Embedded DC sending end area new energy station control method
By optimizing the reactive current ratio coefficient of new energy power plants and utilizing deep reinforcement learning algorithms, the problem of insufficient voltage support in the embedded DC transmission area was solved, thereby improving the voltage stability of the converter bus node and the stable operation of new energy units.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-25
- Publication Date
- 2026-03-10
AI Technical Summary
After a high proportion of new energy sources are connected to the grid, the voltage support capability of the embedded DC transmission end area is weak, resulting in unstable voltage at the converter bus node, especially in the event of a fault, making it difficult to maintain voltage stability.
The reactive current ratio coefficient of new energy power plants is optimized by using deep reinforcement learning algorithms. Objective functions and constraints are constructed, reference new energy power plants are selected, and control parameters are determined by using the TD3 deep reinforcement learning algorithm to optimize the reactive current ratio coefficient of new energy power plants to support the converter bus node voltage of embedded DC rectifier stations.
It improves the insufficient reactive voltage support in the embedded DC sending-end area, ensures the stable operation of new energy units, avoids overvoltage and repeated low voltage ride-through, and enhances the stability of converter bus node voltage.
Smart Images
Figure CN121643006A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to power system safety and stability analysis and control technology, and in particular discloses an embedded DC sending-end regional new energy control method. Background Technology
[0002] With the increasing proportion of new energy sources in the energy structure and the continuous expansion of AC power grids, there is an urgent need to improve the operational flexibility and cross-sectional transmission capacity of provincial / regional AC power grids. However, if AC transmission technology is used to enhance these capabilities, AC power flow is difficult to control effectively, and the overall transmission capacity improvement is limited due to factors such as short-circuit current. Therefore, the Jiangsu power grid has proposed "embedding" DC transmission lines within the provincial AC power grid to improve cross-sectional transmission capacity and promote the consumption of new energy sources.
[0003] Because the sending-end rectifier station and receiving-end inverter station of an embedded DC system are located in the same provincial AC power grid, with strong electrical distance and coupling, when a fault occurs in the sending-end area causing a voltage drop at the converter bus node of the rectifier station, not only will improper interaction of the internal control strategies of the DC system lead to commutation failure on the inverter side, but the AC fault will also directly cause a voltage drop at the converter bus node of the inverter station. Therefore, maintaining voltage stability in the sending-end area of the embedded DC system, especially at the converter bus node of the rectifier station, is crucial for the safe and stable operation of the provincial AC / DC hybrid power grid. However, with the high proportion of renewable energy grid connection and the construction of new power systems, there may be a large number of renewable energy sources in the sending-end area of the embedded DC system. Their reactive power voltage support capacity is weaker than that of traditional generators. How to support the voltage of the converter bus node of the embedded DC rectifier station by optimizing the control parameters of renewable energy sources and taking into account the stability constraints of the renewable energy sources themselves is a technical problem that urgently needs to be solved. Summary of the Invention
[0004] This application provides a new energy control method for embedded DC transmission areas, which solves the problem that the voltage support for embedded DC transmission areas is weak when a high proportion of new energy grid connection is connected in current research.
[0005] This application provides an embedded DC-to-end regional renewable energy power station control method, including:
[0006] The reference new energy power stations are determined based on the voltage support capability of each new energy power station;
[0007] Construct the objective function and constraints for optimizing the control parameters of each of the aforementioned new energy power stations;
[0008] Using a deep reinforcement learning algorithm, the control parameter values are optimized and determined for the reference new energy power station based on the objective function and the constraints.
[0009] Optionally, a reference new energy power station is determined based on the voltage support capability of each new energy power station for the converter bus node of the embedded DC rectifier station, including:
[0010] Calculate the voltage support capability of each new energy power station for the converter bus node of the embedded DC rectifier station;
[0011] The new energy power stations are sorted from largest to smallest according to their reactive power and voltage support capabilities, and the new energy power stations with the highest rankings are selected as the reference new energy power stations.
[0012] Optionally, the voltage support capability of each new energy power station for the converter bus node of the embedded DC rectifier station is calculated, including:
[0013] According to the PQ decomposition method for power flow calculation, the sub-matrices K and J are obtained from the node voltage and node admittance matrices based on the first formula, which is as follows:
[0014]
[0015] Wherein, ΔP is the change in active power, ΔQ is the change in reactive power, Δθ is the change in the phase angle of the node voltage, and ΔV is the change in the magnitude of the node voltage.
[0016] The first formula is converted into the second formula to obtain the voltage change of the converter bus node of the embedded DC rectifier station caused by the reactive power of the new energy power station.
[0017] The second formula is converted into the third formula to obtain the reactive power capacity of the new energy power station.
[0018] Optionally, the first formula is converted into a second formula to obtain the voltage change at the converter bus node of the embedded DC rectifier station caused by the reactive power of the new energy power station, including:
[0019] Convert the first formula into a second formula, which is: ,in, Let J be the element in the i-th row and j-th column of the inverse matrix of submatrix J.
[0020] Optionally, the second formula can be converted into a third formula, wherein the third formula is: S j Q represents the reactive power capacity of the renewable energy power station at node j. j_pu This represents the per-unit value of the reactive power capacity of the new energy power station at node j.
