Outdoor arrangement flywheel UPS power supply system based on data center

Abstract

The invention relates to the technical field of data center power supply systems, and discloses a data center-based outdoor arrangement flywheel UPS power supply system, which comprises a region division module, a flywheel monitoring module, a parameter generation module and a mode adjustment module. The area division module divides a power supply area into a main energy storage verification sub-area and a standby electric energy dispatching sub-area. The flywheel monitoring module is provided with a flywheel energy storage device, collects rotating speed and temperature data, and monitors operation stability through vibration period calibration and a temperature equalization judgment strategy. The parameter generation module obtains a power transmission path and load fluctuation data, and generates dynamic power distribution parameters by applying a dynamic voltage compensation and harmonic suppression strategy; and the mode adjustment module adaptively adjusts the output modes of the standby electric energy dispatching sub-region power conversion device according to the parameters, wherein the output modes comprise redundant power supply, fault switching priority adjustment, harmonic isolation and the like. The system improves the power supply stability, reliability and power quality of the data center, and is suitable for outdoor arrangement scenes.

Description

Technical Field

[0001] This invention relates to the field of data center power supply system technology, specifically to an outdoor flywheel UPS power supply system for data centers. Background Technology

[0002] With the rapid development of information technology, data centers, as the core hubs for information storage, processing, and transmission, are constantly increasing in scale and complexity, placing extremely high demands on the reliability, stability, and efficiency of power supply systems. Traditional UPS (Uninterruptible Power Supply) power supply systems, especially those using batteries as energy storage devices, have many insurmountable drawbacks.

[0003] Batteries have a relatively short lifespan, typically 3-5 years, requiring frequent replacements. This not only increases maintenance costs but can also disrupt the normal operation of data centers if replacements are not timely. Furthermore, batteries generate significant heat during charging and discharging; improper heat dissipation management can easily lead to safety hazards such as thermal runaway. Their low energy density also makes it difficult to meet the ever-increasing power density demands of data centers. In addition, traditional UPS systems lack the flexibility and dynamic adjustment capabilities to handle load fluctuations, failing to quickly and accurately allocate power. This can easily lead to unstable power supply in certain areas, and even problems such as voltage drops and harmonic interference, affecting the normal operation and lifespan of equipment within the data center.

[0004] On the other hand, the outdoor deployment of data centers places more stringent environmental adaptability requirements on power supply systems. The outdoor environment is complex and variable; factors such as temperature, humidity, and vibration can all affect the performance and stability of power supply equipment. Traditional power supply systems deployed outdoors lack effective monitoring and adaptive adjustment mechanisms, making it difficult to detect environmental changes and abnormal equipment operating conditions in real time, and thus unable to take timely countermeasures, increasing the risk of system failure.

[0005] Furthermore, with the promotion of the green data center concept, higher requirements have been placed on the energy efficiency and environmental friendliness of power supply systems. Traditional battery systems cause some pollution to the environment during production, use, and disposal, and their energy conversion efficiency is relatively low, which does not meet the requirements of sustainable development. Therefore, there is an urgent need for a new type of UPS power supply system that can adapt to outdoor data center deployment scenarios and has higher reliability, stability, efficiency, and environmental friendliness. Summary of the Invention

[0006] The purpose of this invention is to provide an outdoor flywheel UPS power supply system for data centers to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an outdoor flywheel UPS power supply system for data centers, the system comprising: The area division module is used to divide the power supply area of ​​the data center into a main energy storage verification sub-area and several backup power dispatch sub-areas. The flywheel monitoring module is used to deploy at least one flywheel energy storage device in the main energy storage verification sub-area, collect real-time speed and temperature data of each flywheel device, and determine the operational stability of the flywheel device based on the flywheel status monitoring strategy. The parameter generation module is used to obtain power transmission path data between the main energy storage verification sub-area and the backup power dispatch sub-area, and combine it with real-time load fluctuation data to import the power conversion control strategy and generate dynamic power allocation parameters. The mode adjustment module is used to adaptively adjust the output mode of all power conversion devices in the standby power dispatch sub-region according to dynamic power distribution parameters.

[0008] Preferably, a power monitoring terminal is provided at the boundary of the main energy storage verification sub-region, and a power distribution terminal is provided at the access node of each backup power dispatch sub-region; The flywheel condition monitoring strategy includes: a flywheel vibration period calibration strategy; The flywheel vibration period calibration strategy is as follows: When the power monitoring terminal detects that the flywheel device has started, it acquires the speed fluctuation data of the flywheel device in the first 10 sampling intervals according to the preset sampling interval, forming a speed fluctuation sequence. The speed fluctuation data includes: the speed difference between adjacent sampling intervals and the speed change acceleration. The sum of the speed difference between three consecutive sampling intervals in the speed fluctuation sequence is calculated as the speed anomaly index. Sampling intervals with speed anomaly indices exceeding the threshold are selected, and two sampling intervals before and after the selected interval are extracted to form the flywheel vibration characterization period.

