CN-121984062-A - Electrical cooperative control method and system of distributed energy storage system and computer readable storage medium

CN121984062ACN 121984062 ACN121984062 ACN 121984062ACN-121984062-A

Abstract

The present invention relates to the field of energy storage technologies, and in particular, to an electrical cooperative control method, system and computer readable storage medium for a distributed energy storage system. The method comprises the steps of firstly collecting global optimization targets of a system and local state information of all energy storage units, calculating an electric sensitivity matrix of all the energy storage units on a public connection point PCC based on a node admittance matrix, dividing virtual cooperative clusters according to the magnitude and the relativity of electric sensitivity, generating a total power instruction of the system according to the global optimization targets, distributing total power instruction distribution weights to all the virtual cooperative clusters to obtain a power instruction of each virtual cooperative cluster, generating a preliminary power instruction of each energy storage unit in each virtual cooperative cluster based on a consistency algorithm, introducing compensation items based on the electric sensitivity to correct the preliminary power instruction to obtain a final power instruction of each energy storage unit, and executing corresponding instructions. The invention obviously improves the control precision and the operation efficiency of the distributed energy storage system.

Inventors

LEI HAODONG
TONG GUOPING
WU JIAN
HU YUANXIAO
ZHAO JUNLIANG
Duan Zhaorong
YANG CHAORAN
BAI PANXING
WEI YU
LV JIAN
HAN HUI
CAO XI
WANG NING
ZHOU LIREN
GU YI
LIU BO
LI XIAOCHEN
CAO CHUANZHAO
PEI JIE
LIU MINGYI

Assignees

中国华能集团清洁能源技术研究院有限公司
华能国际电力股份有限公司上海石洞口第二电厂

Dates

Publication Date: 20260505
Application Date: 20251216

Claims (10)

