Cooperative inertia response control apparatus and method

Abstract

The application discloses a kind of collaborative inertia response control device and method, belong to inertia response control field, the first end of collaborative inertia response control device is connected with the first end of current transformer voltage transformer, the second end of collaborative inertia response control device is connected with the second end of data room, the third end of collaborative inertia response control device is connected with the third end of data room, the second end of collaborative inertia response control device is connected with the second end of energy storage system;Collaborative inertia response control device includes data center measurement module, virtual rotor control module, computing load power consumption scheduling module, energy storage power scheduling module.The application can inhibit power grid frequency change rate, restore steady-state frequency deviation to the allowable range, enhance the stability and anti-disturbance ability of high proportion renewable energy power grid, increase power grid inertia supply source, realize the optimal allocation of power grid resources.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of inertia response control, and particularly relates to a coordinated inertia response control device and method. BACKGROUND

[0002] With the deepening of the "double carbon" strategy, the proportion of new energy power generation represented by wind power and photovoltaic power is increasing, which is accelerating the replacement of traditional synchronous generators. However, new energy units with power electronic converter interfaces cannot provide natural rotational inertia support for the power grid, resulting in a continuous decrease in the equivalent inertia time constant of the new power system, and the frequency stability problem is increasingly prominent. The lack of system inertia makes the rate of change of frequency (ROCOF) increase and the frequency extreme deviation increase when the power grid is subjected to power disturbance, which seriously threatens the safe and stable operation of the power grid. To address this challenge, the power grid side increases the rotational inertia auxiliary service market through policy guidance on the one hand, and actively promotes the virtual synchronous machine (VSG) technology on the other hand, which makes the converter simulate the external characteristics of the synchronous generator through control algorithm.

[0003] However, existing virtual synchronous machine technology relies on power generation side resources such as energy storage, photovoltaic power and wind power to simulate the operation behavior of synchronous generators by occupying valuable power generation capacity, resulting in high initial investment cost and valuable power generation capacity being consumed in regulation function, and failing to achieve optimal allocation of resources. Moreover, the regulation potential of data centers as key infrastructure for digital economy has not been fully tapped. The existing utilization mainly focuses on peak shaving services, and fails to combine them with power grid inertia support, resulting in waste of high-quality flexible resources.

[0004] Therefore, how to utilize the massive flexible load resource of data centers to enable them to provide rapid inertia support for the power grid like power generation resources has become a technical problem to be solved. SUMMARY

[0005] The present application aims to overcome the defects of the prior art and provide a coordinated inertia response control device and method to tap the rapid power regulation potential of data centers, control the coordinated action of data center computing load and energy storage system to realize the complementarity of their outputs, and simulate the rotor motion and speed regulator function of synchronous generators to provide rapid, smooth and accurate inertia response function for the power grid.

[0006] In a first aspect, the application provides a coordinated inertia response control device, a first end of the coordinated inertia response control device being connected to a first end of a voltage transformer and a current transformer, a second end of the voltage transformer and the current transformer being connected to a grid point bus, a third end of the voltage transformer and the current transformer being connected to a first end of a data center room and a first end of an energy storage system, a second end of the coordinated inertia response control device being connected to a second end of the data center room, a third end of the coordinated inertia response control device being connected to a third end of the data center room, and a fourth end of the coordinated inertia response control device being connected to a second end of the energy storage system.

[0007] The coordinated inertia response control device comprises a data center measurement module, a virtual rotor control module, a computing load power consumption scheduling module, and an energy storage power scheduling module. A first end of the data center measurement module is connected to the first end of the voltage transformer and the current transformer. A second end of the data center measurement module is connected to the second end of the data center room. A third end of the data center measurement module is connected to a first end of the virtual rotor control module and a first end of the energy storage power scheduling module. A fourth end of the data center measurement module is connected to a second end of the energy storage power scheduling module. A second end of the virtual rotor control module is connected to a first end of the computing load power consumption scheduling module. A second end of the computing load power consumption scheduling module is connected to the third end of the data center room. A third end of the energy storage power scheduling module is connected to a second end of the virtual rotor control module. A fourth end of the energy storage power scheduling module is connected to a second end of the virtual rotor control module. A fifth end of the energy storage power scheduling module is connected to the second end of the energy storage system.

[0008] Optionally, the data center measurement module comprises a data center and supporting energy storage grid point measurement unit and a data center access end measurement unit. The data center and supporting energy storage grid point measurement unit and the data center access end measurement unit are arranged in parallel and are respectively used to measure the frequency change rate of the grid point bus and the power change amount of the data center room.

[0009] Optionally, the transfer function of the virtual rotor control module is and is expressed as:

[0010]

[0011] wherein, is the frequency change rate of the grid point bus, is the compensation power, is the virtual inertia time constant of the virtual synchronous machine, is the rated frequency of the virtual synchronous machine, is the damping coefficient of the virtual synchronous machine, is the Laplace operator.

[0012] Optionally, the computing load power consumption scheduling module comprises a computing load power consumption mapping model and a computing load scheduling model, a first end of the computing load power consumption mapping model is connected to a second end of the virtual rotor control module, a second end of the computing load power consumption mapping model is connected to a first end of the computing load scheduling model, and a second end of the computing load scheduling model is connected to a third end of the data center.

[0013] Optionally, the energy storage power scheduling module comprises an energy storage primary frequency modulation unit, an energy storage auxiliary compensation unit, and a fusion unit, a first end of the energy storage primary frequency modulation unit is connected to a second end of the data center and the supporting energy storage grid-connected point measurement unit, a second end of the energy storage primary frequency modulation unit is connected to a first end of the fusion unit, a first end of the energy storage auxiliary compensation unit is connected to a second end of the virtual rotor control module, a second end of the energy storage auxiliary compensation unit is connected to a second end of the virtual rotor control module, a third end of the energy storage auxiliary compensation unit is connected to a second end of the data center access end measurement unit, a fourth end of the energy storage auxiliary compensation unit is connected to a second end of the fusion unit, and a third end of the fusion unit is connected to a second end of the energy storage system.

[0014] Optionally, the transfer function of the energy storage primary frequency modulation unit is:

[0015]

[0016] wherein, represents a primary frequency modulation coefficient, is a Laplace operator.

[0017] In a second aspect, the application also provides a coordinated inertia response control method applied to the coordinated inertia response control device in any one of the first aspect, comprising the following steps:

[0018] The data center and the supporting energy storage grid-connected point measurement unit measure the grid-connected point bus to obtain the frequency change rate of the grid-connected point bus;

[0019] The data center access end measurement unit measures the data center access end power change to obtain the power change of the data center;

[0020] The virtual rotor control module receives the frequency change rate of the grid-connected point bus, simulates the rotor motion equation, and obtains the required virtual synchronous generator compensation power of the grid-connected point bus under disturbance through the virtual inertia power compensation control strategy;

[0021] The energy storage power scheduling module performs energy storage auxiliary power compensation and primary frequency modulation power compensation according to the frequency change rate of the grid-connected point bus and the power change of the data center to obtain an energy storage converter scheduling instruction.​

[0022] The computing load power scheduling module performs task scheduling according to the compensation power required when the grid-connected point bus is subjected to disturbance, obtains data room scheduling instructions, and adjusts the data room and the energy storage system according to the energy storage converter scheduling instructions and the data room scheduling instructions, to complete the collaborative inertia response.