[0009] Preferably, the flywheel status monitoring strategy further includes: a temperature equalization determination strategy; The specific temperature equalization determination strategy is as follows: Extract temperature data within the flywheel vibration characterization period and calculate the correlation coefficient between the temperature rise rate and the rotational speed fluctuation data; Retrieve the temperature equalization threshold range of all flywheel devices within the same vibration cycle from historical operating data; If the correlation coefficient between the current rate of temperature rise and the speed fluctuation data exceeds the threshold range, it is determined that the flywheel device has a risk of thermal imbalance, triggering the pre-start command of the backup power dispatch sub-region.

[0010] Preferably, the power conversion control strategy includes: a dynamic voltage compensation strategy; The dynamic voltage compensation strategy is as follows: Calculate the minimum compensation voltage value for each backup power dispatch sub-region based on the line impedance parameters in the power transmission path data; The output voltage fluctuation of the main energy storage verification sub-area is monitored in real time. When the fluctuation exceeds the preset tolerance, the output mode of the power distribution terminal is adjusted to N times the minimum compensation voltage value, where N is a preset constant.

[0011] Preferably, the power conversion control strategy further includes: a harmonic suppression strategy; The harmonic suppression strategy is specifically as follows: The harmonic distortion rate of the current at the output of the power conversion device is collected. If the distortion rate of three consecutive sampling intervals exceeds the threshold, a harmonic suppression command is generated. The pulse width modulation parameters of the power conversion device are adjusted according to the harmonic suppression command until the distortion rate drops below the threshold.

[0012] Preferably, the adaptive adjustment includes: Extract the backup power dispatch sub-region number corresponding to the flywheel device with thermal imbalance risk; When the thermal imbalance risk level is Level 1, the power conversion device of the corresponding backup power dispatch sub-area will be switched to redundant power supply mode, wherein the redundant power supply mode includes: Close the power transmission path from the main energy storage verification sub-area to that sub-area; Enable the cross-power supply link of adjacent backup power dispatch sub-regions and limit the maximum load capacity of the sub-region to 80% of the rated value.

[0013] Preferably, the adaptive adjustment further includes: When the thermal imbalance risk level is level two, historical power supply stability data is retrieved from the power distribution terminal of the backup power dispatch sub-region. A power supply reliability model is constructed based on historical power supply stability data, and the fault switching priority of each sub-region is marked. The output phase of the power conversion device is dynamically adjusted according to the fault switching priority.

[0014] Preferably, the adaptive adjustment further includes: Real-time acquisition of harmonic components of the output current of each power conversion device; Calculate the amplitude proportion of the 2nd to 13th harmonics in the harmonic components. If the amplitude proportion of a certain harmonic exceeds 5%, a harmonic isolation command is triggered. The output circuit of the corresponding power conversion device is cut off according to the harmonic isolation command, and the residual harmonic energy is absorbed by the parallel capacitor bank of the adjacent sub-region.

[0015] Preferably, the power supply reliability model includes: Count the number of voltage drops and their duration in each sub-region over the past 24 hours; If a sub-region experiences more than 3 voltage drops with each drop lasting more than 10 seconds, the fault switching priority for that sub-region will be raised to the highest level.

[0016] Preferably, the power supply reliability model further includes: The load mutation rate of each sub-region is collected in real time, and the correlation coefficient between the mutation rate and the voltage phase deviation is calculated. If the correlation coefficient exceeds the preset threshold, phase compensation parameters are generated based on historical fault data, and the output frequency of the power conversion device is adjusted to the target phase compensation value simultaneously.

[0017] Compared with the prior art, the beneficial effects of the present invention are: The system divides the power supply area into a primary energy storage verification sub-area and a backup power dispatch sub-area through a regional division module, achieving an optimized design of the primary and backup power supply structure. The flywheel energy storage device within the primary energy storage verification sub-area serves as the core energy storage unit, featuring high energy density, long service life, and rapid charging and discharging. This overcomes the problems of short lifespan and high maintenance costs associated with traditional batteries, reducing the need for frequent replacement of energy storage equipment and lowering maintenance workload and costs. Simultaneously, the establishment of the backup power dispatch sub-area forms a redundant power supply system. When the primary energy storage area or a specific backup energy dispatch sub-area malfunctions, other backup energy dispatch sub-areas can quickly take over the power supply task, significantly improving the system's fault tolerance and power supply continuity.

[0018] The flywheel monitoring module collects real-time speed and temperature data from the flywheel assembly and, combined with a flywheel condition monitoring strategy, achieves precise monitoring and evaluation of the flywheel's operating status. Specifically, the flywheel vibration cycle calibration strategy analyzes speed fluctuation data to accurately identify the characteristic cycle of flywheel vibration, promptly detecting mechanical anomalies such as bearing wear and rotor imbalance—early signs of faults—facilitating early maintenance and repair, preventing the fault from escalating, and ensuring stable operation of the flywheel assembly. The temperature equalization judgment strategy calculates the correlation coefficient between the temperature rise rate and speed fluctuation data and compares it with historical temperature equalization threshold ranges to effectively determine if the flywheel assembly is at risk of thermal imbalance. Once a risk is detected, a pre-start command for the backup power dispatch sub-area is triggered promptly to prepare for power switching in advance, ensuring that power switching is completed before a thermal fault occurs in the flywheel assembly, avoiding power outages caused by thermal failures.