1. An electrical cooperative control method of a distributed energy storage system is characterized by comprising the following steps: S1, collecting a global optimization target of a system and local state information of each energy storage unit; s2, calculating an electric sensitivity matrix of each energy storage unit to the public connection point PCC based on the node admittance matrix, and dividing a virtual cooperative cluster according to the magnitude and the correlation of the electric sensitivity; S3, generating a system total power instruction according to the global optimization target, and distributing the total power instruction to each virtual cooperative cluster according to cluster power distribution weight to obtain a power instruction of each virtual cooperative cluster; S4, generating a preliminary power instruction of the energy storage unit based on a consistency algorithm in each virtual cooperative cluster; S5, introducing a compensation term based on electrical sensitivity to correct the preliminary power instruction, obtaining a final power instruction of each energy storage unit, and controlling each energy storage unit to execute the corresponding final power instruction.
2. The method of claim 1, wherein S1 further comprises: S11, determining a global optimization target of a system, and quantifying the target into an executable control parameter, wherein the global optimization target comprises smoothing PCC point power fluctuation, participating in grid frequency modulation or realizing peak clipping and valley filling, the smoothing PCC point power fluctuation comprises setting a time constant of a low-pass filter or a maximum change rate threshold of the power fluctuation, the participating in the grid frequency modulation comprises setting a frequency-power droop coefficient and a frequency deviation dead zone, and the realizing peak clipping and valley filling comprises setting a target power or a time-varying target power curve; S12, periodically acquiring own local state information by utilizing each energy storage unit through a local sensor, and summarizing the local state information.
3. The method of claim 1, wherein S2 further comprises: s21, constructing a node admittance matrix Y of the system based on the topological structure of the power distribution network and the resistance and reactance parameters of each branch; S22, based on the node admittance matrix Y, obtaining a sensitivity matrix of the change of the injection power of each energy storage unit to the change of the voltage amplitude of the PCC at the public connection point by solving a system tide equation or calculating an inverse matrix of a jacobian matrix so as to represent the sensitivity of the active power change of the jth energy storage unit to the ith node voltage, wherein the calculation formula of the sensitivity matrix is as follows: ; Wherein, the Representing the magnitude of the voltage at node i, Representing the active power injected into node j; S23, calculating the electrical distance between any two energy storage units m and n based on the sensitivity matrix S The calculation formula is as follows: Wherein, the And Respectively representing the m-th column and the n-th column vectors in the sensitivity matrix S, namely the sensitivity vectors of two units to all nodes of the whole system, Is the two norms of the vector; s24, constructing an electric distance matrix based on the electric distance between any two energy storage units m and n, and dividing the energy storage units with the electric distances smaller than a first threshold value into the same virtual cooperative cluster according to the electric distance matrix; S25, selecting a leading unit in each virtual cooperative cluster according to one or more indexes of rated capacity, real-time charge state SOC and communication link quality of the energy storage unit so as to perform cooperative calculation and information collection in the clusters.
4. The method of claim 1, wherein S3 further comprises: s31, calculating and obtaining a total power reference value required to be absorbed or released by the whole distributed energy storage system based on a global optimization target; S32, for each virtual cooperative cluster, calculating the power distribution weight of the cluster according to the real-time state information of each energy storage unit in the cluster The power allocation weight The calculation basis comprises the sum of rated capacities of the energy storage units in the cluster, the sum of the currently available maximum charge and discharge power or the deviation of the average state of charge (SOC) of the energy storage units in the cluster and the ideal SOC; s33, distributing weight according to the power of each cluster Total power reference value of system Proportionally distributing the power instructions to each virtual cooperative cluster to obtain the power instructions of each virtual cooperative cluster The power allocation weight calculation formula is: ; Wherein, the Total power reference value.
5. The method of claim 4, wherein S31 further comprises: when the global optimization target is smooth PCC point power fluctuation, the calculation formula of the total power reference value is as follows: ; Wherein, the The power is actually measured for the PCC point, The power reference value is a PCC point power reference value obtained through a low-pass filter; When the global optimization target is participated in grid frequency modulation, the calculation formula of the total power reference value is as follows: ; Wherein, the For the measured frequency of the power grid, For the nominal frequency to be a set value, Is the frequency-power droop coefficient; When the global optimization target is to realize peak clipping and valley filling, the calculation formula of the total power reference value is as follows: ; Wherein, the Is the set target power.
6. The method of claim 1, wherein S4 further comprises: S41, the power instruction of each virtual cooperative cluster is processed Preliminarily distributing the initial power instructions to all the energy storage units in the cluster according to a preset rule to obtain initial power instructions of all the energy storage units The preset rule comprises allocation according to rated capacity proportion or allocation according to maximum charge-discharge power proportion; s42, exchanging charge state SOC and power instruction information of adjacent energy storage units of the cluster, and iteratively updating own power instruction based on a consistency algorithm, wherein the updating rule of the consistency algorithm is as follows: ; Wherein: representing a power instruction of the ith energy storage unit in the kth iteration; A set of neighbor units in direct communication with the ith energy storage unit; Is a communication weight; is a power consistency gain coefficient; The average SOC of all energy storage units in the cluster at the kth iteration; the SOC of the ith energy storage unit at the kth iteration; equalizing gain coefficients for the SOC; S43, generating a preliminary power instruction, namely when the iterative updating process meets the convergence condition or reaches the maximum iterative times, taking the current power instruction of each energy storage unit as the preliminary power instruction And the convergence condition is that the power instruction variation of all the energy storage units in the cluster is smaller than a second threshold value.
7. The method of claim 1, wherein S5 further comprises: s51, for each virtual cooperative cluster, calculating the sensitivity of each energy storage unit in each cluster to the PCC voltage of the common connection point Average sensitivity to clusters Deviation of (2) The calculation formula is as follows: Wherein, the N is the number of energy storage units in the cluster; s52, calculating an electric compensation term based on the sensitivity deviation Calculating the electric compensation term of each energy storage unit The calculation formula is as follows: Wherein, K is a preset compensation coefficient; s53, the electric compensation term And primary power instruction based on consistency algorithm Adding to obtain final power instruction of each energy storage unit ; S54, the final power instruction And performing amplitude limiting treatment to ensure that the maximum charge and discharge power range of the corresponding energy storage unit is not exceeded.
8. The method according to claim 1, wherein the step S5 is performed by each energy storage unit for executing the corresponding final power command by: Final power instruction of each energy storage unit through central controller or master unit of each virtual cooperative cluster The final power instruction is sent to a corresponding energy storage unit local controller to receive Generating a Pulse Width Modulation (PWM) signal of the converter through a double closed-loop control algorithm, and controlling the converter to execute corresponding charge and discharge operations, wherein the double closed-loop control comprises the following steps: outer loop power control based on power command And the actual output power Is used for generating a current reference value through a PI regulator ; Inner loop current control according to current reference value And actual output current And (3) generating a modulation signal through a PI regulator to drive a converter power switching device.
9. An electrical cooperative control system of a distributed energy storage system, comprising, The information acquisition module is used for acquiring global optimization targets of the system and local state information of each energy storage unit; The electric sensitivity calculation and cluster division module calculates an electric sensitivity matrix of each energy storage unit to the public connection point PCC based on the node admittance matrix, and divides a virtual collaborative cluster according to the magnitude and the correlation of the electric sensitivity; the power instruction distribution module generates a system total power instruction according to the global optimization target, distributes the total power instruction to each virtual cooperative cluster according to cluster power distribution weight, and obtains the power instruction of each virtual cooperative cluster; The preliminary power instruction generation module generates a preliminary power instruction of the energy storage unit based on a consistency algorithm in each virtual cooperative cluster; And the final power instruction correction module is used for correcting the preliminary power instruction by introducing a compensation item based on the electrical sensitivity to obtain a final power instruction of each energy storage unit and controlling each energy storage unit to execute the corresponding final power instruction.
10. A computer readable storage medium storing a computer program which, when executed by a processor, implements the method of any one of claims 1-8.

