A flywheel energy storage system control system for data center computing power load energy recovery

Abstract

The present application relates to the technical field of data center energy management, and discloses a flywheel energy storage system control system for computing power load energy recovery in a data center, which comprises an energy storage unit, a load monitoring unit, a flywheel control unit and an energy distribution unit. The energy storage unit is provided with multiple sets of flywheel energy storage devices that can be charged / discharged. The load monitoring unit monitors load data through power sensors covering key nodes of computing power equipment. The flywheel control unit collects flywheel rotating speed and power grid frequency fluctuation data. The energy distribution unit decides charging and discharging based on load data, establishes an alternating control strategy for load and frequency coordination, and dynamically adjusts and predicts the flywheel operating state. The system adopts time sequence synchronous control, the energy distribution unit contains multiple modules, the alternating controller supports dual-mode switching and parameter optimization, and the flywheel unit adopts a ring redundancy or star centralized structure. The system realizes efficient recovery of computing power load energy and optimization of flywheel energy storage scheduling, and improves the energy utilization rate of the data center and the stability of the power grid.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data center energy management, in particular to a flywheel energy storage system control system for data center computing power load energy recovery. BACKGROUND

[0002] With the rapid development of information technology, data centers as the infrastructure of digital economy, their scale and computing power demand are showing explosive growth. During operation, the load of computing power equipment in data centers has significant volatility and intermittency. This unstable power demand not only increases the pressure on power supply, but also causes a lot of energy waste. Traditional data center energy management systems are difficult to efficiently respond to the dynamic changes of computing power load, and cannot realize real-time recovery and reasonable allocation of energy, resulting in low energy utilization and increasing the operating cost and environmental burden of data centers.

[0003] In existing energy storage technologies, battery energy storage systems have problems such as short service life, high maintenance cost, poor environmental adaptability, and are prone to performance degradation and safety hazards during frequent charging and discharging. Flywheel energy storage, as a mechanical energy storage method, has the advantages of high power density, fast charging and discharging speed, long cycle life, and environmental friendliness, and has unique advantages in dealing with short-term power fluctuations and energy recovery. However, the flywheel energy storage system control system for data center computing power load characteristics is not mature yet, and lacks effective load monitoring, flywheel control and energy distribution coordination mechanisms, which cannot realize efficient recovery of computing power load energy and optimal scheduling of flywheel energy storage devices.

[0004] Specifically, the existing technology has the following deficiencies: First, in terms of load monitoring, traditional monitoring methods cannot comprehensively and real-time obtain load data of key nodes of computing power equipment, making it difficult to accurately reflect the real state of computing power load, resulting in a lack of reliable basis for charging and discharging decisions of energy storage systems. Second, in terms of flywheel control, there is a lack of coordinated monitoring and control of flywheel speed and grid frequency, which cannot adjust the flywheel operating state in time according to grid frequency fluctuations, affecting the stability of the system and power quality. Third, in terms of energy distribution, the existing system fails to establish a load and frequency coordinated control strategy, which cannot realize dynamic optimal scheduling of flywheel energy storage devices, resulting in low energy recovery efficiency and failing to meet the requirements of data centers for high reliability and high energy utilization.

[0005] The existing flywheel energy storage system lacks effective timing synchronization control mechanism when multiple flywheels are running cooperatively, and the operating states of each flywheel device are difficult to synchronize, resulting in a decline in overall system performance. At the same time, in the topology design of flywheel units, there is a lack of optimized configuration for the characteristics of data center power networks, which cannot fully utilize the efficiency of flywheel energy storage devices.

[0006] Therefore, there is an urgent need for a flywheel energy storage system control system that can adapt to the data center computing power load characteristics. Through efficient load monitoring, precise flywheel control and intelligent energy distribution, real-time recovery of computing power load energy and optimization of flywheel energy storage devices can be achieved, improving energy utilization in data centers, reducing operating costs and environmental impact. SUMMARY

[0007] The present application aims to provide a flywheel energy storage system control system for data center computing power load energy recovery to solve the problems raised in the background art.

[0008] To achieve the above-mentioned purpose, the present application provides the following technical solution: a flywheel energy storage system control system for data center computing power load energy recovery, the system comprising:

[0009] an energy storage unit, a load monitoring unit, a flywheel control unit and an energy distribution unit;

[0010] The energy storage unit comprises a plurality of flywheel energy storage devices arranged in the data center power network; each flywheel energy storage device is in a charging or discharging state;

[0011] The load monitoring unit comprises a plurality of power sensors deployed on the data center power supply line; each power sensor monitors the power input node in the direction of the computing power equipment, so that at least one power sensor captures real-time load data of a group of computing power equipment;

[0012] The flywheel control unit comprises a flywheel speed sensor and a power grid frequency detector, which respectively measure the speed data of the flywheel energy storage device and the power grid frequency fluctuation data;

[0013] The energy distribution unit uses the computing power equipment load data collected by the load monitoring unit to make flywheel charging and discharging decisions; and establishes an alternating control strategy coordinated with the load and frequency, uses the power grid frequency fluctuation data to dynamically adjust the charging and discharging decisions; according to the adjustment result, the running state of the flywheel energy storage device in the next period is predicted, which is used for fast response of power dispatching in the next cycle.

[0014] Preferably, the energy storage unit, load monitoring unit and flywheel control unit are controlled synchronously in time sequence; specifically including:

[0015] Each flywheel energy storage device in the energy storage unit is numbered, and one of them is used as a master flywheel to generate a synchronization signal Fly SYNC; the remaining flywheels switch modes synchronously after receiving the synchronization signal Fly SYNC, and send the running state data with flywheel numbers to the energy distribution unit;

[0016] The multiple groups of power sensors of the load monitoring unit are numbered; the synchronization signal Fly SYNC adjusts the sampling frequency of each group of power sensors, and sends the adjusted sensor number information to the energy distribution unit;

[0017] The synchronization signal Fly SYNC is transmitted to the flywheel control unit at the same time, to control the flywheel speed sensor and the grid frequency detector to synchronously collect data.

[0018] Preferably, the energy distribution unit comprises a load prediction module, a speed regulation module, an alternating controller and a state estimation module; wherein,

[0019] The load prediction module is used for trend analysis on historical load data of the computing power equipment, to obtain load change characteristics in a future period;

[0020] The speed regulation module is used for difference calculation on the measurement data of the flywheel speed sensor and the grid frequency detector, to determine a matching degree parameter of the flywheel speed and the grid frequency;

[0021] The alternating controller is used for constructing a state space model of alternating control based on the load change characteristics and the matching degree parameter, and performing parameter correction on the controller using real-time load data as a constraint condition; and taking the corrected flywheel charge-discharge rate as an output;

[0022] The state estimation module is used for dynamic integration on the corrected charge-discharge rate, in combination with the real-time matching degree parameter received by the speed regulation module, to predict a running mode threshold value of the flywheel energy storage device in a next period.

