Multi-direct-current sending-end power grid stability control method and related device

Abstract

The invention belongs to the technical field of power systems, and discloses a multi-direct-current sending-end power grid stability control method and a related device. The multi-direct-current sending-end power grid stability control method comprises the steps that the name of a fault direct-current line and the direct-current sending-out power, lost due to a fault, of the fault direct-current line are acquired; inputting the acquired name of the fault DC line and the DC output power of the fault DC line lost due to the fault into an intelligent agent configured by a DC group where the fault DC line is located, and obtaining the emergency power adjustment amount of other DC lines of the DC group where the fault DC line is located through the intelligent agent; unbalanced power caused by faults is balanced through the emergency power adjustment amount of other direct-current lines, and stable control over the multi-direct-current sending-end power grid is achieved. According to the technical scheme disclosed by the invention, the generator tripping amount during stable control is greatly reduced, and the influence on daily production and life can be reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric power systems, and in particular relates to a multi-DC sending-end power grid stability control method and related devices. Background Art

[0002] In recent years, due to the large-scale commissioning of renewable energy power generation and ultra-high voltage direct current transmission, the "three highs" (i.e., high proportion of new energy, high degree of power electronics, and high degree of AC / DC coupling) characteristics of the power grid have become prominent and the stability characteristics are complex; with the commissioning of more ultra-high voltage direct current transmission projects, power grids in many regions have shown the important characteristics of multiple DC transmission or multiple input. Explanatory, the construction of large-scale new energy sending-end power grids will significantly enhance the ability to optimize the allocation of energy resources on a large scale, large scale, and high efficiency, which is of great and far-reaching significance for the large-scale development of clean energy and large energy bases, the large-scale optimization of energy resources, and the guarantee of energy security. In addition, while the sending-end power grid transmits high-power electric energy to the receiving-end power grid, it also provides power support to the local power grid to meet the regional electricity demand. Therefore, the stability of the sending-end power grid has a great impact on the safe and stable operation of the entire power system.

[0003] In the process of development, the power grid has gradually formed a pattern of high proportion of new energy and multi-DC transmission. Under the impact of fault disturbance, the interaction between the AC system and the DC system and the multi-circuit DC system (explanatory, the distance between the DC landing points is relatively close, the electrical coupling is close, the mutual influence is large, and there are complex interactions between DC and DC, and between DC and AC) will have a significant impact on the grid's disturbance behavior characteristics, safety and stability level, and power transmission and receiving capabilities of important sections. Further specific examples and explanations, when a DC project fails, the converter bus voltage of other DC projects with close electrical distances will be disturbed, affecting the transmission power of the DC project; after the DC project has a lockout fault, the power imbalance impact will also cause the power angle stability problem of the DC near the fault; the converter station is equipped with a large number of reactive compensation devices, but when the DC has a lockout fault, the reactive compensation control device in the converter station may not be able to act immediately, which will cause a large amount of reactive power to be fed into the AC network, causing the transient overvoltage of the nearby new energy units to be disconnected from the grid, further deteriorating the operating conditions of the sending end system.

[0004] At present, there are relatively complete emergency control methods for various types of DC faults, which are mainly based on traditional machine cutting and load shedding. The above existing emergency control methods will have a great impact on daily production and life, and a new multi-DC sending-end power grid stability control solution is urgently needed. Summary of the invention

[0005] The purpose of the present invention is to provide a method and related device for stabilizing a multi-DC power grid to solve one or more of the above-mentioned technical problems. The technical solution disclosed in the present invention adjusts the output power of other DCs in the same DC group through DC power modulation to offset the unbalanced power caused by the fault to the DC power grid, helps the multi-DC power grid to quickly restore stability after the fault, can greatly reduce the amount of power cuts during stable control, and can reduce the impact on daily production and life.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a method for stabilizing a multi-DC sending-end power grid, comprising the following steps: Obtain the name of the faulty DC line and the DC output power lost by the faulty DC line; The acquired faulty DC line name and the DC output power lost by the faulty DC line due to the fault are input into an intelligent agent configured in a DC group where the faulty DC line is located. The emergency power adjustment amount of other DC lines in the DC group where the faulty DC line is located is obtained through the intelligent agent. The unbalanced power caused by the fault is balanced through the emergency power adjustment amount of other DC lines to realize stable control of multi-DC sending-end power grids.

[0007] A further improvement of the present invention is that in the step of inputting the acquired faulty DC line name and the DC output power lost by the faulty DC line into the intelligent agent configured in the DC group where the faulty DC line is located, the DC group acquisition step includes: Obtain the topological relationship between nodes and grid parameters of the multi-DC sending-end power grid, and establish a power system simulation model; Based on the power system simulation model, perform power flow calculations on the multi-DC sending-end power grid under various operation modes, and obtain multiple feasible power flow operation modes; Based on the obtained multiple feasible power flow operation modes, short-circuit calculation is carried out on the multi-DC sending-end power grid to obtain the interaction factors between DC lines in the multi-DC sending-end power grid; cluster analysis is carried out based on the interaction factors between DC lines to obtain the DC group clustering results considering multiple DC interactions.

