and the energy flow will gradually transition from disordered or fluctuating state to smooth and orderly rhythm, eventually forming an optimized set of energy flow timing paths.

[0047] Embodiment 2, Figure 2 The present application provides an energy scheduling method and system for a data center, which comprises a data acquisition module, a loop identification module, a channel cutting module, a channel recovery module and a battery adjustment module: The data acquisition module is used to acquire battery unit operation data in the solar power plant energy storage system, and to extract energy accumulation nodes according to the battery unit operation data; A loop identification module is configured to identify an extension direction of a current loop of the energy accumulation node based on a preset wire connection relationship, and obtain an energy echo area; A channel cutting module is configured to cut an echo channel corresponding to the energy echo area according to a preset cutting frequency, and obtain a first cutting area; A channel recovery module is configured to extract a residual current decay feature of the first cutting area, and perform channel recovery on the first cutting area based on the residual current decay feature, and obtain a second cutting area; A battery adjustment module is configured to dynamically adjust a timing path of internal electric energy flow of the battery unit based on the second cutting area.

[0048] As Figure 3 shown in the figure is an electronic device structure schematic diagram of the energy scheduling method of the data center.

[0049] The electronic device can include a processor, a memory, a communication bus, and a communication interface, and can further include a computer program stored in the memory and executable on the processor, such as an energy scheduling program of a data center.

[0050] The processor is the control core (Control Unit) of the electronic device, which connects all components of the electronic device through various interfaces and lines, and executes or runs the programs or modules stored in the memory, and calls the data stored in the memory, to perform various functions of the electronic device and process data.

[0051] The memory includes at least one type of readable storage medium, and in some embodiments, the memory can be an internal storage unit of the electronic device, such as a mobile hard disk of the electronic device. The memory can be used to store application software and various data installed in the electronic device.

[0052] The communication bus is configured to realize the connection and communication between the memory and at least one processor.

[0053] The communication interface is used for communication between the above-mentioned electronic device and other devices, including a network interface and a user interface.

[0054] Only the electronic device with components is shown in the figure, and those skilled in the art can understand that the structure shown in the figure does not constitute a limitation on the electronic device, and can include fewer or more components than shown in the figure, or combine certain components, or different component arrangements.

[0055] It should be understood that the embodiments are only for illustration, and the scope of the patent application is not limited by the structure.

[0056] The memory in the electronic device stores an energy scheduling program of a data center, which is a combination of a plurality of instructions and can implement the steps in the energy scheduling method of the data center when executed in the processor.

[0057] Specifically, the processor can refer to the description of the relevant steps in the corresponding embodiments of the above-mentioned specific implementation system of instructions, and will not be described here.

[0058] The above embodiments can be realized in whole or in part by software, hardware, firmware or any combination thereof. When realized by software, the above embodiments can be realized in whole or in part in the form of a computer program product.

[0059] Those skilled in the art can realize that the modules and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized 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 the present application.

[0060] In addition, the functional modules in each embodiment of the present application can be integrated in one processing module, or each module can exist physically alone, or two or more modules can be integrated in one module.

[0061] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any skilled person in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be included in 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.

[0062] Finally: the above is only the preferred embodiment of the present application, and is not used to limit the present application, any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application, should be included in the protection scope of the present application.

Claims

1. A method for energy scheduling of a data center, the method comprising: The method comprises the following steps: acquiring battery unit operation data in a solar power station energy storage system, and extracting an energy accumulation node according to the battery unit operation data; identifying a current loop extension direction of the energy accumulation node based on a preset wire connection relationship, and obtaining an energy echo area; cutting an echo channel corresponding to the energy echo area according to a preset cutting frequency, and obtaining a first cutting area; extracting a residual current decay characteristic of the first cutting area, and performing channel recovery on the first cutting area based on the residual current decay characteristic, and obtaining a second cutting area; dynamically adjusting a time sequence path of internal electric energy flow of the battery unit based on the second cutting area.

2. The method of Claim 1, wherein, The energy accumulation node is extracted according to the battery unit operation data, specifically: extracting a multi-channel voltage sequence and a current sequence of the battery unit based on the battery unit operation data, and obtaining a battery operation state matrix; extracting a charge accumulation rate value and a discharge fluctuation amplitude of each battery unit based on the battery operation state matrix; calculating an energy difference value within a unit time based on the charge accumulation rate value and the discharge fluctuation amplitude, and identifying an initial energy accumulation point based on the energy difference value; obtaining a physical connection topology between the battery units, and performing cluster analysis on the initial energy accumulation point based on the physical connection topology, and obtaining an energy accumulation area; performing energy density evaluation on the energy accumulation area, and screening the initial energy accumulation point according to the evaluation result, and obtaining the energy accumulation node.