[0023] Preferably, the alternating controller adopts a dual-mode switching strategy, in which,

[0024] When the grid frequency fluctuation exceeds a set threshold value, a flywheel speed compensation amount is calculated based on the matching degree parameter, and a charge-discharge rate correction instruction is generated;

[0025] When the load change rate exceeds a set threshold value, a flywheel charge-discharge priority sequence is generated based on the load prediction result, and a cooperative scheduling mechanism of the energy storage device is triggered.

[0026] Preferably, in the dual-mode switching process, a dynamic weight distribution algorithm is adopted for control parameter optimization.

[0027] Preferably, the alternating controller uses the load variation feature and the matching degree parameter to construct a state vector composed of the flywheel speed compensation amount under the grid frequency reference, the charge-discharge priority coefficient, and the remaining capacity proportion of the energy storage device, and establishes a state transition equation in combination with the load prediction error correction factor, the frequency fluctuation compensation factor, and the mode switching delay parameter; uses real-time load data as a boundary condition to modify the parameters of the controller; and uses the modified flywheel charge-discharge rate as the output.

[0028] Preferably,

[0029] In the control process, the rapid response process under the condition that the flywheel operation mode or the load monitoring object is switched includes:

[0030] 1) According to the synchronization signal Fly SYNC, the alternating controller output charge-discharge rate data of the flywheel or the sensor at the switching time is obtained;

[0031] 2) According to the charge-discharge rate data and the energy capacity parameters of each flywheel energy storage device, the available energy reserve amount of each flywheel group is calculated;

[0032] 3) Based on the real-time data of the load monitoring unit, the matching degree between the energy reserve amount and the load demand is calculated to determine the dispatchable energy threshold of each flywheel energy storage device;

[0033] 4) The actual operating state parameters of each flywheel energy storage device at the switching time are obtained, and the deviation amount between the actual operating state parameters of each flywheel energy storage device and the dispatchable energy threshold of each flywheel energy storage device is calculated; when the deviation amount of a flywheel is lower than a set threshold, the flywheel is included in the current dispatch queue and the charge-discharge rate adjustment instruction is executed.

[0034] Preferably, each flywheel energy storage device includes a plurality of independent flywheel groups; and the plurality of independent flywheels of each flywheel group are connected to the power network in a set topology.

[0035] Preferably, the connection mode of the independent flywheels in each flywheel group is a ring redundancy structure or a star centralized structure; wherein,

[0036] In the ring redundancy structure, three flywheels are connected in series to form a closed loop, and the center node is connected to the power distribution bus;

[0037] In the star centralized structure, four flywheels are connected in parallel to the same power node, and the center node is connected to the output end of the load monitoring unit.

[0038] Preferably, the arrangement position of the power sensor covers the AC input terminal, the DC power distribution cabinet, and the uninterruptible power supply output port of the computing power equipment.

[0039] Compared with the prior art, the present application has the following advantages:

[0040] In terms of load monitoring and data acquisition, the load monitoring unit comprehensively covers the key nodes of the AC input terminals, DC distribution cabinets, and uninterruptible power supply output ports of computing power equipment through multiple power sensors deployed on the data center power supply lines. It can capture real-time load data of each group of computing power equipment in real time and accurately. This comprehensive monitoring coverage ensures the integrity and accuracy of the load data, providing a reliable basis for subsequent charging and discharging decisions, enabling the system to respond to dynamic changes in computing power load in a timely manner.

[0041] The design of the energy storage unit is highly innovative, with multiple flywheel energy storage devices arranged in the data center power network. Each flywheel energy storage device includes multiple independent flywheel sets, connected to the power network in a ring redundancy structure or a star centralized structure. In the ring redundancy structure, three flywheels are connected in series to form a closed loop, with the center node connected to the power distribution bus. This structure improves the reliability and redundancy of the system, so that even if a flywheel fails, other flywheels can still maintain normal operation of the system. In the star centralized structure, four flywheels are connected in parallel to the same power node, with the center node connected to the load monitoring unit output for centralized monitoring and control, improving the system's ability to work together. The coordinated operation of multiple flywheels and the flexible topology design significantly enhance the system's energy storage capacity and regulation ability, better adapting to the energy recovery needs of different scales of computing power load in data centers.

[0042] The flywheel control unit measures the flywheel speed data and grid frequency fluctuation data in real time through the flywheel speed sensor and grid frequency detector, providing key operating state parameters for the system. Based on these parameters, the system can achieve coordinated control of flywheel speed and grid frequency. When the grid frequency fluctuation exceeds the set threshold, the system calculates the flywheel speed compensation amount and generates a charging and discharging rate correction instruction in a timely manner, ensuring the stability of the grid frequency and improving the power quality and stability of the system.

[0043] The energy distribution unit is the core control module of the system, and its innovation point is remarkable. By establishing an alternating control strategy coordinated with load and frequency, and combining the trend analysis of historical load data of the computing power equipment by the load prediction module, the load variation characteristics of the future period can be accurately obtained, and the charging and discharging plan can be made in advance. The speed regulation module calculates the difference between the measured data of the flywheel speed and the grid frequency to determine the matching degree parameter, which provides an important adjustment basis for the alternating controller. The alternating controller adopts a dual-mode switching strategy. When the load change rate exceeds the set threshold, the flywheel charging and discharging priority sequence is generated based on the load prediction result, triggering the cooperative scheduling mechanism of the energy storage device, realizing the dynamic optimization scheduling of the flywheel energy storage device, and improving the energy recovery efficiency and the response speed of the system. At the same time, the alternating controller uses the load variation characteristics and the matching degree parameter to construct the state vector, and establishes the state transition equation combined with various correction factors, and uses real-time load data as boundary conditions for parameter correction, so that the charging and discharging decision is more accurate and scientific, further improving the control performance and energy utilization rate of the system.

[0044] The time sequence synchronization control mechanism is another highlight of the present application. The time sequence synchronization control is adopted for the energy storage unit, the load monitoring unit and the flywheel control unit. The main flywheel generates a synchronization signal Fly SYNC, realizing the synchronous switching of the running state of each flywheel energy storage device, the synchronous adjustment of the sampling frequency of the power sensor and the synchronous acquisition of the flywheel control unit data. This synchronous control ensures the consistency of data interaction and action execution between the units of the system, improves the cooperative working efficiency and overall performance of the system, and reduces the control delay and error caused by asynchronous operation.