[0008] A further improvement of the present invention is that The calculation expression of the interaction factor is: ; In the formula, Indicates DC line , The interaction factors between Indicates that the DC line The DC line caused by the inductive load connected to the commutation bus The voltage drop; Indicates that the DC line The DC line caused by the inductive load connected to the commutation bus Adjacent DC lines The voltage drop; In the step of performing cluster analysis based on the interaction factor between DC lines, if the interaction factor between two DC lines is greater than a preset threshold, the two DC lines belong to the same DC group.

[0009] A further improvement of the present invention is that in the step of inputting the acquired faulty DC line name and the DC output power lost by the faulty DC line into an intelligent agent configured in a DC group where the faulty DC line is located, the step of acquiring the intelligent agent comprises: Initialize the agent and simulation environment, and set training parameters; In each round of training, each decision-making step includes: performing transient stability simulation in the simulation environment to obtain the name of the faulty DC line and the DC output power lost by the faulty DC line due to the fault; inputting the state observation into the intelligent agent, and the neural network returns the emergency power adjustment of the faulty DC line and other DC lines in the DC group; inputting the action into the power system environment, obtaining the reward function and storing it in the experience pool; using the state, action and reward, updating the policy network and value network through the reinforcement learning algorithm, and updating the state observation value; After the training is completed, the agent model is saved and the obtained agent model representation is: ; In the formula, is the power variation of each DC line in the DC group during DC power modulation; It is the DC output power lost due to faults in each DC line in the DC group.

[0010] A further improvement of the present invention is that in the step of inputting the action into the power system environment, obtaining the reward function and storing it in the experience pool, the reward function is calculated based on the power system state observation quantity and the control quantity, and is expressed as: ; In the formula, represents the reward function; For the The power angle of the generator; is the voltage deviation of each node in the system; is the system frequency deviation; is the number of DC lines in the DC group; For the The emergency power adjustment of the DC line during the DC power modulation process; For the Rated output power of the DC line.

[0011] A further improvement of the present invention is that, in the step of inputting the acquired faulty DC line name and the DC output power lost by the faulty DC line due to the fault into an intelligent agent configured in a DC group where the faulty DC line is located, obtaining the emergency power adjustment amount of other DC lines in the DC group where the faulty DC line is located through the intelligent agent, and balancing the unbalanced power caused by the fault through the emergency power adjustment amount of other DC lines to realize the stable control of the multi-DC sending-end power grid, The name of the faulty DC line and the DC output power loss of the faulty DC line due to the fault are expressed as a length of An array of , each value in the array represents a DC line, the corresponding value of a DC line without a fault is 0, and the corresponding value of a faulty DC line is the DC output power lost due to the fault; is the number of DC lines in the DC group; The emergency power adjustment amount of other DC lines in the DC group where the faulty DC line is located is expressed as a length of An array of , where each value in the array represents a DC line. The corresponding value of a DC line without a fault is the emergency power adjustment amount given by the agent, and the corresponding value of a faulty DC line is 0; Among them, when If a DC line fails, the DC group The emergency power adjustment of the DC line is expressed as: ; In the formula, The first The emergency power adjustment of the DC lines, The DC output power lost due to faults in each DC line in the DC group; is the maximum value of the adjustable power of each DC line in the DC group; represents the agent trained by reinforcement learning; Output array for the agent elements.

[0012] A further improvement of the present invention is that the maximum value constraint of the adjustable power of each DC line in the DC group is that after DC power modulation of other DC lines in the same DC group, the output power of the modulated DC line is less than or equal to 120% of the rated output power, and the voltage, power angle and frequency in the vicinity of the modulated DC line are safe.

[0013] A second aspect of the present invention provides a multi-DC sending-end power grid stability control system, comprising: A data acquisition module, used to obtain the name of the faulty DC line and the DC output power lost by the faulty DC line due to the fault; The adjustment control module is used to input the acquired faulty DC line name and the DC output power lost by the faulty DC line due to the fault into an intelligent agent configured in the DC group where the faulty DC line is located, and obtain the emergency power adjustment amount of other DC lines in the DC group where the faulty DC line is located through the intelligent agent. The unbalanced power caused by the fault is balanced through the emergency power adjustment amount of other DC lines to achieve stable control of multi-DC sending-end power grids.

[0014] In a third aspect of the present invention, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a multi-DC sending-end power grid stability control method as described in any one of the first aspect of the present invention is implemented.