[0019] The parameter generation module combines power transmission path data and real-time load fluctuation data, imports power conversion control strategies to generate dynamic power distribution parameters, enabling the system to dynamically adjust the power distribution of each sub-area according to actual load conditions and power transmission conditions. The dynamic voltage compensation strategy calculates the minimum compensation voltage value for each backup electronic area based on line impedance parameters, and adjusts the output mode of the power distribution terminal by a certain multiple when the output voltage fluctuation in the main energy storage area exceeds the preset tolerance, effectively compensating for voltage fluctuations and ensuring that the power supply voltage of each sub-area remains stable within a reasonable range, guaranteeing the normal operation of the data center equipment. The harmonic suppression strategy, by collecting current harmonic distortion rate and adjusting pulse width modulation parameters, can promptly suppress harmonic interference, improve power quality, reduce harmonic damage to equipment, and extend equipment lifespan.

[0020] The mode adjustment module adaptively adjusts the output mode of the power conversion device in the backup power dispatch sub-region based on dynamic power distribution parameters, taking different adjustment measures for different thermal imbalance risk levels and harmonic issues. When the thermal imbalance risk level is Level 1, it switches to redundant power supply mode, shuts down the power transmission path from the main energy storage to the sub-region, enables the cross-power supply link between adjacent sub-regions and limits the load capacity, thus isolating the faulty area while ensuring the basic power supply needs of the sub-region, preventing the operation of the entire system from being affected by local faults. When the thermal imbalance risk level is Level 2, by constructing a power supply reliability model, marking the fault switching priority of each sub-region, and dynamically adjusting the output phase according to the priority, the fault switching process is optimized, improving the system's response speed and power supply reliability under complex fault conditions. For harmonic issues, by triggering harmonic isolation commands to cut off the corresponding output circuits and absorb residual harmonic energy, the impact of harmonics on the system is effectively eliminated, ensuring the stable operation of the power supply system.

[0021] Furthermore, the power supply reliability model, by statistically analyzing the number and duration of voltage sags and examining the correlation coefficient between load mutation rate and voltage phase deviation, can comprehensively assess the power supply reliability of each sub-region, providing a scientific basis for the dynamic adjustment and optimization of the system. Generating phase compensation parameters based on historical data and adjusting the output frequency further improves the system's adaptability to load changes and its power supply stability. Attached Figure Description

[0022] Figure 1 This is a schematic diagram illustrating the working principle of the outdoor flywheel UPS power supply system based on a data center as described in this invention. Figure 2 Workflow diagram for flywheel vibration period calibration strategy; Figure 3 A flowchart illustrating the implementation of a dynamic voltage compensation strategy; Figure 4This is a flowchart of the harmonic suppression strategy and pulse width modulation adjustment. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figures 1-4 The present invention relates to an outdoor flywheel UPS power supply system for data centers, specifically comprising the following components and implementation methods: The system includes a zone division module to divide the data center's power supply area into a primary energy storage verification sub-area and several backup power dispatch sub-areas. The primary energy storage verification sub-area serves as the core power supply unit, responsible for centrally deploying energy storage devices and performing power verification; the backup power dispatch sub-areas serve as distributed power supply units, responsible for distributing power according to the dispatch instructions from the primary area.

[0025] The flywheel monitoring module deploys at least one flywheel energy storage device within the main energy storage verification sub-area. It collects real-time rotational speed and temperature data for each flywheel device via sensors. The collection frequency is determined according to a preset sampling interval. The data is transmitted to the system backend for processing and analysis. Based on the flywheel status monitoring strategy, the operational stability of the flywheel device is determined, providing a basis for system scheduling.

[0026] The parameter generation module acquires power transmission path data between the main energy storage verification sub-region and the backup power dispatch sub-region, including information such as line length, impedance, and topology. Simultaneously, it combines this data with real-time load fluctuation data, imports power conversion control strategies, and generates dynamic power allocation parameters. These dynamic power allocation parameters include power transmission thresholds and voltage adjustment coefficients for each sub-region, guiding the power output of the backup power dispatch sub-region.

[0027] The mode adjustment module adaptively adjusts the output modes of all power conversion devices in the backup power dispatch sub-region based on the dynamic power distribution parameters output by the parameter generation module. Adjustment methods include, but are not limited to, voltage compensation, frequency regulation, and power supply mode switching, to achieve dynamic balance and stable operation of the data center power supply system.

[0028] Example 1: The system deploys power monitoring terminals at the boundaries of the main energy storage verification sub-area. These terminals integrate multi-parameter sensors for voltage, current, frequency, and other parameters, enabling them to collect power signals output from the main area in real time and transmit them to the system control center. Each backup power dispatch sub-area has an access node equipped with a power distribution terminal. This terminal has power receiving, conversion, and distribution functions, and is connected to the system control center via a communication link to receive and execute dynamic power distribution parameter commands.