[0045] In terms of the rapid response capability of the system, when the flywheel operating mode or the load monitoring object is switched, the charging and discharging rate data at the switching time is obtained through the synchronization signal, and the available energy reserve, the dispatchable energy threshold and the deviation are quickly calculated combined with the energy capacity parameter of the flywheel and the real-time load data, so as to timely determine the dispatchable flywheel and execute the charging and discharging rate adjustment instruction. This rapid response mechanism ensures the stability and reliability of the system in the dynamic change process, can effectively cope with the sudden change of the data center computing power load, and further improves the practicality and adaptability of the system. BRIEF DESCRIPTION OF DRAWINGS

[0046] Fig. 1 The working principle diagram of the flywheel energy storage system control system for data center computing power load energy recovery described in the present application;

[0047] Fig. 2 The working principle diagram of the energy distribution unit;

[0048] Fig. 3 The working principle diagram of the alternating controller dual-mode switching strategy. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0050] Please refer to Figs. 1-3 The present application relates to a flywheel energy storage system control system for data center computing power load energy recovery, which comprises an energy storage unit, a load monitoring unit, a flywheel control unit and an energy distribution unit. The specific implementation steps are as follows:

[0051] The energy storage unit comprises a plurality of flywheel energy storage devices arranged in the data center power network, and each flywheel energy storage device can be in a charging or discharging state. The load monitoring unit comprises a plurality of power sensors deployed on the power supply line of the data center, and the monitoring direction of each power sensor covers the power input node of the computing power equipment, so that at least one power sensor can capture real-time load data of a group of computing power equipment. The flywheel control unit comprises a flywheel speed sensor and a power grid frequency detector, which are respectively used to measure the speed data of the flywheel energy storage device and the power grid frequency fluctuation data. The energy distribution unit uses the computing power equipment load data collected by the load monitoring unit to make flywheel charging and discharging decisions, establishes an alternating control strategy coordinated with the load and the frequency, dynamically adjusts the charging and discharging decisions using the power grid frequency fluctuation data, and predicts the operating state of the flywheel energy storage device in the next period according to the adjustment result, for rapid response of power dispatching in the next cycle.

[0052] The technical solutions of the present application will be further described in detail below in combination with specific embodiments. Embodiments

[0053] In this embodiment, time sequence synchronization control is adopted for the energy storage unit, the load monitoring unit and the flywheel control unit, so as to ensure the coordination and consistency of the entire system in the process of data acquisition, state switching and instruction execution. The specific implementation manner is as follows:

[0054] For time sequence synchronization control of the energy storage unit, the energy storage unit comprises a plurality of flywheel energy storage devices arranged in the data center power network, each device is composed of a plurality of independent flywheel machines, and each flywheel energy storage device needs to be switched between charging and discharging states. In order to realize synchronization control, each flywheel energy storage device needs to be uniquely numbered, and the numbering rule can adopt a combination of "group number + single machine serial number", such as "Group1-01" and "Group2-02", to ensure that each flywheel has a unique identifier in the system.

[0055] On a numbered basis, one of the flywheels is selected as the master flywheel. The selection of the master flywheel can be based on preset rules, such as preferentially selecting the flywheel with the highest remaining capacity in the initial state of the charging state, or determined by a self-election mechanism at system initialization. The core function of the master flywheel is to generate a synchronization signal Fly SYNC, which contains key information such as a timestamp, instruction type (such as mode switching instruction, data acquisition instruction, etc.), and synchronization period. The transmission medium of the synchronization signal can use optical fiber or high-speed Ethernet to ensure low delay and high reliability of signal transmission.

[0056] The remaining flywheels are slave flywheels, which, after receiving the synchronization signal Fly SYNC, need to perform the following operations: first, parse the instruction type and timestamp in the signal to confirm whether it is a valid synchronization instruction; if it is a valid instruction, then switch the operating mode according to the instruction type, for example, from the charging mode to the discharging mode, or vice versa; after switching, the slave flywheel needs to pack the operating state data (such as current speed, remaining capacity, charging and discharging rate, etc.) with its own number into a data frame and send it to the energy distribution unit through the designated communication link. The format of the data frame needs to follow a unified communication protocol, such as including frame header, flywheel number, state parameter, and check code fields, to ensure that the energy distribution unit can accurately parse the data.

[0057] Timing synchronization control for the load monitoring unit. The load monitoring unit is deployed with multiple power sensors, which are distributed at key nodes of the data center power supply line, such as the AC input terminals of computing devices, DC power distribution cabinets, and uninterruptible power supply output ports. To achieve synchronized sampling of the sensors, multiple groups of power sensors need to be numbered, which can correspond to the area or device group they monitor, for example, "Sensor-A1" corresponds to the sensor monitoring the first group of computing devices in the A area.

[0058] The synchronization signal Fly SYNC adjusts the sampling frequency of each group of power sensors. Specifically, the synchronization signal generated by the master flywheel contains sampling frequency control parameters, such as setting the sampling period to 100ms or 50ms. After receiving the synchronization signal, each power sensor adjusts its sampling period according to the sampling frequency parameter in it and synchronously collects data at the start of each sampling period. For example, if the synchronization signal sets the sampling period to 100ms, all sensors collect voltage, current, power, and other parameters of the monitored nodes at the start of each 100ms (such as 0ms, 100ms, 200ms, etc.).

[0059] After sampling is completed, the sensor needs to bind the collected data with its own number information to form a data packet containing the sensor number, sampling time, and monitoring data, and send it to the energy distribution unit. To avoid data transmission conflicts, communication mechanisms such as time division multiplexing (TDM) or carrier sense multiple access (CSMA) can be used to ensure that the data of each sensor can be transmitted in order. In addition, the synchronization signal can also be used to calibrate the clock of the sensor, by periodically sending time synchronization instructions to ensure that the internal clocks of all sensors are consistent, avoiding inconsistencies in the timestamps of the sampled data due to clock drift.

[0060] The timing synchronization control of the flywheel control unit. The flywheel control unit includes a flywheel speed sensor and a power grid frequency detector, which are used to measure the speed data of the flywheel energy storage device and the frequency fluctuation data of the power grid, respectively. The synchronization signal FlySYNC needs to be transmitted to the flywheel control unit at the same time to control the synchronous collection of data by the two types of sensors.

[0061] For the flywheel speed sensor, its installation position needs to be close to the rotating shaft of the flywheel, and a non-contact measurement method (such as an optical encoder or a magneto-electric sensor) is used to obtain the real-time speed of the flywheel. After receiving the synchronization signal, the speed sensor triggers sampling at a specified time (such as the rising edge of the synchronization signal), records the current flywheel speed value, and sends the data combined with the flywheel number, sampling time, and other information to the energy distribution unit.

[0062] The power grid frequency detector is used to monitor the frequency fluctuation of the data center power network, and its sampling point is usually set at the entrance of the power network or at the key bus node. When the synchronization signal is transmitted to the frequency detector, the detector samples the power grid frequency at the synchronization time to obtain the current frequency value and the frequency change rate (df / dt), and packages these data into a frequency monitoring data packet and sends it to the energy distribution unit.

[0063] To ensure the timing synchronization of the three types of units (energy storage unit, load monitoring unit, and flywheel control unit), the synchronization signal Fly SYNC needs to have a high-precision time reference. The master flywheel can be equipped with a high-precision clock module (such as a GPS synchronized clock or an atomic clock) to generate a timestamp with nanosecond-level precision. The transmission delay of the synchronization signal needs to be controlled within the microsecond level, which can be calibrated through hardware synchronization mechanisms (such as a dedicated synchronization bus) or software compensation algorithms (such as the NTP time synchronization protocol).

[0064] During the system initialization phase, the configuration of the synchronization parameters needs to be performed, including the setting of the synchronization period, the designation of the master flywheel, and the mapping table of the number of each unit device. The synchronization period can be flexibly adjusted according to the response requirements of the system, for example, in periods of large load fluctuations, the synchronization period can be shortened to improve the real-time performance of the system, and in periods of stable load, the synchronization period can be extended to reduce communication overhead.