[0015] According to a fourth aspect of the present invention, a non-transitory computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, a multi-DC sending-end power grid stability control method as described in any one of the first aspects of the present invention is implemented.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The present invention discloses a method for stabilizing a DC power grid at the sending end, which is a method for stabilizing a DC power grid at the sending end that takes into account the interaction of multiple DCs. It makes full use of the interaction and coupling characteristics between multiple DCs, and uses DC power modulation to urgently adjust the sending power of other DCs in the same DC group to offset the unbalanced power caused by the fault. It can help the DC power grid at the sending end after the fault to restore stability, and at the same time can greatly reduce the amount of machine cutting during stabilization, and can reduce the impact on daily production and life. The proposal and application of the method of the present invention promotes the development of the power system in the direction of intelligence and automation, and provides strong technical support for the construction of new power systems.

[0017] In the present invention, data of multiple DC sending-end power grids are obtained, and power flow calculations are performed to obtain a variety of feasible power flow operation modes. On this basis, short-circuit calculations are performed to obtain interaction factors between DC lines in the multiple DC sending-end power grids. Based on the interaction factors between DC lines, accurate DC group clustering results can be obtained by clustering.

[0018] In the present invention, an intelligent agent is obtained by training through a deep reinforcement learning method, which can realize the generation of DC power adjustment amount, reduce the dependence on manual adjustment, reduce labor costs, and can greatly improve the generation speed of DC power adjustment amount, which can fully avoid the conservatism of the obtained results. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below; obviously, the drawings described below are some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 It is a flow chart of a method for stabilizing a multi-DC sending-end power grid in an embodiment of the present invention.

[0021] Figure 2 It is a flowchart of a multi-DC sending-end power grid stability control method considering multi-DC interactions in a specific embodiment of the present invention.

[0022] Figure 3 It is a schematic diagram of the division results of the DC groups of the Northwest Power Grid based on the interaction factors in a specific embodiment of the present invention.

[0023] Figure 4 It is a schematic diagram comparing the frequencies of the Northwest power grid without stabilization and with stabilization through DC power modulation after the Harbin Heavy DC bipolar blocking in a specific embodiment of the present invention.

[0024] Figure 5 It is a schematic diagram of a multi-DC sending-end power grid stability control system in an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention; it is obvious that the described embodiments and technical solutions are only part of the embodiments of the present invention, not all of the embodiments.

[0026] All other embodiments obtained by those of ordinary skill in the art without creative work based on the technical solutions disclosed in the embodiments of the present invention belong to the scope of protection of the present invention. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0027] See also Figure 1 , a multi-DC sending-end power grid stability control method provided by an embodiment of the present invention comprises the following steps: Step 1: after a DC fault occurs, obtain the name of the faulty DC and the DC output power lost by the faulty DC; Step 2, input the name of the faulty DC obtained in step 1 and the DC transmission power lost by the faulty DC into the intelligent agent configured in the DC group where the faulty DC is located, and obtain the emergency power adjustment amount of other DCs in the DC group where the faulty DC is located through the intelligent agent. The unbalanced power caused by the fault is balanced through the emergency power adjustment amount of other DCs in the DC group where the faulty DC is located to achieve stable control of the multi-DC sending-end power grid.

[0028] The multi-DC power grid stability control method disclosed in the embodiment of the present invention is a new control strategy designed for the characteristics of multi-DC power transmission technology in modern power systems. Explanatory, multi-DC power transmission technology has been widely used in modern power grids due to its high efficiency, flexibility and large-capacity transmission of electric energy; however, multi-DC systems also face complex interactions and potential stability problems, especially when DC faults occur, which may cause serious power imbalance and stability threats to the power grid. The core of the method provided in the embodiment of the present invention is "considering multi-DC interactions" and "DC power modulation emergency adjustment". It not only focuses on the stable control of a single DC line, but considers the multi-DC system as a whole, making full use of the interaction and coupling characteristics between each DC line. When a DC line fails, the method of the embodiment of the present invention can quickly identify and calculate the power imbalance caused by the fault, and quickly offset the unbalanced power caused by the fault by adjusting the output power of other normally operating DC lines in the same DC group, thereby stabilizing the voltage and frequency of the multi-DC power grid. In summary, the technical solution of the embodiment of the present invention can significantly shorten the recovery time of the power grid after a fault and improve the overall stability of the system through timely power modulation; in addition, traditional stability control measures often rely on cutting off generators (cutting off machines) to balance power, but this method will cause a large impact on the power grid and may affect daily production and life. The method of the present invention greatly reduces the need for cutting off machines through DC power modulation, reducing the impact on the economy and society.

[0029] In one embodiment of the present invention, in the step of inputting the name of the faulty DC and the DC output power lost by the faulty DC obtained in step 1 into the intelligent agent configured in the DC group where the faulty DC is located, the step of obtaining the DC group of the multi-DC sending-end power grid specifically includes: Step 1), obtaining the topological relationship between the nodes of the multi-DC sending-end power grid and the data of the grid parameters, and establishing a power system simulation model; based on the power system simulation model, using the power system analysis and calculation software, the power flow calculation under various operation modes of the multi-DC sending-end power grid is performed to obtain various feasible power flow operation modes; further exemplifying the preferred technical solution, the various operation modes refer to various operation modes under different wiring modes (explanatory, including full wiring and partial line maintenance modes, etc.), different generator outputs, different DC transmission powers, and different load powers; Step 2), based on the obtained multiple feasible power flow operation modes, short-circuit calculation is performed on the multi-DC sending-end power grid to obtain the interaction factors between DCs in the multi-DC sending-end power grid, and cluster analysis is performed based on the interaction factors between DCs to obtain the DC group clustering results considering the interaction between multiple DCs.