[0023] Optionally, the virtual rotor control module receives the frequency change rate of the grid-connected point bus, simulates a rotor motion equation, and obtains the virtual synchronous generator compensation power required when the grid-connected point bus is subjected to disturbance through a virtual inertia power compensation control strategy, including:

[0024] determining a rotor motion equation of the virtual synchronous generator;

[0025] based on the rotor motion equation of the virtual synchronous generator, obtaining a power deficiency per unit of the virtual synchronous generator through the relationship between frequency and angular frequency and the relationship between virtual rotational inertia and virtual inertia time constant;

[0026] obtaining the virtual synchronous generator compensation power required when the grid-connected point bus is subjected to disturbance through Laplace transform.

[0027] Optionally, the computing load power scheduling module performs task scheduling according to the compensation power required when the grid-connected point bus is subjected to disturbance, obtains data room scheduling instructions, and adjusts the data room and the energy storage system according to the energy storage converter scheduling instructions and the data room scheduling instructions, to complete the collaborative inertia response, including:

[0028] The computing load power mapping model in the computing load power scheduling module obtains the compensation power output by the virtual rotor control module, and calculates the computing power consumption of the IT equipment;

[0029] The computing load scheduling model in the computing load power scheduling module determines a scheduling objective function and computing task constraints by using the delay tolerance and time flexibility of the task, to obtain data room scheduling instructions;

[0030] The data room and the energy storage system are adjusted according to the energy storage converter scheduling instructions and the data room scheduling instructions, to complete the collaborative inertia response.

[0031] Optionally, the scheduling objective function is:

[0032]

[0033] wherein, is a delay-tolerant computing task of the i th server at time t, is a computing task in the delay of the i th server at time t, P(t-1) is the computing power consumption of the server cluster at time t-1, Pcomp is the compensation power output by the virtual rotor control module, P0(i) is the idle power consumption of the i-th server, Pmax(i) is the maximum power consumption of the i-th server, T(t,i) is the instant computing task of the i-th server at time t, T(t,i) is the instant computing task of the i-th server at time t, Tmax(i) is the maximum computing task of the i-th server, I is the total number of servers in the IT equipment.

[0034] The application provides a collaborative inertia response control device and method. By constructing a virtual rotor control module and a computing load power consumption scheduling module, the device can participate in frequency regulation and power regulation of the power grid, and the data center is transformed from a pure energy consumer into a regulator supporting the stability of the power grid, so as to deeply tap the potential of rapid power regulation of the data center. Through collaborative control of the energy storage power scheduling module and the computing load power consumption scheduling module, the negative influence of other equipment in the data center on the load response of the data center can be overcome, the IT equipment of the data center can be controlled, and the IT equipment and the energy storage system can be synergistically acted to realize complementation of outputs of the IT equipment and the energy storage system, and the rotor motion and the governor function of the synchronous generator are simulated, so as to provide rapid, smooth and accurate inertia response functions for the power grid. The rate of change of the frequency of the power grid is effectively inhibited, and the steady-state frequency deviation is restored to the allowable range, thereby enhancing the stability and anti-disturbance ability of the power grid with a high proportion of renewable energy, increasing the inertia supply source of the power grid, and being beneficial to the optimal allocation of resources of the power grid.

[0035] In order to make the above features and advantages of the application more obvious and easy to understand, the following specific examples are described in detail below, and the accompanying drawings are described as follows. BRIEF DESCRIPTION OF DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments or the related art, the following will briefly introduce the drawings needed to be used in the embodiments or the related art description. Obviously, the drawings in the following description only some embodiments of the application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0037] Figure 1 FIG. 1 is a structural schematic diagram of a collaborative inertia response control device provided in an embodiment of the application.

[0038] Figure 2 FIG. 2 is a specific structural schematic diagram of the collaborative inertia response control device provided in an embodiment of the application.

[0039] Figure 3A schematic diagram of scheduling logic of a load power consumption scheduling module in a coordinated inertia response control device provided in an embodiment of the present application.

[0040] Figure 4 A flowchart of a coordinated inertia response control method provided in an embodiment of the present application.

[0041] Figure 5 A flowchart of step S3 in a coordinated inertia response control method provided in an embodiment of the present application.

[0042] Figure 6 A flowchart of step S5 in a coordinated inertia response control method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0043] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the drawings of the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments of the present application. Based on the described embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort belong to the scope of protection of the present application.

[0044] In one embodiment, referring to Figure 1 The present application provides a coordinated inertia response control device 1, a first end of the coordinated inertia response control device 1 is connected with a first end of a potential transformer (PT) and a current transformer (CT) 2, a second end of the potential transformer and the current transformer 2 is connected with a grid point bus 5, a third end of the potential transformer and the current transformer 2 is connected with a first end of a data machine room 3 and a first end of an energy storage system 4, a second end of the coordinated inertia response control device 1 is connected with a second end of the data machine room 3, a third end of the coordinated inertia response control device 1 is connected with a third end of the data machine room 3, and a fourth end of the coordinated inertia response control device 1 is connected with a second end of the energy storage system 4.

[0045] As an example, the first end and the second end of the coordinated inertia response control device 1 can be signal acquisition interfaces.

[0046] As an example, the third end and the fourth end of the coordinated inertia response control device 1 can be control instruction interfaces.

[0047] As an example, the data room 3 includes an uninterruptible power supply (UPS) 31, IT equipment 32, a data center infrastructure management (DCIM) / battery management system (BMS) system 33, a load scheduling controller 34, a first end of the uninterruptible power supply 31 is connected to a third end of the voltage transformer and current transformer 2 and a first end of the energy storage system 4, a second end of the uninterruptible power supply 31 is connected to a first end of the IT equipment 32; a first end of the DCIM / BMS system 33 is connected to a second end of the collaborative inertia response control device 1, a second end of the DCIM / BMS system 33 is connected to a second end of the IT equipment 32; a first end of the load scheduling controller 34 is connected to a third end of the collaborative inertia response control device 1, a second end of the load scheduling controller 34 is connected to a third end of the IT equipment 32.

[0048] As an example, the DCIM / BMS system 33 can include a high-frequency intelligent power measurement unit (not shown), which can be arranged at the incoming line of the data room 3.

[0049] As an example, the energy storage system 4 includes a power conversion system (PCS) 41, an energy storage element 42, a first end of the power conversion system 41 is connected to a third end of the voltage transformer and current transformer 2, a second end of the power conversion system 41 is connected to a fourth end of the collaborative inertia response control device 1, a third end of the power conversion system 41 is connected to the energy storage element 42.

[0050] As an example, the collaborative inertia response control device 1, the voltage transformer and current transformer 2, the data room 3, the energy storage system 4, and the grid-connected point bus 5 are all located within the data center.