[0029] The flywheel vibration period calibration strategy in the flywheel condition monitoring strategy is implemented as follows: When the power monitoring terminal detects the flywheel device start signal, the system starts the data acquisition program according to a preset sampling interval (e.g., once per second) to acquire the speed fluctuation data of the flywheel device within the first 10 sampling intervals, forming a speed fluctuation sequence containing 10 sets of data. Each set of speed fluctuation data includes the speed difference between adjacent sampling intervals and the speed change acceleration. The speed difference is obtained by subtracting the speed values ​​at adjacent times, as shown in the example. The rotational speed of each sampling interval is , No. The rotational speed of each sampling interval is The speed difference is:

[0030] The acceleration due to rotational speed change is obtained by subtracting the difference between adjacent rotational speeds and then dividing by the sampling interval, i.e.:

[0031] in This represents the sampling interval.

[0032] The system performs real-time analysis of the speed fluctuation sequence, calculating the sum of the speed differences over three consecutive sampling intervals as a speed anomaly index. For example, for the first... , , Each sampling interval, the speed difference is respectively , , The corresponding abnormal speed index is:

[0033] The preset speed anomaly index threshold is set based on engineering experience using parameters such as the rated speed and allowable vibration amplitude of the flywheel assembly. When the speed anomaly index at a certain sampling interval... When the threshold is exceeded, the system determines that there is an abnormal fluctuation in rotational speed within that sampling interval. It then extracts two sampling intervals before and after the abnormal interval, forming a continuous five sampling intervals including the abnormal interval (i.e., ...). , , , , ), which serves as the period for characterizing flywheel vibration.

[0034] In determining the flywheel vibration characterization period, the system uses a sliding window algorithm to iterate through the speed fluctuation sequence, ensuring that every possible abnormal interval is detected. The duration of the vibration characterization period is determined by the sampling interval; for example, when the sampling interval is 1 second, each characterization period corresponds to a 5-second time period. The speed data within this period is used to further analyze the vibration characteristics of the flywheel device, such as vibration frequency and amplitude variation trends, providing crucial information for judging the flywheel's operational stability.

[0035] The power distribution terminal and power monitoring terminal are connected to the system control center via wired or wireless communication, and the communication protocol supports real-time data transmission and command interaction. The deployment location of the power monitoring terminal must meet the requirement of comprehensive monitoring of the output power of the main energy storage verification sub-area, and it is usually set at the entrance of the power transmission line between the main area and the backup power dispatching sub-area; the power distribution terminal is installed at the power input node of each backup power dispatching sub-area to realize real-time adjustment and distribution control of the input power.

[0036] The flywheel vibration period calibration strategy is implemented through system software. The program employs multi-threading technology to ensure real-time data processing and anomaly detection while collecting flywheel speed data. Upon detecting a vibration characterization period, the system automatically triggers the flywheel condition assessment process, further analyzing the flywheel device's operating status in conjunction with temperature data.

[0037] Throughout the implementation process, the system uses standardized data acquisition procedures, preset algorithm logic, and hardware deployment architecture to achieve quantitative monitoring and periodic identification of the vibration state of the flywheel device, providing basic data support for subsequent temperature balance determination and power dispatching strategies, and ensuring the stability and reliability of the data center power supply system.

[0038] Example 2: The temperature equalization judgment strategy in the flywheel condition monitoring strategy is implemented after the flywheel vibration characterization period is determined. By integrating speed fluctuation data and temperature data, a quantitative assessment of the thermal stability of the flywheel device is achieved. The specific process is as follows: After acquiring the flywheel vibration characterization period (e.g., five consecutive sampling intervals), the system simultaneously extracts the temperature data within that period. The temperature data is collected from key components of the flywheel assembly, including temperature sensors deployed at locations such as bearings, stator windings, and rotor core. Sensor types can include thermocouples, resistance temperature detectors (RTDs), or infrared thermometers. The sampling frequency is consistent with the rotational speed data acquisition frequency (e.g., once per second), forming a temperature sequence aligned with the time axis of the rotational speed fluctuation sequence. .

[0039] The rate of temperature rise is calculated using the temperature difference between adjacent sampling intervals, i.e.:

[0040] The unit is ℃ / s. Simultaneously, rotational speed fluctuation data within the vibration characterization period are extracted, including the rotational speed difference at each sampling interval. ( and acceleration due to change in rotational speed ( This creates a two-dimensional dataset that includes speed fluctuations and temperature changes.

[0041] The system calculates the correlation coefficient between the rate of temperature rise and the rotational speed fluctuation data using statistical methods. Specifically, the correlation coefficient is calculated based on the rate of temperature rise. As the dependent variable, the corresponding speed difference is used. (or acceleration due to change in rotational speed) Using as the independent variable, the linear correlation between the two variables is calculated using either the Pearson correlation coefficient or the Spearman rank correlation coefficient algorithm. The range of the correlation coefficient is... The larger the absolute value, the stronger the correlation.

[0042] The system retrieves the temperature equalization threshold range for all flywheel units within the same vibration characterization period from historical operating data. This threshold range is determined through statistical analysis of historical data, including the maximum and minimum values ​​of the correlation coefficient under normal operating conditions. For example, the mean of the correlation coefficient is calculated by collecting 1000 sets of historical data. The standard deviation is The threshold range can then be set to (The specific range is adjusted according to the operating characteristics of the equipment).