[0065] When the master flywheel fails during system operation, a master-slave switching mechanism is required. For example, a new master flywheel is automatically elected from the slave flywheels. Before taking over the task of generating synchronization signals, the new master flywheel needs to synchronize the clock with other slave flywheels to ensure the continuity and accuracy of the synchronization signals during the switching process.

[0066] At the data transmission level, the energy distribution unit needs to set up a dedicated synchronization data receiving interface that can handle multiple sources of synchronization data (such as flywheel state data, sensor monitoring data, and frequency detection data) simultaneously. The interface needs to have data buffering and time alignment functions, such as aligning data from different sources to the same time point based on the timestamps in the data, in order to facilitate subsequent collaborative analysis and control decisions.

[0067] In addition, the timing synchronization control also needs to consider electromagnetic compatibility (EMC) factors to avoid the synchronization signals being interfered by high-frequency devices (such as servers and switches) in the data center. The transmission cable of the synchronization signals needs to use shielded twisted pair or optical fiber, and the wiring path needs to be away from strong electromagnetic sources to ensure the integrity of the signals. Embodiment

[0068] In this embodiment, the energy distribution unit includes a load prediction module, a speed regulation module, an alternating controller, and a state estimation module. The load prediction module performs trend analysis on the historical load data of the computing power devices to obtain the load variation characteristics of the future period, providing forward-looking load information for charging and discharging decisions. The speed regulation module performs difference calculation on the measurement data of the flywheel speed sensor and the grid frequency detector to determine the matching degree parameter of the flywheel speed and the grid frequency, which reflects the coordination degree of the flywheel speed and the grid frequency. The alternating controller constructs a state space model of alternating control based on the load variation characteristics and the matching degree parameter, uses real-time load data as a constraint condition to modify the parameters of the controller, and outputs the modified flywheel charging and discharging rate. The state estimation module dynamically integrates the modified charging and discharging rate with the real-time matching degree parameter received by the speed regulation module to predict the operating mode threshold of the flywheel energy storage device in the next period. Through the collaborative work of each module, the energy distribution unit can more accurately make charging and discharging decisions and operating state predictions, achieving effective control of the flywheel energy storage device.

[0069] In constructing the load prediction module, a complete historical load database needs to be established first. This database should include the computing power device load data of different time periods (such as day, week, month, and quarter), as well as corresponding metadata such as timestamps, device types, and business scenarios. The data collection frequency should be determined according to system requirements and device characteristics, for example, for high-frequency fluctuating computing power devices, a second-level or millisecond-level collection frequency can be used. In the data preprocessing stage, the collected raw data needs to be cleaned to remove outliers and noise data. Outlier detection can use statistical-based methods (such as the 3σ principle) or machine learning algorithms (such as Isolation Forest, DBSCAN clustering). For missing data points, linear interpolation, spline interpolation, or time series prediction methods based on historical data can be used for filling.

[0070] To obtain the load change characteristics, the load prediction module uses multiple time series analysis methods. First, trend analysis is performed to identify long-term trends in load data through moving average, exponential smoothing, and other methods. For example, for periodically fluctuating load data, seasonal decomposition methods can be used to decompose the data into trend components, seasonal components, and residual components. At the same time, the periodicity of the load data is analyzed to determine its period length and fluctuation amplitude. For load data with obvious periodicity, a seasonal time series model (such as SARIMA) can be established for prediction. In addition, the volatility of the load data needs to be analyzed, and statistical indicators such as variance and standard deviation are calculated to assess the degree of load change.

[0071] Based on trend analysis, the load prediction module uses machine learning algorithms to build prediction models. The algorithms that can be used include support vector regression (SVR), random forest (Random Forest), long short-term memory network (LSTM), etc. For different types of computing power devices and business scenarios, appropriate algorithms need to be selected. For example, for load data with long-term dependencies, the LSTM network can better capture long-term patterns in time series. In the model training process, historical load data is divided into training and test sets, the training set is used to train the model, and the test set is used to evaluate the prediction performance of the model. To improve the generalization ability of the model, cross-validation, regularization, and other techniques can be used.

[0072] The load prediction module also needs to consider the impact of external factors on the load of computing power devices. These external factors include weekdays / holidays, holidays, weather conditions, business activities, etc. These external factors are input as features into the prediction model, which can further improve the accuracy of the prediction. For example, on weekdays and holidays, the load patterns of computing power devices usually differ significantly, and including the date type as a feature can help the model better capture this difference. For continuous external factors such as weather conditions, discretization or feature engineering methods can be used to extract valuable information.

[0073] The core task of the rotational speed regulation module is to calculate the matching degree parameter of the flywheel rotational speed and the grid frequency. First, the flywheel rotational speed sensor collects real-time rotational speed data of the flywheel, and the grid frequency detector monitors real-time grid frequency fluctuation data. These two sets of data need to be time-synchronized to ensure that the timestamps of the data are consistent for accurate difference calculation. Time synchronization can be calibrated by the synchronization signal Fly SYNC generated by the main control flywheel using a similar method as in Embodiment 1.

[0074] When calculating the difference, the rotational speed regulation module first converts the flywheel rotational speed data into the same physical dimension as the grid frequency. For example, the flywheel rotational speed (unit: revolutions per minute, RPM) is converted to frequency (unit: hertz, Hz). The conversion formula is: frequency (Hz) = flywheel rotational speed (RPM) / 60 × pole pairs. The pole pairs are an important parameter of the flywheel motor and depend on the design of the motor. The converted flywheel frequency is then subtracted from the grid frequency to obtain the frequency deviation value.

[0075] To more comprehensively reflect the matching degree of the flywheel rotational speed and the grid frequency, the rotational speed regulation module not only calculates the frequency deviation value, but also considers the frequency change rate (df / dt). The frequency change rate reflects the dynamic change characteristics of the grid frequency, which is crucial for a flywheel energy storage system that responds quickly. By performing a differential operation on the grid frequency data, the frequency change rate can be obtained. At the same time, a differential operation is performed on the flywheel rotational speed data to obtain the flywheel frequency change rate. Comparing these two change rates, the difference and ratio are calculated as part of the matching degree parameter.

[0076] The rotational speed regulation module also introduces a weight coefficient to comprehensively evaluate the matching degree. Different weights are assigned to the frequency deviation value and the frequency change rate difference. The size of the weight is dynamically adjusted according to the operating state of the system and control requirements. For example, in the case of large fluctuations in the grid frequency, the weight of the frequency change rate difference can be increased to pay more attention to the dynamic response performance of the system; in the case of relatively stable grid frequency, the weight of the frequency deviation value can be increased to improve the steady-state accuracy of the system. Finally, a comprehensive matching degree parameter is obtained by weighted summation. The smaller the parameter value, the higher the matching degree of the flywheel rotational speed and the grid frequency.

[0077] The alternate controller constructs a state space model for alternate control based on the load variation characteristics provided by the load prediction module and the matching degree parameter calculated by the rotational speed regulation module. The state space model is a commonly used control system modeling method that can comprehensively describe the dynamic characteristics of the system. In this system, the state variables of the state space model include the rotational speed, remaining energy, and charging and discharging rate of the flywheel energy storage device, the input variables include the load prediction value and the matching degree parameter, and the output variable is the corrected flywheel charging and discharging rate.