[0051] As an example, the collaborative inertia response control device 1 is connected to the grid-connected point bus 5 through the voltage transformer and current transformer 2, and real-time acquisition of the frequency change rate of the grid-connected point bus 5 ROCOF ( s ) is performed; the collaborative inertia response control device 1 communicates with the DCIM / BMS system 33 through a field bus to obtain the power change amount ; through the network communication or hardwired connection to the load scheduling controller 34, the inertia response control device 1 completes the calculation of load scheduling calculation through the network communication or hardwired to the load scheduling controller 34 sends data room scheduling instructions, by the load scheduling controller 34 control IT equipment 32 power regulation; through the high-speed communication link and power electronic drive circuit connected to the main controller of the energy storage converter 41, to the energy storage converter 41 sends energy storage converter scheduling instructions, energy storage converter 41 according to the energy storage converter scheduling instructions control energy storage element 42 to absorb or release active power.

[0052] As an example, the energy storage converter scheduling instructions can include real-time power set value or frequency-power regulation curve instructions.

[0053] As an example, please refer to Figure 2 , the inertia response control device 1 includes, data center measurement module 11, virtual rotor control module 12, computing load power consumption scheduling module 13, energy storage power scheduling module 14, the first end of the data center measurement module 11 is connected with the first end of the voltage transformer and current transformer 2, the second end of the data center measurement module 11 is connected with the second end of the data center 3, the third end of the data center measurement module 11 is connected with the first end of the virtual rotor control module 12 and the first end of the energy storage power scheduling module 14, the fourth end of the data center measurement module 11 is connected with the second end of the energy storage power scheduling module 14; the second end of the virtual rotor control module 12 is connected with the first end of the computing load power consumption scheduling module 13, the second end of the computing load power consumption scheduling module 13 is connected with the third end of the data center 3; the third end of the energy storage power scheduling module 14 is connected with the second end of the virtual rotor control module 12, the fourth end of the energy storage power scheduling module 14 is connected with the second end of the virtual rotor control module 12, the fifth end of the energy storage power scheduling module 14 is connected with the second end of the energy storage system 4.

[0054] As an example, please continue to refer to Figure 2 , the data center measurement module 11 includes data center and supporting energy storage grid connection point measurement unit 111, data center access end measurement unit 112, data center and supporting energy storage grid connection point measurement unit 111, data center access end measurement unit 112 are arranged in parallel, respectively for measuring the frequency change rate of the grid connection point bus 5 and the power change of the data center 3.

[0055] As an example, the data center and supporting energy storage grid connection point measurement unit 111 converts the high voltage on the grid connection point bus 5 into a low voltage signal proportionally through voltage transformers and current transformers 2 connected to the grid connection point bus 5. The secondary output cables of the voltage transformers and current transformers 2 are connected to the first terminal of the data center and supporting energy storage grid connection point measurement unit 111. The data center and supporting energy storage grid connection point measurement unit 111 discretizes the low voltage signal at a first sampling rate to obtain a digital sampling sequence. Then, the digital sampling sequence is sent to the processor of the data center and supporting energy storage grid connection point measurement unit 111, and the frequency change rate of the grid connection point bus 5 is measured by the zero-crossing detection method. ROCOF ( k ); the rate of change of frequency of bus 5 at the time-domain grid connection point. Performing a Laplace transform, we obtain the frequency change rate of bus 5 at the grid connection point in the s-domain. .

[0056] As an example, the first terminal of the measurement unit 111 for the data center and its supporting energy storage grid connection point can be a frequency measurement input terminal.

[0057] As an example, for frequency measurements, the first sampling rate does not need to be extremely high, but it must satisfy the Nyquist sampling theorem.

[0058] As an example, the first terminal of the data center access measurement unit 112 is connected to the high-frequency intelligent power measurement unit of the DCIM / BMS system 33 to detect the power change of the data center 3. The power variation of time-domain data center 3 Perform a Laplace transform to obtain the power change in data center 3 in the s-domain. .

[0059] As an example, the first terminal of the data center access measurement unit 112 can be a power measurement input terminal.

[0060] As an example, the virtual rotor control module 12 simulates the rotor motion equations of a virtual synchronous machine (VSG) and uses a virtual inertial power compensation control strategy to solve for the virtual synchronous motor compensation power required for the inertial response when the grid connection point bus 5 is disturbed. and will compensate power The data is transmitted to the computing load power scheduling module 13 and the energy storage power scheduling module 14.

[0061] Specifically, please refer to Figure 2 The virtual inertial power compensation control strategy is as follows: the virtual rotor control module 12 receives the frequency change rate of the grid connection point bus 5 output by the data center and the supporting energy storage grid connection point measurement unit 111. Virtual inertial time constant via virtual synchronizer The ratio of the rated frequency of the virtual synchronizer The proportional element is inertia compensated to obtain the first component; the frequency change rate of the grid connection point bus 5. First, go through the points process Accumulated frequency deviation, then damped by the virtual synchronizer. Damping compensation is performed to suppress subsequent oscillations in the grid frequency and prevent frequency fluctuations near a stable value, resulting in a second component. The first and second components are then fused to obtain the compensation power. and will compensate power The data is transmitted to the computing load power scheduling module 13 and the energy storage power scheduling module 14.

[0062] As an example, the transfer function of the virtual rotor control module 12 It can be represented as:

[0063]

[0064] in, The virtual inertial time constant of the virtual synchronizer. The rated frequency of the virtual synchronizer. Here is the damping coefficient of the virtual synchronizer. For the Laplace operator.

[0065] As an example, please continue reading Figure 2 The computational load power consumption scheduling module 13 includes a computational load power consumption mapping model 131 and a computational load scheduling model 132. The first end of the computational load power consumption mapping model 131 is connected to the second end of the virtual rotor control module 12, the second end of the computational load power consumption mapping model 131 is connected to the first end of the computational load scheduling model 132, and the second end of the computational load scheduling model 132 is connected to the third end of the data center 3.

[0066] As an example, IT devices are the direct executors of computing tasks, and their power consumption is most closely related to their load. Virtual inertial power compensation can be further achieved by constructing a mapping curve between the power consumption and load of IT devices.

[0067] As an example, the computational load power consumption mapping model 131 is used to calculate the computational power consumption of IT equipment, and the computational load scheduling model 132 uses the latency tolerance and time elasticity of the computational task to adjust the power consumed by its own IT equipment.

[0068] For details, please continue reading Figure 2 The load power consumption mapping model 131 is used to obtain the compensation power output of the virtual rotor control module 12. Calculate the computing power consumption of IT equipment before scheduling. Afterwards, the output is sent to the computing load scheduling model 132, which adjusts the computing power consumption of the IT equipment according to the delay tolerance and time flexibility of the computing task, to obtain the computing power consumption of the IT equipment , and is output to the load scheduling controller 34, which controls the IT equipment power adjustment.