[0030] Explanatory, the interaction factor is used to characterize the interaction between multiple DC lines in a multi-transmission system, reflecting the strength of the interaction between the converter station buses. The calculation method is as follows: ; In the formula, Represents the commutation bus , The interaction factors between Indicates that the commutation bus The commutation busbar is connected to an inductive load. voltage drop (about 1%), which will inevitably cause the commutation bus Other adjacent converter stations converter busbars The voltage drop is The ratio of the two is the commutation bus and commutation bus The interaction factors between them.

[0031] In one embodiment of the present invention, when performing cluster analysis, a clustering algorithm such as K-means or K-medoids may be used to divide multiple DC lines into several DC groups; wherein, when the interaction factor between two DC lines is greater than a certain threshold, they are generally considered to belong to the same DC group, which is expressed as, ; In the formula, For DC lines With DC line The interaction factors between is the critical value for classifying two DC lines into the same group.

[0032] In the preferred scheme of the embodiment of the present invention, by deeply analyzing the electrical connections and dynamic responses between multiple DC systems, the mutual influence between different DC lines can be identified and quantified; using these coupling characteristics, how changes in one DC line affect other lines can be considered in the control strategy, thereby achieving more refined and coordinated control.

[0033] In one embodiment of the present invention, in the step of inputting the name of the faulty DC and the DC output power lost by the faulty DC obtained in step 1 into an intelligent agent configured in the DC group where the faulty DC is located, the intelligent agent acquisition step includes: Initialize the agent and simulation environment, and set the training parameters; in each round of training, the steps for making decisions at each step include: Step (1), firstly, perform transient stability simulation in a simulation environment to obtain the name of the fault DC and the DC output power lost due to the fault; Step (2), input the state observation into the intelligent agent, and the neural network returns the emergency power adjustment of other DCs in the same DC group; Step (3), inputting the action into the power system environment, obtaining the reward function, and storing it in the experience pool; Step (4), using the state, action and reward, through the reinforcement learning algorithm, update the policy network and value network, and update the state observation value; After training, save the agent model. The obtained agent model can be represented as: ; In the formula, is the power change of each DC line in the DC group during DC power modulation, It is the DC output power lost due to faults in each DC line in the DC group.

[0034] In the embodiment of the present invention, the reward function refers to a reward function calculated according to the power system state observation quantity and the control quantity. There are different calculation methods according to the transient stability of the power system after control, which are as follows: If the power angle stability problem of the power system still exists after control, the reward function is a large negative value. The reward function is taken as the inverse of the absolute value of the maximum value of the slope of the power angle curve of each generator. The smaller the reward function, the greater the instability of the power system, which is expressed as: ; In the formula, For the The power angle of the generator, , is the total number of generators.

[0035] If there is no power system power angle stability problem after control, but there is voltage or frequency insecurity, the reward function is taken as the inverse of the maximum value of the voltage deviation and frequency deviation, expressed as: ; In the formula, is the voltage deviation of each node in the system, which is an array; is the system frequency deviation.

[0036] If there is no power angle stability problem and voltage and frequency insecurity problem after control, the control effect and control cost are considered comprehensively when designing the reward function, which can be expressed as: ; In the formula, is the number of DC lines in the DC group; For the The emergency power adjustment amount of the DC line during the DC power modulation process; For the Rated output power of the DC line.

[0037] In summary, the reward function can be expressed as follows:

[0038] In the preferred technical solution of the embodiment of the present invention, the generation of DC power adjustment amount by an intelligent agent trained by a deep reinforcement learning method can reduce the dependence on manual adjustment, reduce labor costs, and greatly improve the generation speed of DC power adjustment amount, and can fully avoid the conservatism of the obtained results. The proposal and application of the method of the present invention can promote the development of power systems in the direction of intelligence and automation, and can provide strong technical support for the construction of new power systems.