3. The method of Claim 2, wherein, The current loop extension direction of the energy accumulation node is identified based on the preset wire connection relationship, and the energy echo area is obtained, specifically: obtaining a wire connection topology diagram of the solar power station energy storage system, and starting from the energy accumulation node, traversing the wire connection topology diagram in a forward direction and a reverse direction respectively, and obtaining a plurality of current flow paths; calculating an impedance value and a current phase delay value of each current flow path, and identifying a closed-loop current loop based on the impedance value and the current phase delay value; calculating an active power density vector between nodes of the closed-loop current loop, and obtaining a main loop and a branch loop of the closed-loop current loop based on the active power density vector; superimposing the energy accumulation nodes of the main loop to obtain an energy echo core area; extending the boundary of the energy echo core area based on the branch loop to obtain the energy echo area.

4. The method of Claim 3, wherein, The boundary of the energy echo core area is extended based on the branch loop to obtain the energy echo area, specifically: extracting a current oscillation characteristic of the branch loop, and performing cluster analysis on the branch loop according to the current oscillation characteristic to obtain an oscillation cluster; extending the boundary along the oscillation cluster direction with the energy echo core area as the center to obtain a first extension zone; calculating an avoidance radius based on the first extension zone, and adjusting the boundary of the energy echo core area according to the avoidance radius to obtain the energy echo area.

5. The method of Claim 1, wherein, The echo channel corresponding to the energy echo area is cut according to a preset cutting frequency to obtain a first cutting area, specifically: obtaining an echo channel in the energy echo area, and classifying the channel according to a preset harmonic component based on a current amplitude and a frequency characteristic of the echo channel to obtain a classified echo channel; obtaining a harmonic resonance characteristic of the classified echo channel, and presetting a plurality of cutting frequency points according to the harmonic resonance characteristic; For each classified echo channel, the channel is sequentially cut off according to a preset cut-off frequency point to obtain a first cut-off area.

6. The method of Claim 1, wherein, The residual current decay characteristics of the first cut-off area are extracted, and the first cut-off area is recovered based on the residual current decay characteristics to obtain a second cut-off area, specifically: The residual current waveform and the current response curve of the first cut-off area are collected, and the residual current waveform is exponentially fitted to obtain a residual decay time; The residual energy dissipation rate is calculated according to the residual decay time, and the recoverable channel is identified based on the dissipation rate distribution diagram; The rising rate of the voltage excitation is adjusted according to the slope change of the current response curve, and the voltage excitation is applied to the recoverable channel to obtain the second cut-off area.

7. The method of Claim 1, wherein, The timing path of the internal electric energy flow of the battery unit is dynamically adjusted based on the second cut-off area, specifically: The three-dimensional coordinates of the second cut-off area are obtained to obtain a spatial distribution matrix of the second cut-off area; The preset energy distribution matrix of the matrix battery unit is reconstructed based on the spatial distribution matrix of the second cut-off area to obtain a reconstructed energy distribution matrix; Based on the reconstructed energy distribution matrix, a variable length time window is allocated to each battery unit, and voltage ripple and current total harmonic distortion data are collected in real time; The start delay and duration of the timing path are dynamically calibrated according to the voltage ripple and harmonic data to obtain an optimized set of electric energy flow timing paths.

8. A system for energy scheduling using the data center of any of claims 1-7, wherein, It includes: A data acquisition module is configured to acquire battery unit operation data in a solar power station energy storage system and extract energy accumulation nodes based on the battery unit operation data; A loop identification module is configured to identify the current loop extension direction of the energy accumulation nodes based on a preset wire connection relationship to obtain an energy echo area; A channel cutting module is configured to cut the echo channel corresponding to the energy echo area according to a preset cut-off frequency to obtain a first cut-off area; A channel recovery module is configured to extract residual current decay characteristics of the first cut-off area and recover the first cut-off area based on the residual current decay characteristics to obtain a second cut-off area; A battery adjustment module is configured to dynamically adjust the timing path of the internal electric energy flow of the battery unit based on the second cut-off area.

9. An electronic device, comprising: The electronic device includes: At least one processor; and a memory connected in communication with the at least one processor; The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the energy scheduling method of the data center as claimed in any one of claims 1 to 7.

Images