[0069] As an example, the data center can be used as a rotor of a virtual synchronous machine. When the grid point bus 5 is in power shortage, the computing load can only be delayed for computing and cannot be abandoned for execution, so the data center can only delay the computing load for a certain time to realize power consumption reduction within the inertia response time, simulating the release of kinetic energy of the rotor of the synchronous machine. At the same time, the part of the delayed load will be calculated after the inertia response, simulating the absorption of kinetic energy of the rotor of the synchronous machine. Similarly, when the power grid is in power shortage, the data center can only calculate the computing load in advance within the inertia response time to realize power consumption increase, which is exactly in line with the power absorption and release law of the inertia response.

[0070] As an example, please refer to Figure 3 , Figure 3 The scheduling logic diagram of the computing load power consumption scheduling module is shown in the figure, wherein the vertical axis is the frequency of the grid point bus 5, the horizontal axis is time, the blue square below is the basic load power consumption of the data center before scheduling, the blue square in the middle is the load power consumption of the data center after scheduling, and the red dashed box represents the amplitude and timing of the load adjustment. When the grid point bus 5 is in power shortage and the frequency rapidly decreases with oscillation, the frequency change rate The data center releases the virtual rotor kinetic energy by reducing its own power consumption, which is equivalent to releasing power to the grid point bus 5 to help reduce the amplitude of the frequency decrease of the grid point bus 5; on the contrary, when the frequency of the grid point bus 5 rises, the frequency change rate , indicating that the grid point bus 5 is in power redundancy, and the data center absorbs the virtual rotor kinetic energy by increasing the power consumption to suppress the excessive rise and oscillation of the frequency of the grid point bus 5, and finally stabilizes the frequency of the grid point bus 5 near the new equilibrium value.

[0071] As an example, please refer to Figure 2, the energy storage power scheduling module 14 can include an energy storage primary frequency modulation unit 141, an energy storage auxiliary compensation unit 142, a fusion unit 143, the first end of the energy storage primary frequency modulation unit 141 is connected to the second end of the data center and the supporting energy storage grid connection point measurement unit 111, the second end of the energy storage primary frequency modulation unit 141 is connected to the first end of the fusion unit 143, the first end of the energy storage auxiliary compensation unit 142 is connected to the second end of the virtual rotor control module 12, the second end of the energy storage auxiliary compensation unit 142 is connected to the second end of the virtual rotor control module 12, the third end of the energy storage auxiliary compensation unit 142 is connected to the second end of the data center access end measurement unit 112, the fourth end of the energy storage auxiliary compensation unit 142 is connected to the second end of the fusion unit 143, and the third end of the fusion unit 143 is connected to the second end of the energy storage system 4.

[0072] As an example, the energy storage primary frequency modulation unit 141 is used to simulate the prime mover of the virtual synchronous machine, and the grid connection point bus 5 frequency is stabilized around the new balance value.

[0073] Specifically, the energy storage primary frequency modulation unit 141 obtains the frequency change rate of the grid connection point bus 5 output by the data center and the supporting energy storage grid connection point measurement unit 111 , first through the integral element Accumulate the frequency deviation, and then pass through the primary frequency modulation coefficient , get the primary frequency modulation power of the energy storage primary frequency modulation unit 141 .

[0074] As an example, the transfer function of the energy storage primary frequency modulation unit 141 is:

[0075]

[0076] Wherein, represents the primary frequency modulation coefficient, is the Laplace operator.

[0077] As an example, the energy storage auxiliary compensation unit 142 is used to build a dynamic power buffer layer, which can accurately compensate for the instruction tracking error caused by the delay of computing load scheduling. After the virtual rotor control module issues a power instruction, the energy storage auxiliary compensation unit 142 responds to all instruction powers in advance through millisecond-level fast charging and discharging, thereby offsetting the inherent error of computing task scheduling and ensuring the immediacy and accuracy of the system's external power output; it can also effectively isolate the interference of random power fluctuations of other electrical equipment in the data center on the regulation process, for example, the use of air conditioners, lighting, water pumps and other equipment will cause errors between the total power consumption of the data center and the power it should adjust. The energy storage auxiliary compensation unit 142 compensates for such uncontrolled power fluctuations in real time, so that the net power output of the data center strictly follows the virtual synchronous machine instruction, thereby ensuring the reliability and accuracy of the entire system participating in grid frequency regulation.

[0078] In particular, please continue to refer to Figure 2 , the energy storage auxiliary compensation unit 142 calculates the deviation between the compensation power output by the virtual rotor control module 12 and the power change amount of the data center 3 output by the data center access end measurement unit 112 ; and inputs the compensation power output by the virtual rotor control module 12 into the switch model, and the output value of the switch model is multiplied by the deviation between the compensation power output by the virtual rotor control module 12 and the power change amount of the data center 3 output by the data center access end measurement unit 112 to obtain the compensation power of the energy storage auxiliary compensation unit 142 .

[0079] As an example, the switch model controls whether the energy storage auxiliary compensation unit 142 is started according to the value of the compensation power output by the virtual rotor control module 12 ; when the compensation power output by the virtual rotor control module 12 is not equal to zero, the energy storage auxiliary compensation unit 142 is started; when the compensation power output by the virtual rotor control module 12 is equal to zero, the energy storage auxiliary compensation unit 142 is not started, which can avoid the situation that the energy storage power scheduling module is misoperated when the system does not have a disturbance and the data center has a power change.

[0080] Further, the primary frequency modulation power of the energy storage primary frequency modulation unit 141 and the compensation power of the energy storage auxiliary compensation unit 142 are fused by the fusion unit 143 to obtain the compensation power of the energy storage power scheduling module 14 , output to the main controller of the energy storage converter 41, send the energy storage converter scheduling instruction to the energy storage converter 41, and the energy storage converter 41 controls the energy storage element 42 to absorb or release active power according to the energy storage converter scheduling instruction.

[0081] As an example, the energy storage power scheduling module 14 has the function of auxiliary compensation, specifically, the energy storage power scheduling module 14 can act as a power compensation unit, can dynamically make up for the data center computing load power error through the millisecond-level fast charge and discharge response, and ensure the rapidity and accuracy of the overall output power of the virtual synchronous machine; the energy storage power scheduling module 14 can also act as a primary frequency regulation execution unit, can simulate the active-frequency droop characteristic of the virtual synchronous generator through the instruction issued by the virtual rotor control module 12, and provide power support for the power grid.

[0082] In the above-mentioned cooperative inertia response control device, the frequency change rate of the grid-connected point bus 5 and the power change amount of the data center are respectively collected by the data center and the supporting energy storage grid-connected point measurement unit 111 and the data center access end measurement unit 112, which can provide real-time and accurate basic data for subsequent regulation and control, and avoid control errors caused by measurement errors; the virtual rotor control module 12 simulates the virtual synchronous machine rotor motion equation, combines inertia compensation and damping compensation to solve compensation power, so that the device has synchronous machine type inertia response capability, and can effectively suppress frequency oscillation after power grid disturbance; the load power consumption scheduling module 13 adjusts power by using the delay tolerance of IT equipment, and the primary frequency regulation unit 141 and the energy storage auxiliary compensation unit 142 of the energy storage power scheduling module 14, so as to realize load-energy storage cooperative response, tap the elastic potential of IT equipment, rely on the rapidity of energy storage to ensure the accuracy and immediacy of power output, ultimately improve the reliability of the data center participating in the inertia response of the power grid, effectively suppress the frequency oscillation of the power grid, and restore the steady-state frequency deviation to the allowable range, thereby enhancing the stability and anti-disturbance ability of the power grid with a high proportion of renewable energy, and providing strong support for the frequency stability of the power grid.