[0078] The first step in constructing the state-space model is to determine the state equation and the output equation of the system. The state equation describes the relationship between the system state variables and time, while the output equation describes the relationship between the system output variables and the state variables and input variables. When establishing the equations, the physical characteristics and control logic of the system need to be considered. For example, there is a certain physical relationship between the speed change of the flywheel and the charging and discharging power, which can be established according to the law of conservation of energy. At the same time, considering the nonlinear characteristics of the system, the nonlinear equation can be linearized by methods such as Taylor expansion.

[0079] In order to improve the accuracy of the model, the alternate controller uses real-time load data as a constraint condition to modify the parameters of the state-space model. Specifically, the real-time collected load data is compared with the predicted value of the load prediction module to calculate the prediction error. According to the size and direction of the prediction error, the parameters of the state-space model are adjusted so that the model can better reflect the actual operation of the system. Parameter modification can use adaptive control algorithms (such as model reference adaptive control, self-tuning PID control) or optimization algorithms (such as least squares method, gradient descent method).

[0080] The alternate controller also needs to consider the constraints of the system. These constraints include the maximum charging and discharging power limit of the flywheel energy storage device, the upper and lower limits of the speed, the remaining energy range, etc. When solving the state-space model, these constraints are used as constraints of the optimization problem to ensure that the charging and discharging rate output by the controller is within the allowable range of the system. For example, when the remaining energy of the flywheel is close to the lower limit, the discharging rate is limited to avoid excessive discharging; when the speed of the flywheel is close to the upper limit, the charging rate is limited to prevent the flywheel from running at high speed.

[0081] The main function of the state estimation module is to predict the operating mode threshold of the flywheel energy storage device in the next period. This module dynamically integrates the modified charging and discharging rate output by the alternate controller, combined with the real-time matching degree parameter received by the speed regulation module. The dynamic integration process takes into account the inertia and delay characteristics of the system, and by integrating the charging and discharging rate over time, the energy change of the flywheel energy storage device in the future period is obtained.

[0082] During the integration process, the state estimation module introduces a decay factor to simulate the energy loss of the system. The size of the decay factor depends on the efficiency of the flywheel energy storage device, heat dissipation, and other factors. The specific value of the decay factor is determined through experiments or theoretical analysis and is dynamically adjusted during the integration process. At the same time, considering the uncertainty of the grid frequency fluctuation, the state estimation module performs probability analysis on the matching degree parameter to calculate the energy change range under different matching degrees.

[0083] Based on the dynamic integration results, the state prediction module calculates the operating mode threshold of the flywheel energy storage device in the next period. The operating mode threshold includes the charging threshold, the discharging threshold and the standby threshold. When the predicted energy change exceeds the charging threshold, the system will switch to the charging mode; when the predicted energy change is lower than the discharging threshold, the system will switch to the discharging mode; when the predicted energy change is within the standby threshold range, the system will maintain the current state or enter the standby mode.

[0084] The state prediction module also needs to consider the response time and safety margin of the system. In order to ensure that the system can respond to load changes and grid frequency fluctuations in time, a certain safety margin is reserved when calculating the threshold. For example, set the charging threshold slightly higher than the actual need, and set the discharging threshold slightly lower than the actual need, to avoid frequent switching of the system operating mode. At the same time, according to the response time characteristics of the system, adjust the size of the threshold to ensure that the system can complete mode switching within the specified time.

[0085] In actual operation, the state prediction module continuously updates the prediction results and adjusts the operating mode threshold according to the latest prediction. When the operating state of the system changes, such as sudden increase of load or large fluctuation of grid frequency, the state prediction module can quickly recalculate the threshold to provide timely and accurate decision basis for the energy distribution unit. Through this dynamic prediction and adjustment mechanism, the system can make preparations in advance, improve the response ability to load changes and grid fluctuations, and realize more efficient energy distribution and control.

[0086] Through the collaborative work of the load prediction module, the speed regulation module, the alternating controller and the state prediction module, the energy distribution unit can realize precise control of the flywheel energy storage device. The load prediction module provides forward-looking load information to help the system plan charging and discharging strategies in advance; the speed regulation module calculates the matching degree parameter of flywheel speed and grid frequency to provide key input for the controller; the alternating controller makes parameter correction and control decision based on state space model and real-time data; the state prediction module predicts the future operating mode threshold to guide the dynamic adjustment of the system. This multi-level, multi-dimensional control architecture makes the system adapt to different operating conditions and external environmental changes, realizes efficient recovery and utilization of data center computing power load energy, and improves the stability and reliability of the entire system. EMBODIMENT

[0087] The alternative controller of this embodiment adopts a dual-mode switching strategy. When the grid frequency fluctuation exceeds the set threshold, the flywheel speed compensation amount is calculated based on the matching degree parameter, and the charge and discharge rate correction instruction is generated to respond to the impact of grid frequency fluctuation on the system, ensuring the stability of the system. When the load change rate exceeds the set threshold, the flywheel charge and discharge priority sequence is generated based on the load prediction result, and the cooperative scheduling mechanism of the energy storage device is triggered, so that the energy storage device can adjust the charge and discharge strategy in time according to the rapid change of the load, improving the response ability of the system to the load change. In the dual-mode switching process, a dynamic weight distribution algorithm is used for control parameter optimization, and the weight of the control parameter is dynamically adjusted according to different operating conditions and parameter changes, further improving the performance and adaptability of the controller.

[0088] When constructing the dual-mode switching strategy of the alternative controller, the grid frequency fluctuation threshold and the load change rate threshold need to be set first. The setting of these thresholds needs to consider the stability requirements of the grid, the performance limitations of the flywheel energy storage device, and the operating requirements of the data center computing equipment. The grid frequency fluctuation threshold is usually determined according to the nominal frequency of the grid (such as 50Hz or 60Hz) and the allowed fluctuation range, for example, set to ±0.5Hz or ±1Hz. The load change rate threshold is obtained by statistical analysis of the historical load data of the data center according to the business characteristics and load data, for example, set to not more than 10% or 20% of the rated load per minute.

[0089] When the grid frequency fluctuation exceeds the set threshold, the alternative controller enters the frequency compensation mode. In this mode, the flywheel speed compensation amount is further calculated based on the matching degree parameter calculated by the speed regulation module. The matching degree parameter reflects the coordination degree of the flywheel speed and the grid frequency, including the frequency deviation value and the frequency change rate difference value. By analyzing these parameters, it is determined how much the flywheel speed needs to be adjusted to compensate for the fluctuation of the grid frequency.

[0090] When calculating the flywheel speed compensation amount, multiple factors need to be considered. First, the fluctuation amplitude and direction of the grid frequency, the larger the fluctuation amplitude, the larger the compensation amount needed; the fluctuation direction determines whether the flywheel speed needs to be increased or decreased. Second, the current state of the flywheel energy storage device, including speed, remaining energy, etc. If the flywheel speed is close to the upper or lower limit, or the remaining energy is insufficient, the size of the compensation amount may need to be limited to ensure the safe operation of the system. In addition, the response time and inertia characteristics of the system also need to be considered to avoid overcompensation or insufficient compensation leading to system oscillation.