[0043] When the calculated correlation coefficient exceeds the threshold range, the system determines that the flywheel device has a risk of thermal imbalance. Potential causes of thermal imbalance risk include insufficient bearing lubrication, poor winding heat dissipation, and rotor dynamic balance failure, leading to an abnormal coupling relationship between temperature changes and speed fluctuations. At this time, the system triggers a pre-start command for the backup power dispatch sub-region. This command is transmitted to the power distribution terminal of the target sub-region through the communication network, initiating the pre-charging process of the backup power supply link. Simultaneously, it activates the cross-power supply switching logic of adjacent sub-regions to ensure rapid switching to backup power in the event of a failure of the main energy storage device.

[0044] The deployment location of temperature sensors must meet the requirements for effective monitoring of heat-generating parts of the flywheel assembly. They are typically densely arranged in areas prone to heat generation, such as bearing housings, stator slots, and rotor end rings, to ensure the accuracy and representativeness of temperature data. Historical operating data is stored in the system database, categorized by flywheel assembly number, vibration cycle type, and operating period, facilitating rapid retrieval and comparison.

[0045] The calculation of correlation coefficients and threshold comparison are achieved through the system's built-in data analysis module. This module employs real-time data stream processing technology to ensure that calculations and judgments are completed within one sampling interval after the vibration characterization period is determined. The execution process of the pre-start command includes steps such as self-testing of the backup power inverter, status confirmation of the static switch, and pre-synchronization detection of the load-side voltage, ensuring that the backup power supply link is ready to be activated at any time.

[0046] The entire temperature balance determination strategy achieves early warning of the thermal state of the flywheel device through multi-source data fusion, historical data comparison and real-time algorithm analysis. Combined with the pre-start mechanism of the backup power scheduling sub-area, it improves the response capability of the data center power supply system to abnormalities of energy storage equipment and ensures the continuity and stability of power supply.

[0047] Example 3: The dynamic voltage compensation strategy and harmonic suppression strategy in the power conversion control strategy achieve dynamic adjustment of the supply voltage of the standby power dispatch sub-region and effective control of harmonic pollution through real-time analysis of power transmission path parameters and output power quality. The specific implementation methods are as follows: ① Implementation of dynamic voltage compensation strategy: The system first acquires power transmission path data between the main energy storage verification sub-area and the backup power dispatch sub-area, including parameters such as line length, conductor cross-sectional area, and material. Based on Ohm's law, it calculates the line impedance parameters (resistance). With reactance For each backup power dispatch sub-region, based on its current load current. (Data collected in real time by the power distribution terminal), calculate the line voltage drop:

[0048] This determines the minimum compensation voltage value for this sub-region:

[0049] ( To allow for a safety margin, it is usually set at 1.1-1.3.

[0050] The system monitors the output voltage fluctuation of the main energy storage verification sub-area in real time through a power monitoring terminal, specifically the percentage deviation between the measured voltage value and the rated voltage value. When the fluctuation exceeds a preset tolerance (e.g., ±5%), a dynamic voltage compensation mechanism is triggered. At this time, the system compensates for the minimum voltage value. times ( A voltage adjustment command is generated (as a preset constant, such as 1.2) and sent to the power distribution terminal of the corresponding backup power dispatch sub-region.

[0051] Upon receiving an instruction, the power distribution terminal adjusts the output voltage by regulating the transformation ratio of power conversion devices (such as UPS inverters) or controlling the number of reactive power compensation devices (such as capacitor banks and reactors) in operation. For example, when a voltage drop in the main area causes the input voltage in the sub-area to fall below the rated value, the power distribution terminal increases the output voltage to the rated value. It compensates for the combined effects of line voltage drop and main area voltage fluctuation; conversely, when the main area voltage is too high, it reduces the output voltage to a safe range to avoid damage to the load equipment due to overvoltage.

[0052] ② Implementation of harmonic suppression strategies: The system acquires the output current signal in real time through a current sensor installed at the output end of the power conversion device, and calculates the current harmonic distortion rate (THD) using the Fast Fourier Transform (FFT) algorithm. When the THD exceeds a preset threshold (e.g., 5%) for three consecutive sampling intervals (e.g., 3 cycles, assuming a sampling frequency of 50Hz, then the duration is 60ms), it is determined that there is a risk of harmonic pollution and a harmonic suppression command is generated.

[0053] Upon triggering the harmonic suppression command, the system adjusts the pulse width modulation (PWM) parameters of the power conversion device, including carrier frequency, modulation ratio, and pulse phase. For example, increasing the carrier frequency raises the harmonic frequency, making it easier for the output filter to suppress; or adjusting the modulation ratio optimizes the output voltage waveform, reducing low-order harmonic content. Simultaneously, the system monitors the THD value in real time. If the THD does not drop below the threshold after adjustment, the PWM parameters are further adjusted until the harmonic distortion rate meets the requirements.

[0054] In terms of hardware implementation, the power conversion device uses fully controlled power electronic devices (such as IGBTs) to support real-time dynamic adjustment of PWM parameters. An LC filter is configured at the output, with parameters (inductance...)... ,capacitance Designed based on common harmonic frequencies (such as the 5th and 7th harmonics), it can effectively attenuate harmonics of specific frequencies. The current sensor uses a high-precision Hall element to ensure the accuracy of harmonic component acquisition.