[0083] In yet another embodiment, referring to Figure 4 The application also provides a cooperative inertia response control method, which can include the following steps: steps S1-S5.

[0084] Step S1: The data center and the supporting energy storage grid-connected point measurement unit measure the grid-connected point bus to obtain the frequency change rate of the grid-connected point bus.

[0085] Step S2: The data center access end measurement unit measures the power change amount of the data center access end to obtain the power change amount of the data center.

[0086] Step S3: The virtual rotor control module receives the frequency rate of change of the grid-connected point bus, simulates the rotor motion equation, and obtains the required virtual synchronous motor compensation power of the grid-connected point bus under disturbance through the virtual inertia power compensation control strategy.

[0087] Step S4: The energy storage power scheduling module performs energy storage auxiliary power compensation and primary frequency regulation power compensation according to the frequency rate of change of the grid-connected point bus and the power change amount of the data center room, and obtains the energy storage converter scheduling instruction.

[0088] Step S5: The load power consumption scheduling module executes task scheduling according to the required compensation power of the grid-connected point bus under disturbance, obtains the data center room scheduling instruction, adjusts the data center room and the energy storage system according to the energy storage converter scheduling instruction and the data center room scheduling instruction respectively, and completes the cooperative inertia response.

[0089] In the cooperative inertia response control method of the present application, the grid-connected point bus frequency rate of change and the data center room power change amount are accurately obtained through the data center and the supporting energy storage grid-connected point measurement unit and the data center access end measurement unit, which can provide real-time and accurate basic data support for subsequent regulation and control, avoid control inaccuracy caused by measurement deviation, and ensure the initial reliability of the regulation and control logic. By simulating the rotor motion equation and using the virtual inertia power compensation control strategy to solve the compensation power, the system can have the inertia response capability of a synchronous machine, effectively cope with grid disturbances, provide accurate target basis for subsequent power compensation, and suppress the large fluctuations of the grid frequency after disturbance. Through the energy storage power scheduling module combined with the frequency rate of change and the power change amount, energy storage auxiliary power compensation and primary frequency regulation power compensation are simultaneously realized, which can rely on the millisecond charging and discharging characteristics of energy storage to make up for potential scheduling delay and isolate non-controlled power fluctuations, while stabilizing the frequency within the allowed range, enhancing the stability and anti-disturbance capability of the high-proportion renewable energy grid, and ensuring the immediacy and accuracy of power output. Through the load power consumption scheduling module, task scheduling is performed according to the compensation power, and the energy storage converter scheduling instruction is coordinated, which can fully tap the flexibility potential of IT equipment, convert massive data center loads into high-quality grid adjustable resources, significantly improve the overall efficiency of data centers participating in grid inertia response, and effectively ensure grid frequency stability.

[0090] In step S1, please refer to step S1 in Figure 4 The data center and the supporting energy storage grid-connected point measurement unit measure the grid-connected point bus to obtain the frequency rate of change of the grid-connected point bus.

[0091] Specifically, the voltage signal on the grid connection point bus 5 is discretized and sampled at a first sampling rate using the data center and supporting energy storage grid connection point measurement unit 111 in the data center measurement module 11 to obtain a digital sampling sequence. The digital sampling sequence is then sent to the processor of the data center and supporting energy storage grid connection point measurement unit 111, and the frequency change rate of the grid connection point bus 5 is measured using the zero-crossing detection method. ROCOF ( k ); the rate of change of frequency of bus 5 at the time-domain grid connection point. Performing a Laplace transform, we obtain the frequency change rate of bus 5 at the grid connection point in the s-domain. .

[0092] As an example, the first sampling rate must satisfy the Nyquist sampling theorem.

[0093] As an example, zero-crossing detection can detect the time interval between two consecutive zero-crossings of a low-voltage signal. A sinusoidal period has two zero-crossings, therefore the measurement time... T corresponding frequency The frequency change rate of bus 5 at the grid connection point ROCOF It is the derivative of frequency with respect to time, i.e. In discrete systems, the rate of change of frequency of the grid-connected bus 5 can be approximated by the difference. ,in, This represents the k-th frequency measurement value. This represents the frequency measurement value at the (k+1)th time.

[0094] In step S2, please refer to Figure 4 In step S2, the data center access end measurement unit measures the power change of the data center access end to obtain the power change of the data center.

[0095] Specifically, the data center access measurement unit 112 in the data center measurement module 11 is used to measure the power change at the data center access point. Measurements were performed to obtain the power change in data center 3 in the time domain. .

[0096] As an example, you can set the measurement time. T Inside l There are 1 sampling points, with the initial sampling point being 1. l 0 The power change in data center 3 The expression is:

[0097]

[0098] in, Indicates the first k The power of data center 3 represents the first k the power of the data center 3.

[0099] As an example, the first k the power of the data center 3 The expression is:

[0100]

[0101] wherein, represents the first k measurement time, represents the voltage value of the first sampling point, represents the current value of the first sampling point, represents the initial sampling point.

[0102] Further, the Laplace transform is performed on the time-domain data center 3 power variation to obtain the s-domain data center 3 power variation .

[0103] In step S3, please refer to Figure 4 step S3, the virtual rotor control module receives the frequency rate of change of the grid connection point bus, simulates the rotor motion equation, and obtains the required compensation power of the virtual synchronous machine required by the grid connection point bus when subjected to disturbance through the virtual inertia power compensation control strategy.

[0104] As an example, please refer to Figure 5 , step S3 can include the following steps: steps S31-S33.

[0105] Step S31: determine the rotor motion equation of the virtual synchronous machine.

[0106] Step S32: based on the rotor motion equation of the virtual synchronous machine, through the relationship between frequency and angular frequency and the relationship between virtual rotational inertia and virtual inertia time constant, obtain the power shortage per unit value of the virtual synchronous machine.

[0107] Step S33: through the Laplace transform, obtain the required compensation power of the virtual synchronous machine required by the grid connection point bus when subjected to disturbance.

[0108] Specifically, in step S31, the virtual rotor control module 12 simulates the rotor motion equation of the virtual synchronous machine, and the expression is:

[0109]

[0110] wherein, represents the virtual mechanical power of the virtual synchronous machine, represents the electromagnetic power of the virtual synchronous machine, This indicates the current angular frequency of the virtual synchronizer. This indicates the rated angular frequency of the virtual synchronizer. This represents the virtual moment of inertia of the virtual synchronizer. This represents the damping coefficient of the virtual synchronizer.

[0111] As an example, in step S32, based on the relationship between frequency and angular frequency, the virtual moment of inertia of the virtual synchronizer is... Virtual inertial time constant of the virtual synchronizer The relationship can be represented as:

[0112]

[0113] in, The current angular frequency of the virtual synchronizer. , The current frequency of the virtual synchronizer; This is the rated power generation capacity of the virtual synchronous machine.