[0091] Based on the calculated flywheel speed compensation amount, the alternate controller generates a charge-discharge rate correction instruction. This instruction will be sent to the flywheel control unit to control the flywheel energy storage device to adjust the charge-discharge rate, thereby achieving the regulation of the flywheel speed. When generating the correction instruction, the charge-discharge efficiency of the flywheel and the energy conversion loss need to be considered. Different charge-discharge rates may correspond to different efficiencies, so it is necessary to select a charge-discharge rate that can meet the speed regulation demand and ensure high energy conversion efficiency.

[0092] When the load change rate exceeds the set threshold, the alternate controller enters the load response mode. In this mode, based on the load prediction results provided by the load prediction module, a flywheel charge-discharge priority sequence is generated. The load prediction results include the load change trend, peak and valley of the future period and other information. According to these information, the ability and adaptability of each flywheel energy storage device in response to load changes are evaluated, so as to determine their charge-discharge priority.

[0093] When generating the charge-discharge priority sequence, multiple factors need to be considered. First of all, the remaining energy level of the flywheel energy storage device, the higher the remaining energy of the flywheel, the more energy support it can provide in discharge mode, so the priority is higher; in charge mode, the flywheel with low remaining energy needs to be charged first to restore the energy storage capacity. Secondly, the response speed and adjustment ability of the flywheel, different flywheels may have different response time and adjustment range, the flywheel with fast response speed and strong adjustment ability has high priority. In addition, the historical running state and health condition of the flywheel also need to be considered to avoid overuse of some flywheels leading to premature damage.

[0094] Based on the generated charge-discharge priority sequence, the alternate controller triggers the cooperative scheduling mechanism of the energy storage device. According to the priority sequence, the mechanism coordinates the charge-discharge operation of multiple flywheel energy storage devices, so that they can work cooperatively to jointly respond to the rapid change of load. In the process of cooperative scheduling, it is necessary to ensure that the charge-discharge operation of each flywheel does not interfere with each other, while maximizing the overall response ability and energy utilization efficiency of the system.

[0095] In the dual-mode switching process, the alternate controller uses a dynamic weight allocation algorithm to optimize the control parameters. This algorithm dynamically adjusts the weight of the control parameters according to different operating conditions and parameter changes. Operating conditions include grid frequency fluctuation, load change rate, state of flywheel energy storage device and other factors. By monitoring these factors in real time, the current operating state of the system is evaluated, and then the weight of the control parameters is dynamically adjusted according to the preset rules or learning algorithm.

[0096] The core of the dynamic weight allocation algorithm is to establish a weight adjustment model. This model can be based on rules, machine learning or optimization algorithms. The rule-based model adjusts the weight according to the preset conditions and rules, such as increasing the weight of the frequency compensation mode related parameters when the grid frequency fluctuates greatly; when the load change rate is large, increase the weight of the load response mode related parameters. The machine learning model automatically discovers the optimal weight allocation scheme under different operating conditions through learning from historical data. The optimization algorithm model converts the weight adjustment problem into an optimization problem, and determines the weight by solving the optimal solution.

[0097] In practical applications, the dynamic weight allocation algorithm needs to obtain the system's running state data in real time and make quick calculations and decisions. In order to improve the efficiency and real-time performance of the algorithm, distributed computing architecture or hardware acceleration technology can be used. At the same time, the algorithm needs to be updated and optimized regularly to adapt to the changing operating environment and system requirements.

[0098] The alternating controller also needs to consider the smooth transition problem in the mode switching process. When the system switches from one mode to another, if the switching process is too abrupt, it may cause the system to be unstable or produce large fluctuations. In order to avoid this situation, the alternating controller adopts a smooth transition strategy, gradually adjusts the control parameters and instructions during the mode switching process, so that the system can smoothly transition from one operating state to another.

[0099] The smooth transition strategy can be implemented in several ways. One method is to pre-calculate and prepare the mode to be entered before mode switching, and then gradually apply new control parameters and instructions at the switching time. Another method is to use an interpolation algorithm to interpolate between the control parameters of the two modes, generating a smooth transition curve so that the system's response can gradually change along this curve. In addition, a transition time period can be set, during which the influence of both modes is considered, then gradually reduce the influence of the original mode and increase the influence of the new mode until the complete switching to the new mode.

[0100] The alternating controller also needs to have fault detection and handling capabilities. During the dual-mode switching process, various faults may occur, such as sensor faults, communication faults, actuator faults, etc. The alternating controller needs to monitor the system's operating state in real time, detect faults in a timely manner and take appropriate handling measures. For example, when a sensor fault is detected, the data of the redundant sensor or the estimated value based on the model can be used to replace it; when a communication fault occurs, it can try to re-establish the communication connection or switch to the backup communication channel; when the actuator fails, it can promptly shut down the faulty device and adjust the operating state of other devices to ensure the safe and stable operation of the system.

[0101] To improve the reliability and maintainability of the alternative controller, a perfect log recording and diagnosis function is also needed. The log recording function can record the running state, control parameters, fault information, etc. of the system, providing the basis for subsequent analysis and maintenance. The diagnosis function can evaluate the running state of the system in real time, find potential problems and provide corresponding suggestions and solutions. Through these functions, the operation and maintenance personnel can timely understand the running situation of the system, quickly locate and solve problems, and reduce the downtime and maintenance cost of the system.

[0102] Through the dual-mode switching strategy and dynamic weight distribution algorithm of the alternative controller, the system can automatically adjust the control strategy according to different operating conditions, improving the response ability to grid frequency fluctuations and load changes. The frequency compensation mode can effectively respond to grid frequency fluctuations and maintain the stability of the system; the load response mode can quickly respond to load changes and ensure the normal operation of the data center computing power equipment. The dynamic weight distribution algorithm further optimizes the control parameters, so that the system can achieve the best control effect under various operating conditions. This intelligent and adaptive control strategy provides strong technical support for the flywheel energy storage system of the data center computing power load energy recovery, making it better meet the requirements of modern data centers for energy efficiency and stability. Embodiment

[0103] The alternative controller of this embodiment uses load change characteristics and matching degree parameters to construct a state vector, which is composed of flywheel speed compensation under grid frequency reference, charge-discharge priority coefficient, and remaining capacity proportion of energy storage device. At the same time, combined with load prediction error correction factor, frequency fluctuation compensation factor and mode switching delay parameter to establish state transition equation. Using real-time load data as boundary condition to modify the parameters of the controller, and the modified flywheel charge-discharge rate as output. In the control process, when the flywheel operating mode or load monitoring object switches, the following steps are taken to achieve fast response:

[0104] Firstly, the state vector construction of the alternative controller needs to integrate multi-dimensional information. The flywheel speed compensation under the grid frequency reference is generated based on the frequency deviation and rate of change calculated by the speed regulation module, reflecting the dynamic matching demand of flywheel speed and grid frequency. The charge-discharge priority coefficient is determined by the load change trend output by the load prediction module and the current state of the energy storage device (such as remaining capacity, speed), which is used to quantify the priority order of different flywheels in charge-discharge scheduling. The remaining capacity proportion of the energy storage device is directly related to the available energy of the flywheel, which is a key indicator to determine whether it can participate in the current charge-discharge task. These three components are the core elements of the state vector, which are normalized to ensure dimensional consistency, making it easier for subsequent mathematical modeling.