Description

Electrical cooperative control method and system of distributed energy storage system and computer readable storage medium Technical Field The present invention relates to the field of energy storage technologies, and in particular, to an electrical cooperative control method, system and computer readable storage medium for a distributed energy storage system. Background Along with the continuous improvement of the permeability of renewable energy sources (such as photovoltaic and wind power) and the increasing of the peak-valley difference of the load of the power system, the distributed energy storage system is used as a key flexible regulating resource, and plays an increasingly important role in the aspects of improving the stability of a power grid, improving the quality of electric energy, improving the utilization efficiency of the energy sources and the like. Distributed energy storage systems are typically composed of a plurality of energy storage units distributed in a distributed manner, and how to control these units efficiently and cooperatively to achieve the optimization objective at the system level is the focus of research in the technical field. At present, control methods for distributed energy storage systems can be mainly divided into two main types, namely centralized control and decentralized control. The centralized control method relies on a central controller which collects state information of all energy storage units and calculates control instructions of each unit based on a global optimization model and then issues and executes the control instructions. Although the method can realize global optimization, the method has obvious defects that firstly, the method is highly dependent on a high-speed and high-reliability communication network, the investment and maintenance cost is high, secondly, a central controller becomes a single-point fault source of the system, once the system fails, the whole system is out of control, the robustness is poor, finally, the expandability of the system is poor, and a new energy storage unit or a reduced energy storage unit is required to reconfigure a central control strategy, so that the flexibility is insufficient. The decentralized control method (or referred to as local control) then the energy storage units respond autonomously only on the basis of local measurement information (e.g. local frequency, voltage). This approach, while highly reliable and requiring low communication, lacks coordination among the units. This can easily lead to multiple energy storage units responding homogeneously to the same system disturbance, creating a "power oscillation" or "overcorrected" phenomenon. For example, when the frequency drops, all cells are discharged at high power at the same time, which may cause frequency overshoot and even cause a new round of instability. Furthermore, decentralized control has difficulty achieving advanced optimization objectives such as state of charge equalization, power economy allocation, etc., because the state of other units cannot be known. In order to achieve both reliability and optimization capability, a distributed cooperative control method based on a consistency algorithm has appeared in recent years. The method enables the system to achieve global consensus without a central controller through local information interaction among neighbor units. However, the existing cooperative control research focuses on the common knowledge (such as SOC equalization) of the information layer, and often ignores the difference of the electrical connection relationship of the energy storage units in the physical power grid. The line impedance and electrical distance difference in the power grid exist objectively, so that even if the same power command is output, the influence effect of the energy storage units at different positions on the public connection point or the key bus is different. The lack of the electrical coupling relation can cause two main problems that firstly, the power distribution cannot be accurately matched with the requirement of a global target of the system, the control effect is reduced, and secondly, unexpected circulation can be generated between units due to the difference of outlet voltages, so that not only is the system loss increased, but also the overload of equipment can be caused, and the safe and stable operation of the system is threatened. Therefore, there is an urgent need in the art for a control method of a distributed energy storage system that can achieve efficient collaborative optimization, and also has high reliability, while being able to fully consider and overcome the influence of electrical coupling differences. Disclosure of Invention The present invention aims to solve at least one of the technical problems in the related art to some extent. To this end, a first object of the present invention is to propose an electrical cooperative control method of a distributed energy stora