[0114] Furthermore, based on the relationship between frequency and angular frequency and the virtual moment of inertia... With virtual inertial time constant The relationship allows us to transform the rotor motion equations of the virtual synchronous machine into:

[0115]

[0116] in, This represents the per-unit value of the power deficit of the virtual synchronizer; This represents the virtual mechanical power of the virtual synchronizer. Indicates the electromagnetic power of the virtual synchronizer; This indicates the rated generating capacity of the virtual synchronous machine; This represents the virtual inertial time constant of the virtual synchronizer; Indicates the current frequency of the virtual synchronizer; Indicates the rated frequency of the virtual synchronizer; This represents the difference between the virtual synchronizer's rated frequency and its current frequency.

[0117] It is worth noting that, This refers to the rated power generation capacity of the virtual synchronous machine. In this application, the virtual synchronous machine is constructed using a data center and a supporting energy storage grid connection point measurement unit 111. Therefore, in this application, the rated power generation capacity of the virtual synchronous machine is... Rated load capacity of data center server clusters With supporting energy storage rated power generation capacity The sum is: .

[0118] Further, in step S33, the power deficiency per unit of the virtual synchronous machine is obtained by taking Laplace transform of the required compensation power of the virtual synchronous machine to respond to the disturbance of the grid-connected point bus 5 , and the virtual synchronous machine compensation power is transmitted to the load power consumption scheduling module 13 and the energy storage power scheduling module 14.

[0119] As an example, when the disturbance occurs, the power deficiency per unit of the virtual synchronous machine is compensated by the rotor kinetic energy release simulated by the data center and the supporting energy storage grid-connected point measurement unit 111. Since the data center can actually act as a load, it can be considered here that the frequency of the data center is consistent with the frequency change trend of the grid-connected point bus 5, so that is numerically consistent with . Therefore, when taking Laplace transform, the can be replaced by .

[0120] As an example, the transfer function between the virtual synchronous machine compensation power and the frequency change rate of the grid-connected point bus 5 can be represented as:

[0121]

[0122] wherein is the virtual inertia time constant of the virtual synchronous machine, is the rated frequency of the virtual synchronous machine, is the damping coefficient of the virtual synchronous machine, and is the Laplace operator.

[0123] In step S4, please refer to step S4 in Figure 4 , the energy storage power scheduling module performs energy storage auxiliary power compensation and primary frequency modulation power compensation according to the frequency change rate of the grid-connected point bus and the power change amount of the data center to obtain the energy storage converter scheduling instruction.

[0124] Specifically, the energy storage primary frequency modulation unit 141 in the energy storage power scheduling module 14 acquires the frequency change rate of the grid-connected point bus 5 output by the data center and the supporting energy storage grid-connected point measurement unit 111, first accumulates the frequency deviation through the integral element , and then simulates the primary frequency modulation of the virtual synchronous machine through the primary frequency modulation coefficient to obtain the primary frequency modulation power of the energy storage primary frequency modulation unit 141.

[0125] ​As an example, the primary frequency modulation power of the energy storage primary frequency modulation unit 141 is expressed as:

[0126]

[0127] wherein, denotes the primary frequency modulation coefficient, denotes the initial time, denotes the current frequency of the virtual synchronous machine, denotes the current time.

[0128] As an example, the transfer function of the energy storage primary frequency modulation unit 141 is:

[0129]

[0130] wherein, denotes the primary frequency modulation power, denotes the primary frequency modulation coefficient, is the Laplace operator.

[0131] Further, the energy storage auxiliary compensation unit 142 in the energy storage power scheduling module 14 calculates the deviation between the compensation power output by the virtual rotor control module 12 and the power change amount of the data center 3 output by the data center access end measurement unit 112 ; and inputs the compensation power output by the virtual rotor control module 12 into the switch model, and the output value of the switch model is processed by a multiplier with the deviation between the compensation power output by the virtual rotor control module 12 and the power change amount of the data center 3 output by the data center access end measurement unit 112 , to obtain the compensation power of the energy storage auxiliary compensation unit 142.

[0132] As an example, the compensation power of the energy storage auxiliary compensation unit 142 is expressed as:

[0133]

[0134] wherein, denotes the compensation power output by the virtual rotor control module 12 at time t; denotes the power change amount of the data center 3 output by the data center access end measurement unit 112 at time t, denotes the transfer function of the switch model.

[0135] As an example, the switch model takes the compensation power output by the virtual rotor control module 12 at time t As a signal, when t moment by virtual rotor control module 12 output compensation power When not equal to 0, the energy storage power scheduling module 14 is started, which avoids the situation that the energy storage power scheduling module 14 is misoperated due to the power change of the data center without system disturbance, the transfer function of the switching model The expression is:

[0136] .

[0137] Further, the primary frequency modulation power of the energy storage primary frequency modulation unit 141 The compensation power of the energy storage auxiliary compensation unit 142 Fusion through the fusion unit 143, get the compensation power of the energy storage power scheduling module 14 That is, the energy storage converter scheduling instruction, the energy storage converter scheduling instruction is output to the main controller of the energy storage converter 41, controls the real-time power setting value or the instruction of the frequency-power regulation curve in the main controller of the energy storage converter 41, controls the energy storage converter 41 to absorb or release active power.

[0138] In step S5, please refer to Figure 4 The load power consumption scheduling module calculates the compensation power required when the grid point bus is disturbed to perform task scheduling, obtains the data center scheduling instruction, adjusts the data center and the energy storage system according to the energy storage converter scheduling instruction and the data center scheduling instruction, and completes the collaborative inertia response.

[0139] As an example, please refer to Figure 6 Step S5 can include the following steps: steps S51-S53.

[0140] Step S51: The calculation load power consumption mapping model in the load power consumption scheduling module obtains the compensation power output by the virtual rotor control module, and calculates the computing power consumption of the IT equipment.

[0141] Step S52: The calculation load scheduling model in the load power consumption scheduling module determines the scheduling objective function and the calculation task constraint by using the delay tolerance and the time flexibility of the task, and obtains the data center scheduling instruction.

[0142] Step S53: According to the energy storage converter scheduling instruction and the data center scheduling instruction, the data center and the energy storage system are adjusted respectively, and the collaborative inertia response is completed.

[0143] Specifically, in step S51, the calculation load power consumption mapping model 131 in the load power consumption scheduling module 13 obtains the compensation power output by the virtual rotor control module 12 , calculates the computing power consumption of the IT equipment before scheduling Afterwards, the output is sent to the computing load scheduling model 132.

[0144] As an example, the IT equipment can include one or more servers, which can be online servers.

[0145] As an example, a server can be modeled as a basic unit first, and then aggregated to a cabinet and the entire data center.

[0146] Specifically, since frequent start and stop of the online server will have a negative impact on the switching loss and service life of the online server, the present application does not consider adjusting the number of servers, and does not consider the start and stop of the server. Therefore, the computing power consumption of a server at time t is which is related to the load rate thereof and can be expressed as:

[0147]

[0148] wherein, is the computing power consumption of the server at time t; is the idle power consumption of the server; is the maximum power consumption of the server; is the computing task of the server at time t, is the maximum computing task of the server.