[0039] In one embodiment of the present invention, the name of the faulty DC obtained in step 1 and the DC output power lost by the faulty DC due to the fault are input into an intelligent agent configured in the DC group where the faulty DC is located, and in the step of obtaining the emergency power adjustment amount of other DCs in the DC group where the faulty DC is located through the intelligent agent, The name of the faulty DC and the DC output power lost due to the fault refer to: DC, the faulty DC and the DC output power lost due to the fault are unified by a DC with a length of An array representation of , where each value represents a DC line. For a DC line without a fault, its corresponding value is 0, and for a faulty line, its corresponding value is the DC transmission power lost due to the fault; The emergency power adjustment of other DC in the DC group refers to the assumption that there are If there is a DC line, the emergency power adjustment of other DC lines in the same DC group is unified by a line with a length of , where each value represents a DC line. For a DC line without a fault, the corresponding value is the emergency power adjustment given by the agent. For a faulty line, the corresponding value should be 0. Explanatory, the emergency power adjustment of other DC lines in the same DC group is given by the agent obtained by offline training, and is subject to the maximum value of the DC adjustable power. Assuming that there are DC lines, If a DC line fails, the DC group The emergency power adjustment of the DC line is expressed as: ; In the formula, The first The emergency power adjustment of the DC lines, The DC output power lost due to faults in each DC line in the DC group; is the maximum value of the adjustable power of each DC line in the DC group, Represents the agent trained by reinforcement learning, and its output is an array; The output array is elements.

[0040] In one embodiment of the present invention, the step of obtaining the maximum value constraint of the DC adjustable power includes: Based on various power flow operation modes, transient simulation is performed on each DC line under various DC fault scenarios to screen out transient instability and frequency or voltage insecurity samples; For unstable and unsafe samples, consider using DC power modulation to urgently adjust the output power of other DC lines in the same DC group to offset the unbalanced power caused by the fault on the sending-end system, and determine the maximum adjustable power of each DC line; in, Various DC fault types include DC unipolar blocking, DC bipolar blocking, DC two full-voltage restart success, DC two full-voltage restart failure, DC blocking after four consecutive commutation failures, DC 12 consecutive commutation failures and other common DC faults; Determining the maximum value of the adjustable power of each DC line means that after DC power modulation is performed on other DC lines in the same DC group, the output power of the modulated DC line cannot be higher than 120% of the rated output power, and the voltage, power angle and frequency of the modulated DC line near the DC line need to be safe, which can be expressed as: ; In the formula, is the output power of the modulated DC after modulation, is the output power of the modulated DC before modulation, is the maximum value of the modulated DC adjustable power, is the rated output power of the modulated DC.

[0041] See also Figures 2 to 4 In a specific embodiment of the present invention, a method for stabilizing a multi-DC power grid sending end considering multi-DC interactions is provided, comprising the following steps: S1. Acquire data such as the topological relationship and grid parameters between nodes of a multi-DC sending-end power grid, establish a power system simulation model, and use power system analysis and calculation software to perform power flow calculations on the multi-DC sending-end power grid under various operation modes to obtain a variety of feasible power flow operation modes; wherein, a specific embodiment of the present invention selects the Northwest China regional power grid, and the calculation data adopts the basic data of the Northwest China power grid in a certain year. By using the power system analysis and calculation software PSASP (Power System Analysis Software Package), by adjusting the power flow calculation data, power flow simulation calculations are performed under different wiring modes, different generator outputs, different DC transmission powers, and different load powers to obtain a variety of feasible power flow operation modes.

[0042] S2. Based on a variety of power flow operation modes, short-circuit calculations are performed on multiple DC sending-end power grids to obtain interaction factors, and cluster analysis is performed to obtain DC group clustering results that consider multiple DC interactions; illustratively, as shown in Table 1, Table 1 shows the interaction factors between the DC line commutation buses calculated under a certain power flow operation mode.

[0043] Table 1. Interaction factors between commutation buses of various DC lines

[0044] Among them, the correspondence between the DC lines and the converter busbars is as follows: the converter busbar on the rectifier side of the "Harbin-Zhongshan DC" is the "New Hami North Converter", the converter busbar on the rectifier side of the "Jiquan DC" is the "New Changji Converter Station", the converter busbar on the rectifier side of the "Tianzhong DC" is the "New Hami Converter", the converter busbar on the rectifier side of the "Qi Shao DC" is the "Gan Qi Lian Converter", the converter busbar on the rectifier side of the "Qinghai-Henan DC" is the "Qingqing South", the converter busbar on the rectifier side of the "Lingshao DC" is the "Ninglingzhou Converter Station", the converter busbar on the rectifier side of the "Yindong DC" is the "Ningyinchuan Converter Station", and the converter busbar on the rectifier side of the "Zhaoyi DC" is the "Ninglingzhou Converter Station". The rectifier busbar on the line is "Ning Yikezhao Converter Station", the rectifier busbar on the "Ningxiang DC" is "Zhongning Converter Station", the rectifier busbar on the "Longdong DC" is "Gan Longdong Converter", the rectifier busbar on the "Debao DC" is "Shaanxi Baoji", the rectifier busbar on the "Lingbao DC" is "Shaanxi Luofu" and "Shaanxi Dongnan Suburb", the rectifier busbar on the "Shaanxi Wu DC" is "Shaanxi Bei Converter", and the rectifier busbar on the "Chaila DC" is "Qing Qaidam".