[0055] In terms of data interaction, relevant parameters of dynamic voltage compensation and harmonic suppression strategies (such as line impedance, compensation multiple, THD threshold, etc.) are stored in the system database and can be configured and modified through the human-machine interface. The power monitoring terminal and the power distribution terminal are connected via industrial Ethernet or fieldbus (such as Modbus, CAN) to ensure the real-time performance and reliability of command transmission.

[0056] Throughout the implementation process, the dynamic voltage compensation strategy achieves precise adjustment of the power supply voltage of each sub-region through a closed-loop control process of "parameter calculation - real-time monitoring - command adjustment"; the harmonic suppression strategy forms a feedback control loop through real-time harmonic detection and PWM parameter optimization, effectively suppressing harmonic components in the output current. Together, they ensure the power quality and operational stability of the data center power supply system.

[0057] Example 4: The adaptive adjustment function of the mode adjustment module implements graded control based on the level of thermal imbalance risk. Through power transmission path switching, power supply mode reconstruction, and load capacity management, it achieves dynamic fault tolerance and stable operation of the data center power supply system. The specific implementation method is as follows: ① Adjustment strategy when the risk level of thermal imbalance is Level 1: When the system determines that a certain flywheel device has a level-one thermal imbalance risk through a temperature equalization judgment strategy, it first extracts the backup power dispatch sub-region number (such as sub-region A) corresponding to the flywheel device. The system sends a mode switching command to the power distribution terminal of sub-region A through the communication link, and at the same time triggers the cross-power supply link preparation process of adjacent backup power dispatch sub-regions (such as sub-region B and sub-region C).

[0058] The first step is to shut down the power transmission path from the main energy storage verification sub-area to sub-area A. This involves disconnecting the circuit breaker or solid-state switch between the main area and sub-area A, cutting off the main power supply path, and preventing the risk of thermal imbalance from spreading to the main area. At this point, the power input to sub-area A is switched to the backup path.

[0059] The second step is to activate the cross-power supply link between adjacent backup power dispatch sub-regions. The system controls the closure of the interconnection switch between sub-region A and sub-regions B and C, establishing a cross-power supply channel. The cross-power supply link adopts a pre-designed ring topology or mesh topology, and the link capacity is designed based on 80% of the rated load capacity of sub-region A to ensure that adjacent sub-regions have support capabilities during normal operation.

[0060] The third step is to limit the maximum load capacity of sub-area A to 80% of its rated value. The system scans the load devices within sub-area A through the load management module and automatically cuts off low-priority loads based on preset load priorities (e.g., critical business loads have higher priority than non-critical loads), keeping the total load power below 80% of its rated value. Load shedding commands are executed via intelligent PDUs (Power Distribution Units) or remote control switches to ensure that critical business operations are not affected during power limiting.

[0061] ② Adjustment strategy when the thermal imbalance risk level is level two: When the thermal imbalance risk level rises to level two, the system retrieves historical power supply stability data from the power distribution terminal in the backup power dispatch sub-region. This data includes the number of voltage drops in each sub-region over the past 24 hours, the duration of each drop, and records of load mutation events. Based on this data, the system constructs a power supply reliability model using a reliability analysis algorithm.

[0062] The power supply reliability model employs a multi-factor evaluation method, using parameters such as the number of voltage sags, duration, and load mutation rate as evaluation indicators. For example, if a sub-region experiences more than three voltage sags in the past 24 hours, with each sag lasting more than 10 seconds, the sub-region is deemed to have poor power supply stability, and its fault switching priority is raised to the highest level. The priority evaluation results are marked in the form of numerical codes (e.g., priority 1 is the highest, and priority 5 is the lowest) and stored in the system database.

[0063] The system dynamically adjusts the output phase of the power conversion device based on fault switching priority. Specifically, it collects voltage phase data for each sub-region in real time using a phase detection module and calculates the phase difference between the target sub-region and its adjacent sub-regions. When power switching is required, the highest-priority sub-region with the smallest phase difference from the target sub-region is selected as the backup power source. The output frequency is then gradually adjusted using the power conversion device's frequency regulation function to synchronize the voltage phase of the target sub-region with the backup power source phase, thus preventing inrush current during switching.

[0064] In terms of hardware deployment, cross-power supply links are connected using insulated cables or busbars, and tie switches are selected from circuit breakers or contactors with fast opening and closing capabilities. The load management module and intelligent PDU are connected via RS-485 or IP network, supporting remote load status query and control. The power conversion device is equipped with a high-precision phase detection chip and a digital signal processor (DSP), which can calculate the phase difference in real time and generate adjustment commands.

[0065] In the data processing flow, information such as the determination result of thermal imbalance risk level, sub-region number, and historical power supply stability data are transmitted to the mode adjustment module in real time via a message queue to ensure rapid response to control commands. The fault switching priority is updated every 1 hour to ensure that the model can reflect the latest system operating status.

[0066] The entire adaptive adjustment process uses risk level identification, historical data-driven priority assessment, and phase synchronization control technology to achieve differentiated handling of thermal imbalance risks of varying severity. While ensuring power supply continuity, it minimizes the impact of faults and improves the fault tolerance and reliability of the data center power supply system.