[0149] Further, when aggregated to a cabinet and the entire data center, the IT equipment can be considered as a server cluster including I servers. Therefore, the computing power consumption of the IT equipment at time t is which is related to the load rate thereof and can be expressed as:

[0150]

[0151] wherein, is the computing power consumption of the server cluster at time t; is the idle power consumption of the i-th server; is the maximum power consumption of the i-th server; is the computing task of the i-th server at time t, is the maximum computing task of the i-th server.

[0152] Further, in step S52, when the data center needs to adjust the power consumption of the IT equipment thereof, the computing load scheduling model 132 in the computing load scheduling model 13 adjusts the computing task by using the delay tolerance and time flexibility of the computing task.

[0153] As an example, the computing task of the i-th server at time t can be expressed as:

[0154]

[0155] wherein, is the instant computing task of the i-th server at time t, is the time-critical delay-tolerant computing task of the i-th server at time t, is the delay-in-computing task of the i-th server at time t, is the delay-tolerant computing task of the i-th server at time t, is the maximum computing task of the i-th server.

[0156] As an example, the instant computing task refers to a workload that needs to give an immediate response; the delay-tolerant computing task refers to a workload that is first submitted to the task queue of the data center and can wait until the data center has sufficient resources to respond.

[0157] As an example, the instant computing task of the i-th server at time t and the time-critical delay-tolerant computing task of the i-th server at time t both need to perform computing immediately, which is the non-adjustable part; the delay-tolerant computing task of the i-th server at time t still has a delay time, and the delay-in-computing task of the i-th server at time t can end the delay computing in advance, which is the adjustable part. Therefore, the scheduling variable of the computing load scheduling model 132 can be set as the delay-tolerant computing task of the i-th server at time t and the delay-in-computing task of the i-th server at time t .

[0158] As an example, when the grid point bus 5 is disturbed, the data center and the supporting energy storage grid point measurement unit 111 capture the frequency change rate of the grid point bus 5, and calculate the corresponding compensation power according to the compensation power , the scheduling strategy of the computing load scheduling model 132 can be set. The computing load scheduling model 132 in the computing load power consumption scheduling module 13 determines the scheduling objective function and the computing task constraint by using the delay tolerance and the time flexibility of the computing task, and obtains the data machine room scheduling instruction.

[0159] Specifically, the scheduling objective of the computing load scheduling model 132 can be set as minimizing the difference between the compensation power output by the virtual rotor control module 12 and the actual adjustment power adjustment amount of the data center, and the corresponding scheduling objective function is:

[0160]

[0161] wherein, is the instant computing task of the i-th server at time t, is the in-delay computing task of the i-th server at time t, is the computing power consumption of the server cluster at time t-1, is the compensation power output by the virtual rotor control module 12, is the idle power consumption of the i-th server, is the maximum power consumption of the i-th server, is the instant computing task of the i-th server at time t, is the time-critical in-delay computing task of the i-th server at time t, is the maximum computing task of the i-th server, is the total number of I servers in the IT equipment.

[0162] As an example, the computing task constraints of the computing load scheduling model 132 can include: maximum computing capacity constraints of servers, computing task dependency constraints, and task deadline constraints.

[0163] Specifically, the maximum computing capacity constraint of a server is that, for each server, the amount of computing tasks at time t shall not exceed the corresponding maximum computing capacity , the expression is:

[0164]

[0165] wherein, is the instant computing task of the i-th server at time t; is the time-critical in-delay computing task of the i-th server at time t; is the instant computing task of the i-th server at time t; is the in-delay computing task of the i-th server at time t; represents a safety threshold coefficient, which avoids the server being in full load and thus causing its performance to sharply decrease.

[0166] As an example, there is usually a dependency relationship between computing tasks, and a computing task must start after all its preceding computing tasks are completed, forming a directed acyclic graph (DAG). Therefore, the computing task dependency constraint can be set as: when scheduling, the delay time of the instant computing task and the in-delay computing task shall be less than or equal to the execution time of their corresponding dependent computing tasks, the expression is:

[0167]

[0168]

[0169] wherein, is the delayable computing task of the i-th server at time t; is the in-delay computing task of the i-th server at time t, denotes the delay time of computing task x, denotes the execution time of computing task x, denotes the computing task that depends on computing task x.

[0170] As an example, for the delayable computing task and the in-delay computing task of the adjustable part, each computing task has a deadline. The task deadline constraint can be set as: the delayable computing task and the in-delay computing task must be completed before the corresponding deadline to ensure the timeliness of the computing task, and the expression is:

[0171]

[0172]

[0173] wherein, is the delayable computing task of the i-th server at time t; i is the in-delay computing task of the i-th server at time t, denotes the deadline of computing task x, i denotes the time required for computing task x. Further, according to the scheduling objective function and the computing task constraint, the specific parameters of the delayable task and the in-delay task

[0174] that need to be adjusted are optimized to generate the data center scheduling instruction. As an example, the specific parameters of the delayable task and the in-delay task

[0175] in the data center scheduling instruction can include server or cabinet number, task type, start time of scheduling operation, duration or expected completion time, expected power consumption, etc.

[0176] ​​As an example, in step S53, the energy storage inverter scheduling instruction is connected to the energy storage system 4 through a high-speed communication link and a power electronic drive circuit, sends a real-time power setting value or a frequency-power regulation curve instruction to the energy storage system 4, commands it to absorb or release active power, realizes fast compensation and steady-state support; the data center scheduling instruction is connected to the data center 3 through network communication or hardwiring (dry contact signal), adjusts the IT equipment 32 of the data center 3, realizes task migration and power consumption adjustment, and thus completes the coordinated inertia response of the data center and the matching energy storage.

[0177] In the coordinated inertia response control method of the present application, by focusing on the adjustable potential of the massive data center load, and taking the unique "computing time migration" characteristic as the core power compensation means, the traditional inertia supply path relying on energy storage or new energy equipment is replaced, the inertia response and primary frequency modulation key characteristics of the synchronous generator can be accurately simulated; without additional new energy storage hardware or modification of new energy equipment, the data center load can be flexibly scheduled to effectively suppress the grid frequency oscillation, smooth the frequency change rate, and at the same time, the steady-state frequency deviation is stably restored to the allowable range, thereby enhancing the stability and anti-disturbance ability of the high-proportion renewable energy power grid, fundamentally improving the grid frequency regulation effect; it can also significantly broaden the source of grid inertia supply, break the limitation of traditional single resource, and help the optimization of adjustable resources of the power grid, especially in the high-proportion renewable energy grid-connected scenario, it can effectively enhance the anti-disturbance ability of the power grid, further guarantee the stability and safety of the power grid operation, and provide key technical support for the efficient operation of the new type of power system.