[0045] In cluster analysis, the critical value is taken by the embodiment of the present invention , the DC lines of the Northwest Power Grid can be divided into three major DC groups: Xinjiang and interconnected channel DC group, Ningxia DC group and Shaanxi DC group. Figure 3As shown; among them, Xinjiang's DC group with interconnected channels includes Harbin-Zhongli DC, Tianzhong DC, Jiquan DC, Qishao DC and Qinghai-Henan DC, Ningxia DC group includes Zhaoyi DC, Yindong DC, Ningxiang DC and Lingshao DC, Shaanxi DC group includes Longdong DC, Lingbao DC, Debao DC and Shaanwu DC.

[0046] S3. Based on various power flow operation modes, transient simulation is performed on each DC line under various DC fault scenarios to screen out transient instability and frequency or voltage insecurity samples; various DC fault types include DC unipolar blocking, DC bipolar blocking, DC two full-voltage restart success, DC two full-voltage restart failure, DC blocking after four consecutive commutation failures, DC 12 consecutive commutation failures and other common DC faults. In the embodiment of the present invention, in typical mode 1: the total output of Xinjiang power grid is 79.24424 million kilowatts, the output of thermal power units is 64.72 million kilowatts, the output of wind turbine units is 9.0865 million kilowatts, and the output of photovoltaic units is 5.43774 million kilowatts. The Tianzhong DC transmits 6.5 million kilowatts of power through the new Hami converter station, the Jiquan DC transmits 12 million kilowatts of power through the new Changji converter station, and the Harbin-Zhongqing DC transmits 8 million kilowatts of power through the new Hami North converter station; the total output of the Gansu power grid is 49.9674 million kilowatts, the output of thermal power units is 39.3059 million kilowatts, the output of wind turbine units is 8.1535 million kilowatts, and the output of photovoltaic units is 2.508 million kilowatts. The Qi Shao DC transmits 6.5 million kilowatts of power through the Gan Qilian converter station, and the Longdong DC transmits 8 million kilowatts of power through the Gan Longdong converter station; the total output of the Ningxia power grid is 41.5105 million kilowatts, the output of thermal power units is 36.131 million kilowatts, the output of wind turbine units is 299,500 kilowatts, and the output of photovoltaic units is 5.08 million kilowatts. Ningxiang DC sent out 8 million kilowatts of power through Zhongning Converter Station, Lingshao DC sent out 8 million kilowatts of power through Ninglingzhou Converter Station, Yindong DC sent out 4 million kilowatts of power through Ninglingzhou Converter Station, and Zhaoyi DC sent out 9 million kilowatts of power through Ningyikezhao Converter Station. In this typical mode, when Harbin-Zhongzhong DC, Longdong DC, Ningxiang DC, Tianzhong DC, Jiquan DC, and Zhaoyi DC had bipolar blocking faults, the Northwest Power Grid all experienced high-frequency phenomena to varying degrees, and overvoltage problems occurred when Ningxiang DC and Longdong DC failed to restart at full voltage twice.

[0047] S4. For unstable and unsafe samples, consider using DC power modulation to urgently adjust the output power of other DC lines in the same DC group to offset the unbalanced power caused by the fault to the sending-end system, and determine the maximum value of the adjustable power of each DC line; in this embodiment of the present invention, for the five DC lines in the Xinjiang and interconnected channel DC group under typical mode 1, it can be determined that the maximum values ​​of the adjustable power of Harbin-Chongqing DC, Tianzhong DC, Jiquan DC, Qishao DC and Qinghai-Henan DC are 1.6 million kilowatts, 2 million kilowatts, 2.4 million kilowatts, 2 million kilowatts and 1.6 million kilowatts respectively.

[0048] S5. Configure a neural network as an intelligent agent for each DC group, train the intelligent agent through reinforcement learning method, save the intelligent agent model after the training reaches convergence, and put it into online application after passing the test.

[0049] S6. In the online application, after a DC fault occurs, the name of the faulty DC and the DC output power lost due to the fault are obtained, and the emergency power adjustment amount of other DCs in the same DC group is input into the intelligent agent to perform emergency power modulation control on the DC. Among them, in the embodiment of the present invention, under the typical mode 1, when the DC bipolar locking fault occurs in the Harbin Heavy DC, the Harbin Heavy DC loses 8 million kilowatts, so the array representing the faulty DC and the DC output power lost due to the fault is , the five numbers in the array correspond to Harbin-Zhongshan DC, Tianzhong DC, Jiquan DC, Qishao DC and Qingyu DC respectively; the array is input into the offline trained intelligent agent, which gives the emergency power adjustment amount of other DCs in the same DC group; the emergency power adjustment amount of other DCs in the same DC group given by the intelligent agent is The five numbers in the array correspond to the emergency power adjustment amounts of Harbin-Zhongshan DC, Tianzhong DC, Jiquan DC, Qishao DC and Qinghe DC.