[0067] Example 5: The adaptive adjustment function of the mode adjustment module also includes harmonic mitigation and power supply reliability optimization. Through real-time harmonic monitoring, loop isolation, and phase compensation technologies, it achieves dynamic control of power quality and improves power supply stability. The specific implementation method is as follows: ① Implementation of harmonic mitigation: The system acquires the output current signal in real time through current acquisition modules installed at the output terminals of each power conversion device. After the acquisition module converts the analog signal into a digital signal, it transmits it to the system's harmonic analysis unit. The harmonic analysis unit uses a spectrum analysis algorithm to identify the harmonic components in the current and calculates the amplitude ratio of the 2nd to 13th harmonics (i.e., the ratio of the amplitude of each harmonic to the amplitude of the fundamental wave).

[0068] When the amplitude of a harmonic exceeds 5%, the system determines that the harmonic component may interfere with the power supply system and load equipment, triggering a harmonic isolation command. The harmonic isolation command is transmitted to the controller of the corresponding power conversion device via the communication bus. The controller performs the following operations: First, it sends a trip signal to the circuit breaker at the output of the device to disconnect the output circuit and prevent further harmonic propagation; second, it sends a switching command to the parallel capacitor bank in the adjacent sub-region to activate the capacitor bank's absorption function for residual harmonic energy.

[0069] Parallel capacitor banks are pre-configured at the power input nodes of each backup power dispatch sub-region. Their capacitance values ​​are designed based on common harmonic frequencies (e.g., 250Hz for the 5th harmonic and 350Hz for the 7th harmonic). Through the LC resonance principle, they can create a low-impedance path for harmonics of specific frequencies, guiding harmonic currents into the capacitor bank rather than the load side, thereby reducing the harmonic content in the system. The switching of the capacitor banks is controlled by intelligent switches with a switching response time of less than 10ms, ensuring rapid initiation of the harmonic absorption process after a faulty circuit is disconnected.

[0070] ② Extended applications of the power supply reliability model: The system collects the load mutation rate of each sub-region in real time. The load mutation rate is calculated as the ratio of the change in load power per unit time to the rated power (the specific calculation process does not involve formulas, but is only evaluated through data change trends). At the same time, the system collects voltage phase deviation data of each sub-region. The phase deviation is obtained by comparing it with the reference phase of the main region through a real-time phase detection module.

[0071] The system uses a correlation analysis algorithm to calculate the correlation coefficient between load mutation rate and voltage phase deviation. The range of the correlation coefficient reflects the degree of correlation between the two. When the correlation coefficient exceeds a preset threshold, it indicates that the load mutation has a significant impact on the voltage phase, potentially leading to power supply instability. At this point, the system retrieves records from the historical fault database that are similar to the current load mutation pattern and generates phase compensation parameters. The phase compensation parameters include frequency adjustment and phase adjustment angle, used to guide the output adjustment of the power conversion device.

[0072] After receiving the phase compensation parameters, the power conversion unit adjusts the output frequency through its built-in frequency control module, gradually locking the voltage phase to the target compensation value. The frequency adjustment process employs a PID control algorithm to ensure a smooth adjustment and avoid system oscillations caused by sudden frequency changes. Simultaneously, the system continuously monitors the phase deviation and load change rate until the correlation coefficient drops below the threshold, at which point normal operation is restored.

[0073] ③ Hardware and data interaction: The current acquisition module uses a high-precision Hall current sensor, supporting synchronous acquisition of harmonics from 0 to 50th order. The harmonic analysis unit integrates a dedicated digital signal processing chip, capable of calculating harmonic components within 20ms. The parallel capacitor bank and intelligent switch are integrated into a single design, providing overcurrent protection and switching frequency statistics. The historical fault database is categorized and stored according to sub-region number, load change type, phase deviation amplitude, and other dimensions, supporting fuzzy search and fast retrieval.

[0074] The phase detection module, based on phase-locked loop (PLL) technology, tracks voltage phase changes in real time with a detection accuracy of 0.1 degrees. The frequency control module of the power conversion unit supports frequency adjustment resolution at the 0.01Hz level, ensuring the accuracy of phase compensation. Load change data and phase deviation data for each sub-region are stored and synchronized through a real-time database with an update cycle of 50ms, meeting the timeliness requirements of dynamic control.

[0075] Throughout the implementation process, harmonic mitigation achieves rapid response to harmonic pollution through a closed-loop process of "real-time monitoring - threshold determination - loop isolation - harmonic absorption"; power supply reliability optimization improves the system's adaptability to dynamic loads by analyzing the correlation between load mutations and phase deviations, combined with historical data-driven compensation strategies. Together, these two aspects ensure the power quality and operational stability of the data center power supply system.

[0076] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0077] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An outdoor flywheel UPS power supply system for data centers, characterized in that, The system includes: The area division module is used to divide the power supply area of ​​the data center into a main energy storage verification sub-area and several backup power dispatch sub-areas. The flywheel monitoring module is used to deploy at least one flywheel energy storage device in the main energy storage verification sub-area, collect real-time speed and temperature data of each flywheel device, and determine the operational stability of the flywheel device based on the flywheel status monitoring strategy. The parameter generation module is used to obtain power transmission path data between the main energy storage verification sub-area and the backup power dispatch sub-area, and combine it with real-time load fluctuation data to import the power conversion control strategy and generate dynamic power allocation parameters. The mode adjustment module is used to adaptively adjust the output mode of all power conversion devices in the standby power dispatch sub-region according to dynamic power distribution parameters.