[0105] The state transition equation is established by considering the dynamic characteristics of the system and external disturbance factors. The load prediction error correction factor is used to compensate for the deviation between the predicted value and the real-time load data, and the equation parameters are dynamically adjusted by monitoring the load fluctuation amplitude in real time. The frequency fluctuation compensation factor is based on the measured data of the grid frequency detector, and reflects the influence of frequency fluctuation on the flywheel speed and energy conversion efficiency. The mode switching delay parameter is used to simulate the time lag of the controller from receiving the switching instruction to completing the state adjustment, and the specific value is determined by historical data statistics or hardware characteristic test. The state transition equation describes the evolution of the state vector over time in the form of a differential equation or a difference equation, for example, by simulating the delay characteristics of flywheel speed change through a first-order inertia link, or reflecting the coupling relationship between parameters through a linear combination model.

[0106] When the real-time load data is input, the alternating controller uses it as a boundary condition to modify the parameters of the state transition equation. The specific process is as follows: First, calculate the absolute error and relative error between the predicted value and the measured value of the load, if the error exceeds the preset threshold (such as 5% of the rated load), the correction mechanism is triggered. The correction method includes adjusting the weight coefficient of each component in the state vector, updating the gain parameter of the state transition equation, or reinitializing the integral term of the equation. For example, when the measured load is significantly higher than the predicted value, increase the weight of the charge-discharge priority coefficient to speed up the discharge rate of the high-priority flywheel; if the grid frequency fluctuation is intensified, increase the value of the frequency fluctuation compensation factor to enhance the dynamic adjustment of the flywheel speed.

[0107] In the flywheel operation mode or load monitoring object switching scenario, the fast response mechanism is realized through the following steps:

[0108] Data acquisition and synchronization: According to the timestamp of the synchronization signal Fly SYNC, capture the charge-discharge rate data output by the alternating controller at the switching time. This data includes the current charge-discharge rate instruction value and state vector parameters of each flywheel, and is transmitted to the buffer module of the energy distribution unit with a delay of microseconds through a dedicated communication channel, ensuring the real-time and integrity of the data.

[0109] Available energy reserve calculation: Based on the charge-discharge rate data and the energy capacity parameters of each flywheel energy storage device (such as rated capacity, energy conversion efficiency), the available energy reserve of each flywheel group at the switching time is calculated by integral operation. The calculation formula is: available energy = initial residual energy + charge-discharge rate x switching time interval x efficiency factor. The efficiency factor takes into account the energy loss in the charging and discharging process (such as mechanical loss, electromagnetic loss), which is determined by the equipment nameplate parameters or historical operation data.

[0110] Matching degree calculation and threshold determination: Combine the real-time data of the load monitoring unit (such as the active power and reactive power demand of the current computing device) to calculate the matching degree of the energy reserve and the load demand. The matching degree index can be defined as the ratio of available energy to load demand, or the Euclidean distance to measure the deviation of energy supply and demand. According to the matching degree calculation result, set the dispatchable energy threshold for each flywheel energy storage device, which needs to reserve a certain safety margin (such as 10% of the rated capacity) to avoid excessive charging and discharging that can damage the device.

[0111] Deviation calculation and dispatch queue generation: Obtain the actual operating state parameters of each flywheel at the switching time (such as real-time speed, remaining energy, charging and discharging current), and compare them with the adjustable energy threshold to calculate the deviation. The deviation calculation formula is: Deviation = |actual operating state parameter - threshold parameter| / threshold parameter x 100%. When the deviation of a certain flywheel is lower than the set threshold (such as 15%), it is determined to be in a dispatchable state, and it is included in the current dispatch queue, and the rate adjustment instruction is generated according to the charging and discharging priority coefficient; if the deviation exceeds the threshold, the warning mechanism is triggered, prompting the operation and maintenance personnel to check the device status.

[0112] In the rapid response process, the time synchronization of each link needs to be ensured. The synchronization signal Fly SYNC is not only used for data acquisition triggering, but also as the time reference for each calculation step, through the global clock calibration mechanism (such as IEEE 1588 precision clock protocol) to ensure that the timestamp error between different modules is less than 1 microsecond. In addition, the energy distribution unit needs to have high-speed data processing capability, through field programmable gate array (FPGA) or graphics processing unit (GPU) to accelerate the calculation process, to ensure that the total delay from data acquisition to instruction output is controlled within the millisecond level.

[0113] The parameters of the state vector and the state transition equation need to be optimized offline regularly. Through historical operation data review, intelligent optimization algorithms such as particle swarm optimization (PSO) and genetic algorithm (GA) are used to iteratively optimize the weight coefficients of the state vector and the gain parameters of the state transition equation to adapt to the long-term changes in data center load characteristics (such as changes in load patterns caused by business expansion and equipment upgrades). During the optimization process, constraints need to be set to ensure that the parameter values are within the safe operating range of the device and to avoid excessive optimization that can reduce system robustness.

[0114] In summary, embodiment 4 realizes fine modeling and control of flywheel energy storage devices by constructing a state vector containing multi-dimensional parameters and a dynamically adjusted state transition equation. The rapid response mechanism ensures that the system can quickly adjust the scheduling strategy and maintain energy supply and demand balance when the operating mode or monitoring object is switched through synchronous data acquisition, energy reserve calculation, matching degree analysis, and deviation determination. This scheme deeply couples load changes, grid frequency fluctuations, and flywheel states, forming a closed-loop control system, which provides key technical support for the efficient and stable operation of data center computing power load energy recovery systems. Embodiments

[0115] In this embodiment, each group of flywheel energy storage devices contains multiple independent flywheel units, and each flywheel unit is connected to the data center power network in a set topology. The connection mode within each flywheel group is either a ring redundancy structure or a star centralized structure, and the power sensor arrangement covers the key power nodes of the computing equipment to achieve comprehensive monitoring of the load.

[0116] I. Flywheel unit topology design

[0117] ① Ring redundancy structure

[0118] In this structure, three independent flywheels (denoted as F1, F2, and F3) are connected in series through power electronic interfaces to form a closed loop, and the center node (denoted as Node_C) is connected to the power distribution bus of the data center. The specific connection method is as follows: the output of F1 is connected to the input of F2, the output of F2 is connected to the input of F3, and the output of F3 is connected in reverse to the input of F1, forming a ring circuit. The center node Node_C is drawn from the connection point of F1 and F2 and connected to the power network through a bus cable. The advantage of this structure is redundancy: when any one flywheel fails (such as F2), the ring circuit automatically disconnects the fault point, and the remaining two flywheels (F1 and F3) can continue to supply power to the bus through Node_C, maintaining system operation. At the same time, the ring structure can balance the charge and discharge load of each flywheel, avoiding overload of a single device.