[0050] Harbin-Zhongshan DC is a fault line, and the DC power adjustment is 0. The emergency power adjustment of Tianzhong DC, Jiquan DC, Qishao DC and Qingyu DC are 1.854 million kW, 550,000 kW, 3 million kW and 1.414 million kW respectively. The frequency change comparison of the Northwest Power Grid without stabilization and after emergency DC power modulation and stabilization is shown in the attached figure. Figure 4 As shown, it can be seen that after making full use of the interactive coupling characteristics between multiple DCs and urgently adjusting the output power of other DCs in the same DC group through DC power modulation to offset the unbalanced power caused by the fault to the sending-end system, the high-frequency problem of the Northwest Power Grid has been significantly alleviated, and the frequency has returned to the regulation requirements.

[0051] In summary, the embodiment of the present invention discloses a DC sending-end power grid stability control method considering the interaction of multiple DC lines, including: obtaining data such as topological relationships and grid parameters between nodes of the multiple DC sending-end power grids, performing power flow calculations on the multiple DC sending-end power grids under multiple operation modes, and obtaining multiple feasible power flow operation modes; performing short-circuit calculations on the multiple DC sending-end power grids to obtain interaction factors of multiple DC systems, and performing cluster analysis to obtain DC group clustering results considering the interaction of multiple DC lines; performing transient simulations on each DC line under multiple DC fault scenarios to screen out transient instability and frequency or voltage insecurity samples; and performing transient simulations on the instability and For unsafe samples, consider using DC power modulation to urgently adjust the output power of other DCs in the same DC group to offset the unbalanced power caused by the fault to the sending end system, and determine the maximum value of the adjustable power of each DC; configure a neural network as an intelligent agent for each DC group, train the intelligent agent through a reinforcement learning method, save the intelligent agent model after the training reaches convergence, and put it into online application after the test passes; in the online application, after a DC fault occurs, obtain the name of the faulty DC and the DC output power lost due to the fault, input it into the intelligent agent to obtain the emergency power adjustment amount of other DCs in the same DC group, and perform emergency power modulation control on the DC. The technical solution provided by the embodiment of the present invention makes full use of the interactive coupling characteristics between multiple DCs, and through DC power modulation, urgently adjusts the output power of other DCs in the same DC group to offset the unbalanced power caused by the fault to the sending end system, which can help the DC sending end power grid after the fault to restore stability, and at the same time greatly reduces the amount of machine cutting during stabilization control. The generation of DC power adjustment amount by the intelligent agent obtained through deep reinforcement learning method training greatly improves the generation speed of DC power adjustment amount, and fully avoids the conservatism of the obtained results.

[0052] The following are device embodiments of the present invention, which can be used to implement the method embodiments of the present invention. For details not disclosed in the device embodiments, please refer to the method embodiments of the present invention.

[0053] See also Figure 5 In an embodiment of the present invention, a multi-DC sending-end power grid stability control system is provided, comprising: A data acquisition module, used to obtain the name of the faulty DC line and the DC output power lost by the faulty DC line due to the fault; The adjustment control module is used to input the acquired faulty DC line name and the DC output power lost by the faulty DC line due to the fault into an intelligent agent configured in the DC group where the faulty DC line is located, and obtain the emergency power adjustment amount of other DC lines in the DC group where the faulty DC line is located through the intelligent agent. The unbalanced power caused by the fault is balanced through the emergency power adjustment amount of other DC lines to achieve stable control of multi-DC sending-end power grids.

[0054] In one embodiment of the present invention, a computer device is provided, the computer device comprising a processor and a memory, the memory being used to store a computer program, the computer program comprising program instructions, and the processor being used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions in a computer storage medium to implement a corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used to perform the operation of the multi-DC sending-end power grid stability control method.

[0055] In one embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It is understandable that the computer-readable storage medium here can include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM (Random Access Memory) memory, or a non-volatile memory (non-volatile memory), such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the multi-DC sending-end power grid stability control method in the above embodiment.

[0056] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, optical storage, etc.) containing computer-usable program codes.

[0057] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0058] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0059] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for stabilizing a multi-DC power grid, characterized in that: The following steps are involved: Obtain the name of the faulty DC line and the DC output power lost by the faulty DC line; The acquired faulty DC line name and the DC output power lost by the faulty DC line due to the fault are input into an intelligent agent configured in a DC group where the faulty DC line is located. The emergency power adjustment amount of other DC lines in the DC group where the faulty DC line is located is obtained through the intelligent agent. The unbalanced power caused by the fault is balanced through the emergency power adjustment amount of other DC lines to realize stable control of multi-DC sending-end power grids.

2. A method for stabilizing a multi-DC power grid according to claim 1, characterized in that: In the step of inputting the acquired faulty DC line name and the DC output power lost by the faulty DC line into the intelligent agent configured in the DC group where the faulty DC line is located, the DC group acquisition step includes: Obtain the topological relationship between nodes and grid parameters of the multi-DC sending-end power grid, and establish a power system simulation model; Based on the power system simulation model, perform power flow calculations on the multi-DC sending-end power grid under various operation modes, and obtain multiple feasible power flow operation modes; Based on the obtained multiple feasible power flow operation modes, short-circuit calculation is carried out on the multi-DC sending-end power grid to obtain the interaction factors between DC lines in the multi-DC sending-end power grid; cluster analysis is carried out based on the interaction factors between DC lines to obtain the DC group clustering results considering multiple DC interactions.