2. The outdoor flywheel UPS power supply system for data centers according to claim 1, characterized in that, A power monitoring terminal is installed at the boundary of the main energy storage verification sub-region, and a power distribution terminal is installed at the access node of each backup power dispatch sub-region. The flywheel condition monitoring strategy includes: a flywheel vibration period calibration strategy; The flywheel vibration period calibration strategy is as follows: When the power monitoring terminal detects that the flywheel device has started, it acquires the speed fluctuation data of the flywheel device in the first 10 sampling intervals according to the preset sampling interval, forming a speed fluctuation sequence. The speed fluctuation data includes: the speed difference between adjacent sampling intervals and the speed change acceleration. The sum of the speed difference between three consecutive sampling intervals in the speed fluctuation sequence is calculated as the speed anomaly index. Sampling intervals with speed anomaly indices exceeding the threshold are selected, and two sampling intervals before and after the selected interval are extracted to form the flywheel vibration characterization period.

3. The outdoor flywheel UPS power supply system based on a data center according to claim 2, characterized in that, The flywheel status monitoring strategy also includes: a temperature equalization determination strategy; The specific temperature equalization determination strategy is as follows: Extract temperature data within the flywheel vibration characterization period and calculate the correlation coefficient between the temperature rise rate and the rotational speed fluctuation data; Retrieve the temperature equalization threshold range of all flywheel devices within the same vibration cycle from historical operating data; If the correlation coefficient between the current rate of temperature rise and the speed fluctuation data exceeds the threshold range, it is determined that the flywheel device has a risk of thermal imbalance, triggering the pre-start command of the backup power dispatch sub-region.

4. The outdoor flywheel UPS power supply system for data centers according to claim 1, characterized in that, The power conversion control strategy includes: a dynamic voltage compensation strategy; The dynamic voltage compensation strategy is as follows: Calculate the minimum compensation voltage value for each backup power dispatch sub-region based on the line impedance parameters in the power transmission path data; The output voltage fluctuation of the main energy storage verification sub-area is monitored in real time. When the fluctuation exceeds the preset tolerance, the output mode of the power distribution terminal is adjusted to N times the minimum compensation voltage value, where N is a preset constant.

5. The outdoor flywheel UPS power supply system for data centers according to claim 4, characterized in that, The power conversion control strategy also includes: a harmonic suppression strategy; The harmonic suppression strategy is specifically as follows: The harmonic distortion rate of the current at the output of the power conversion device is collected. If the distortion rate of three consecutive sampling intervals exceeds the threshold, a harmonic suppression command is generated. The pulse width modulation parameters of the power conversion device are adjusted according to the harmonic suppression command until the distortion rate drops below the threshold.

6. The outdoor flywheel UPS power supply system for data centers according to claim 1, characterized in that, The adaptive adjustment includes: Extract the backup power dispatch sub-region number corresponding to the flywheel device with thermal imbalance risk; When the thermal imbalance risk level is Level 1, the power conversion device of the corresponding backup power dispatch sub-area will be switched to redundant power supply mode, wherein the redundant power supply mode includes: Close the power transmission path from the main energy storage verification sub-area to that sub-area; Enable the cross-power supply link of adjacent backup power dispatch sub-regions and limit the maximum load capacity of the sub-region to 80% of the rated value.

7. The outdoor flywheel UPS power supply system for data centers according to claim 6, characterized in that, The adaptive adjustment also includes: When the thermal imbalance risk level is level two, historical power supply stability data is retrieved from the power distribution terminal of the backup power dispatch sub-region. A power supply reliability model is constructed based on historical power supply stability data, and the fault switching priority of each sub-region is marked. The output phase of the power conversion device is dynamically adjusted according to the fault switching priority.

8. The outdoor flywheel UPS power supply system for data centers according to claim 7, characterized in that, The adaptive adjustment also includes: Real-time acquisition of harmonic components of the output current of each power conversion device; Calculate the amplitude proportion of the 2nd to 13th harmonics in the harmonic components. If the amplitude proportion of a certain harmonic exceeds 5%, a harmonic isolation command is triggered. The output circuit of the corresponding power conversion device is cut off according to the harmonic isolation command, and the residual harmonic energy is absorbed by the parallel capacitor bank of the adjacent sub-region.

9. The outdoor flywheel UPS power supply system based on a data center according to claim 7, characterized in that, The power supply reliability model includes: Count the number of voltage drops and their duration in each sub-region over the past 24 hours; If a sub-region experiences more than 3 voltage drops with each drop lasting more than 10 seconds, the fault switching priority for that sub-region will be raised to the highest level.

10. The outdoor flywheel UPS power supply system based on a data center according to claim 9, characterized in that, The power supply reliability model also includes: The load mutation rate of each sub-region is collected in real time, and the correlation coefficient between the mutation rate and the voltage phase deviation is calculated. If the correlation coefficient exceeds the preset threshold, phase compensation parameters are generated based on historical fault data, and the output frequency of the power conversion device is adjusted to the target phase compensation value simultaneously.

Images