[0119] ② Star centralized structure

[0120] The structure contains four independent flywheels (denoted as F4, F5, F6, F7), all in parallel to the same power node (denoted as Node_S). The input / output end of each flywheel is directly connected to Node_S through an independent cable, and the central node is led out from Node_S, while the output end of the load monitoring unit is connected. The star structure is characterized by centralized control: the load monitoring unit can obtain the voltage and current data of all flywheels in real time through Node_S, which facilitates the rapid calculation of the total energy storage capacity and the charging and discharging rate. When the number of flywheels needs to be expanded, the new flywheels can be simply connected in parallel to Node_S, without the need to change the original topology, and the expansibility is strong. However, the reliability of the central node Node_S is relied on, and a redundant power switch needs to be configured to prevent single-point failure.

[0121] II. Power sensor arrangement scheme

[0122] The power sensor is deployed at the three-level monitoring node of the data center power supply line, covering the full-link power input of the computing power equipment:

[0123] Primary node: AC input terminal

[0124] A miniature power sensor is deployed at the AC input port (such as IEC C13 / C14 interface) of the computing power server, and each sensor monitors the real-time power (active power P, reactive power Q, and apparent power S) of a single server. The sensor is fixed to the outside of the terminal through magnetic or buckle structure, supports plug-and-play, and the sampling frequency is not less than 1 kHz to capture high-frequency load fluctuations (such as CPU instantaneous peak power consumption).

[0125] Secondary node: DC power distribution cabinet

[0126] Hall current sensors and voltage transmitters are installed in the output branch of the DC power distribution cabinet to monitor the DC current I_dc and voltage U_dc of each branch. The sensors are connected to the load monitoring unit through RS-485 bus with a communication period of 100 ms, which is used to calculate the regional load (such as the total power consumption of all servers in a cabinet).

[0127] Third-level node: Uninterruptible power supply (UPS) output port

[0128] A high-precision power analyzer is deployed on the AC output bus of the UPS to monitor the total output power P_total and frequency f. The sensor has harmonic analysis function and can detect voltage distortion rate THD_u and current distortion rate THD_i with a sampling period of 1 s, which is used to evaluate the power quality of the entire data center and the compensation effect of the flywheel energy storage system.

[0129] Topology structure and monitoring node coordination

[0130] The topology of the flywheel unit and the arrangement of the power sensor form a data closed loop:

[0131] ①In the ring redundancy structure, when the load fluctuation of a certain computing device group is captured by the first-level node sensor, the load monitoring unit analyzes the real-time power data to determine whether to trigger flywheel charging and discharging. If discharging is needed, the flywheel with the highest remaining capacity in the ring loop (such as F1) is preferentially selected to release energy by adjusting its speed, while the states of other flywheels (F2, F3) are monitored as backups.

[0132] ②In the star centralized structure, when the UPS output frequency fluctuation is detected by the third-level node sensor (such as f deviating from 50Hz ± 0.2Hz), the flywheel control unit adjusts the speed of all parallel flywheels according to the frequency detector data. For example, when the frequency is lower than the nominal value, the flywheel is controlled to accelerate and discharge to inject active power into the grid; when the frequency is higher than the nominal value, the flywheel is controlled to decelerate and charge to absorb excess energy.

[0133] Formula explanation (only contains 1 structure-related formula)

[0134] In the ring redundancy structure, the equivalent output voltage U_ring of the flywheel group can be represented as:

[0135] Where: U_node_C is the output voltage of the center node Node_C of the ring structure (unit: V); U_F1, U_F2, and U_F3 are the terminal voltages of flywheels F1, F2, and F3 respectively (unit: V), and their polarity is determined by the connection direction (F1 and F2 are connected in forward direction, and F3 is connected in reverse direction).

[0136] This formula reflects the vector superposition characteristics of the voltages of the flywheels in the ring structure. By adjusting the speed of different flywheels (changing the back electromotive force), the output voltage of Node_C can be dynamically adjusted to match the power grid demand or the power supply voltage fluctuation of the computing device.

[0137] It should be noted that, in this document, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0138] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the exact construction and arrangement described. It is intended, therefore, to cover all modifications and adaptions that fall within the scope of the claims and their equivalents.

Claims

1. A flywheel energy storage system control system for data center hash power load energy recovery, characterized in that, The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application relates to a data center energy storage system. The application (4) Obtain the switching time, the actual operating state parameters of each flywheel energy storage device, and calculate the deviation between the actual operating state parameters of each flywheel energy storage device and the dispatchable energy threshold of each flywheel energy storage device; When the deviation of a flywheel is lower than the set threshold, the flywheel is included in the current dispatch queue and the charge and discharge rate adjustment instruction is executed.

2. The flywheel energy storage system control system for data center computing power load energy recovery of claim 1, wherein, The timing synchronization control is adopted for the energy storage unit, the load monitoring unit and the flywheel control unit; Specifically, it includes: Each flywheel energy storage device in the energy storage unit is numbered, and one of them is used as the master flywheel to generate the synchronization signal Fly SYNC; The remaining flywheels synchronize the switching mode after receiving the synchronization signal Fly SYNC, and send the operating state data with flywheel number to the energy distribution unit; The multiple groups of power sensors of the load monitoring unit are numbered; The sampling frequency of each group of power sensors is adjusted by the synchronization signal Fly SYNC, and the adjusted sensor number information is sent to the energy distribution unit; The synchronization signal Fly SYNC is transmitted to the flywheel control unit at the same time, and the flywheel speed sensor and the grid frequency detector are controlled to synchronize data collection.

3. The flywheel energy storage system control system for data center computing power load energy recovery according to claim 1, characterized in that: The alternating controller adopts a dual-mode switching strategy, in which, When the grid frequency fluctuation exceeds the set threshold, the flywheel speed compensation is calculated based on the matching degree parameter, and the charge and discharge rate correction instruction is generated; When the load change rate exceeds the set threshold, the flywheel charge and discharge priority sequence is generated based on the load prediction result, and the cooperative scheduling mechanism of the energy storage device is triggered.

4. The flywheel energy storage system control system for data center computational load energy recovery of claim 3, wherein, In the dual-mode switching process, a dynamic weight distribution algorithm is used for control parameter optimization.

5. The flywheel energy storage system control system for data center computational load energy recovery of any one of claims 1-4, wherein, Each group of flywheel energy storage devices includes multiple independent flywheel groups; and the multiple independent flywheels of each group of flywheel groups are connected to the power network in a set topology.

6. The flywheel energy storage system control system for data center computational load energy recovery of claim 1, wherein, The connection mode of the independent flywheels in each group of flywheel groups is a ring redundancy structure or a star centralized structure; wherein, In the ring redundancy structure, three flywheels are connected in series to form a closed loop, and the center node is connected to the power distribution bus; In the star centralized structure, four flywheels are connected in parallel to the same power node, and the center node is connected to the output end of the load monitoring unit.

7. The flywheel energy storage system control system for data center computational load energy recovery of claim 1, wherein, The arrangement position of the power sensor covers the AC input terminal, the DC power distribution cabinet and the uninterruptible power supply output port of the computing power equipment.

Images