3. A method for controlling the stability of a multi-DC sending-end power grid according to claim 2, characterized in that: The calculation expression of the interaction factor is: ； In the formula, Indicates DC line , The interaction factors between Indicates that the DC line The DC line caused by the inductive load connected to the commutation bus The voltage drop; Indicates that the DC line The DC line caused by the inductive load connected to the commutation bus Adjacent DC lines The voltage drop; In the step of performing cluster analysis based on the interaction factor between DC lines, if the interaction factor between two DC lines is greater than a preset threshold, the two DC lines belong to the same DC group.

4. A method for controlling the stability of a multi-DC sending-end power grid according to claim 1, characterized in that: In the step of inputting the acquired faulty DC line name and the DC output power lost by the faulty DC line into the intelligent agent configured in the DC group where the faulty DC line is located, the intelligent agent acquisition step includes: Initialize the agent and simulation environment, and set training parameters; In each round of training, each decision-making step includes: performing transient stability simulation in the simulation environment to obtain the name of the faulty DC line and the DC output power lost by the faulty DC line due to the fault; inputting the state observation into the intelligent agent, and the neural network returns the emergency power adjustment of the faulty DC line and other DC lines in the DC group; inputting the action into the power system environment, obtaining the reward function and storing it in the experience pool; using the state, action and reward, updating the policy network and value network through the reinforcement learning algorithm, and updating the state observation value; After the training is completed, the agent model is saved and the obtained agent model representation is: ; In the formula, is the power variation of each DC line in the DC group during DC power modulation; It is the DC output power lost due to faults in each DC line in the DC group.

5. A method for controlling the stability of a multi-DC sending-end power grid according to claim 4, characterized in that: In the step of inputting the action into the power system environment, obtaining the reward function and storing it in the experience pool, the reward function is calculated based on the power system state observation quantity and control quantity, and is expressed as: ； In the formula, represents the reward function; For the The power angle of the generator; is the voltage deviation of each node in the system; is the system frequency deviation; is the number of DC lines in the DC group; For the The emergency power adjustment of the DC line during the DC power modulation process; For the Rated output power of the DC line.

6. The method for controlling the stability of a multi-DC sending-end power grid according to claim 1, characterized in that: In the step of inputting the acquired faulty DC line name and the DC output power lost by the faulty DC line due to the fault into an intelligent agent configured in the DC group where the faulty DC line is located, obtaining the emergency power adjustment amount of other DC lines in the DC group where the faulty DC line is located through the intelligent agent, and balancing the unbalanced power caused by the fault through the emergency power adjustment amount of other DC lines to realize the stable control of the multi-DC sending-end power grid, The name of the faulty DC line and the DC output power loss of the faulty DC line due to the fault are expressed as a length of An array of , each value in the array represents a DC line, the corresponding value of a DC line without a fault is 0, and the corresponding value of a faulty DC line is the DC output power lost due to the fault; is the number of DC lines in the DC group; The emergency power adjustment amount of other DC lines in the DC group where the faulty DC line is located is expressed as a length of An array of , where each value in the array represents a DC line. The corresponding value of a DC line without a fault is the emergency power adjustment amount given by the agent, and the corresponding value of a faulty DC line is 0; Among them, when If a DC line fails, the DC group The emergency power adjustment of the DC line is expressed as: ； In the formula, The first The emergency power adjustment of the DC lines, The DC output power lost due to faults in each DC line in the DC group; is the maximum value of the adjustable power of each DC line in the DC group; represents the agent trained by reinforcement learning; Output array for the agent elements.

7. A method for controlling the stability of a multi-DC sending-end power grid according to claim 6, characterized in that: The maximum value constraint of the adjustable power of each DC line in the DC group is that, after DC power modulation of other DC lines in the same DC group, the output power of the modulated DC line is less than or equal to 120% of the rated output power, and the voltage, power angle and frequency in the vicinity of the modulated DC line are safe.

8. A multi-DC sending-end power grid stability control system, characterized in that: include: A data acquisition module, used to obtain the name of the faulty DC line and the DC output power lost by the faulty DC line due to the fault; The adjustment control module is used to input the acquired faulty DC line name and the DC output power lost by the faulty DC line due to the fault into an intelligent agent configured in the DC group where the faulty DC line is located, and obtain the emergency power adjustment amount of other DC lines in the DC group where the faulty DC line is located through the intelligent agent. The unbalanced power caused by the fault is balanced through the emergency power adjustment amount of other DC lines to achieve stable control of multi-DC sending-end power grids.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the multi-DC sending-end power grid stability control method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the multi-DC sending-end power grid stability control method according to any one of claims 1 to 7 is implemented.

Images