[0178] It should be understood that, although each step in the flowchart of the accompanying drawings is shown in sequence according to the direction of the arrow, these steps are not necessarily executed in sequence according to the direction of the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other sequences. Moreover, at least part of the steps in the accompanying drawings can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of these sub-steps or stages is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or sub-steps or stages of other steps.

[0179] The technical features of the above embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not exist contradictions, they should be considered as the scope of the present application.

[0180] While the application has been described by way of example, it should be appreciated that any modification and / or alteration of this application be deemed within the scope and spirit of the application unless otherwise indicated.

Claims

1. A cooperative inertia response control device, characterized in that, The first end of the cooperative inertia response control device is connected to the first end of the voltage transformer and the current transformer. The second end of the voltage transformer and the current transformer is connected to the grid connection point bus. The third end of the voltage transformer and the current transformer is connected to the first end of the data center and the first end of the energy storage system. The second end of the cooperative inertia response control device is connected to the second end of the data center. The third end of the cooperative inertia response control device is connected to the third end of the data center. The fourth end of the cooperative inertia response control device is connected to the second end of the energy storage system. The cooperative inertia response control device includes a data center measurement module, a virtual rotor control module, a computing load power consumption scheduling module, and an energy storage power scheduling module. The first end of the data center measurement module is connected to the first ends of the voltage transformer and current transformer; the second end of the data center measurement module is connected to the second end of the data center; the third end of the data center measurement module is connected to the first end of the virtual rotor control module and the first end of the energy storage power scheduling module; the fourth end of the data center measurement module is connected to the second end of the energy storage power scheduling module; the second end of the virtual rotor control module is connected to the first end of the computing load power consumption scheduling module; the second end of the computing load power consumption scheduling module is connected to the third end of the data center; the third end of the energy storage power scheduling module is connected to the second end of the virtual rotor control module; the fourth end of the energy storage power scheduling module is connected to the second end of the virtual rotor control module; and the fifth end of the energy storage power scheduling module is connected to the second end of the energy storage system. The energy storage power scheduling module includes an energy storage primary frequency regulation unit, an energy storage auxiliary compensation unit, and a fusion unit. The first end of the energy storage primary frequency regulation unit is connected to the second end of the measurement unit at the data center and the supporting energy storage grid connection point. The second end of the energy storage primary frequency regulation unit is connected to the first end of the fusion unit. The first end of the energy storage auxiliary compensation unit is connected to the second end of the virtual rotor control module. The second end of the energy storage auxiliary compensation unit is connected to the second end of the virtual rotor control module. The third end of the energy storage auxiliary compensation unit is connected to the second end of the measurement unit at the data center access end. The fourth end of the energy storage auxiliary compensation unit is connected to the second end of the fusion unit. The third end of the fusion unit is connected to the second end of the energy storage system.

2. The cooperative inertia response control device according to claim 1, characterized in that, The data center measurement module includes a data center and supporting energy storage grid connection point measurement unit and a data center access end measurement unit. The data center and supporting energy storage grid connection point measurement unit and the data center access end measurement unit are set in parallel and are used to measure the frequency change rate of the grid connection point bus and the power change of the data center, respectively.

3. The cooperative inertia response control device according to claim 1, characterized in that, The transfer function of the virtual rotor control module Represented as: in, The frequency variation rate of the bus at the grid connection point. To compensate for power, The virtual inertial time constant of the virtual synchronizer. The rated frequency of the virtual synchronizer. Here is the damping coefficient of the virtual synchronizer. For the Laplace operator.

4. The cooperative inertia response control device according to claim 1, characterized in that, The computing load power consumption scheduling module includes a computing load power consumption mapping model and a computing load scheduling model. The first end of the computing load power consumption mapping model is connected to the second end of the virtual rotor control module, the second end of the computing load power consumption mapping model is connected to the first end of the computing load scheduling model, and the second end of the computing load scheduling model is connected to the third end of the data center.

5. The cooperative inertia response control device according to claim 1, characterized in that, The transfer function of the energy storage primary frequency regulation unit for: in, This represents the primary frequency modulation coefficient. For the Laplace operator.

6. A cooperative inertia response control method, characterized in that, The cooperative inertia response control device as described in any one of claims 1 to 5 includes the following steps: The measurement unit at the data center and its supporting energy storage grid connection point measures the grid connection point bus to obtain the frequency change rate of the grid connection point bus; The data center access measurement unit measures the power change at the data center access point to obtain the power change of the data center. The virtual rotor control module receives the frequency change rate of the grid connection point bus, simulates the rotor motion equation, and obtains the virtual synchronous motor compensation power required when the grid connection point bus is disturbed through a virtual inertial power compensation control strategy. The energy storage power scheduling module performs energy storage auxiliary power compensation and primary frequency regulation power compensation based on the frequency change rate of the grid connection point bus and the power change of the data center, and obtains the energy storage converter scheduling command. The load power consumption scheduling module performs task scheduling based on the compensation power required when the grid connection point bus is disturbed, obtains the data center scheduling instruction, and adjusts the data center and energy storage system according to the energy storage converter scheduling instruction and the data center scheduling instruction respectively to complete the coordinated inertia response.

7. The cooperative inertia response control method according to claim 6, characterized in that, The virtual rotor control module receives the frequency change rate of the grid connection point bus, simulates the rotor motion equation, and obtains the virtual synchronous motor compensation power required when the grid connection point bus is disturbed through a virtual inertial power compensation control strategy, including: Determine the rotor motion equation of the virtual synchronous motor; Based on the rotor motion equation of the virtual synchronous motor, the per-unit power deficit of the virtual synchronous motor is obtained through the relationship between frequency and angular frequency and the relationship between virtual moment of inertia and virtual inertial time constant. The virtual synchronous motor compensation power required when the grid connection point bus is disturbed is obtained by using the Laplace transform.

8. The cooperative inertia response control method according to claim 6, characterized in that, The load power consumption scheduling module performs task scheduling based on the compensation power required when the grid connection point bus is disturbed, obtains the data center scheduling instruction, and adjusts the data center and energy storage system according to the energy storage converter scheduling instruction and the data center scheduling instruction respectively to complete the coordinated inertia response, including: The computational load power consumption mapping model in the computational load power consumption scheduling module obtains the compensation power output by the virtual rotor control module and calculates the computational power consumption of the IT equipment. The computational load scheduling model in the computational load power consumption scheduling module utilizes the latency tolerance and time flexibility of tasks to determine the scheduling objective function and computational task constraints, thereby obtaining data center scheduling instructions. Adjustments are made to the data center and energy storage system according to the scheduling instructions of the energy storage converter and the data center, respectively, to complete the coordinated inertia response.

9. The cooperative inertia response control method according to claim 8, characterized in that, The scheduling objective function for: in, Let be the time-delayable computing task of the i-th server at time t. Let be the computation task of the i-th server during the delay at time t. The compensation power output by the virtual rotor control module. Let i be the idle power consumption of the i-th server. The computational power consumption of the server cluster at time t-1. Let be the maximum power consumption of the i-th server. Let be the real-time computing task of the i-th server at time t. Let be the time-delay calculation task for the i-th server at time t. For the maximum computing task of the i-th server, There are I servers in the IT equipment.

Images