[0021] Optionally, the renewable energy power stations are sorted from highest to lowest according to their reactive power and voltage support capabilities, and a predetermined proportion of the renewable energy power stations with the highest ranking are selected as the reference renewable energy power stations, including:
[0022] The reactive power capacity of each of the aforementioned new energy power stations is sorted from largest to smallest, and the top 30% of the new energy power stations in terms of reactive power capacity are selected as the reference new energy power stations.
[0023] Optionally, an objective function for optimizing the control parameters of each of the aforementioned new energy power stations is constructed:
[0024]
[0025] Where F represents the objective function value, t0 represents the start time of the fault simulation, and V i0 V represents the voltage at node i of the converter bus in the embedded DC rectifier station before a fault. i (t) represents the voltage after a fault occurs at node i of the converter bus in the embedded DC rectifier station;
[0026] Based on the objective function, multiple constraints on the grid connection voltage of the new energy power station under various operating conditions are determined.
[0027] Optionally, based on the objective function, several constraints on the grid connection voltage of the new energy power station are determined, including:
[0028] Based on the objective function, after the fault is cleared, to prevent the equipment in the new energy power station from disconnecting from the grid, an overvoltage constraint condition is determined for the grid connection point of the new energy power station. The overvoltage constraint condition is as follows: , among which, U w T is the voltage at the grid connection point of the new energy power station. Uw>1.1 This indicates the time during which the grid connection point voltage exceeds 1.1 pu;
[0029] Based on the objective function, after the fault is cleared, the low-voltage constraint conditions for the repeated grid connections of the new energy power station are determined: n represents the number of times the voltage at the grid connection point of the new energy power station passes through 0.9 pu in a decreasing trend.
[0030] Optionally, the voltage at the grid connection point of the new energy power station decreases by 0.9 pu, and this is determined by the following method:
[0031]
[0032] in, This represents the voltage at the new energy grid connection point during the k-th data read after the fault. This indicates the voltage at the new energy grid connection point during the (k+1)th data reading after the fault, with a data reading interval of 0.01s.
[0033] Optionally, a deep reinforcement learning algorithm is used to optimize and determine the control parameter values for the reference renewable energy power station based on the objective function and the constraints, including:
[0034] The reactive current proportionality coefficient k in the low-voltage ride-through control of the new energy power station is optimized using the TD3 deep reinforcement learning algorithm. q The correction quantity is mapped to an action, which involves setting the voltage status of the converter bus node of the embedded DC rectifier station after the fault and the current reactive current ratio coefficient k of the new energy source. q The value of is mapped to the state, and the reward function is constructed using the following formula:
[0035]
[0036] Wherein, F0 is the objective function value under any given initial value of the reactive current ratio coefficient of the new energy power station.
[0037] The technical solution provided in this application can improve the problem of insufficient reactive voltage support in the embedded DC sending-end area by optimizing the reactive current ratio coefficient in the low voltage ride-through control of new energy power plants. At the same time, it takes into account the overvoltage constraints and repeated low voltage ride-through constraints to ensure the stable operation of new energy units, and can effectively support the stability of the converter bus node voltage of the embedded DC rectifier station. Attached Figure Description
[0038] Figure 1 This is a flowchart illustrating an embedded DC-to-end regional new energy power station control method provided in an embodiment of this application;
[0039] Figure 2 yes Figure 1 The flowchart of step S1 in the embedded DC transmission area new energy power station control method is shown.
[0040] Figure 3 yes Figure 2 The flowchart of step S11 in the embedded DC transmission end area new energy power station control method is shown.
[0041] Figure 4 yes Figure 1 The flowchart of step S2 in the embedded DC transmission area new energy power station control method is shown.
[0042] Figure 5 Figure 4 The flowchart of step S22 in the embedded DC transmission area new energy power station control method is shown.
[0043] Figure 6 This is a schematic diagram of an embedded DC power supply and its sending-end renewable energy provided in a specific embodiment of this application;
[0044] Figure 7 This is a comparison of the simulation curves of the load node voltage in the embedded DC sending-end area after a fault, in a specific embodiment of this application.
[0045] Figure 8 This is a comparison of the simulation curves of the converter bus node voltage of the embedded DC rectifier station after a fault in a specific embodiment of this application.
[0046] Figure 9 This is a comparison of the simulated voltage curves of the new energy grid connection point before and after taking into account constraints in a specific embodiment of this application. Detailed Implementation
[0047] The present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0048] Figure 1 This is a flowchart illustrating an embedded DC-to-end regional renewable energy power station control method provided in an embodiment of this application. Figure 1 As shown in this embodiment, the embedded DC transmission-end area new energy power station control method includes:
[0049] S1: The reference new energy power stations are determined based on the voltage support capability of each new energy power station.
[0050] Figure 2 It shows Figure 1 The flowchart illustrates step S1 in the embedded DC transmission-end area new energy power station control method shown. (See attached diagram.) Figure 2 As shown, in the new energy power station control method provided in this embodiment, step S1 includes:
[0051] S11: Calculate the voltage support capability of each new energy power station for the converter bus node of the embedded DC rectifier station.
[0052] Figure 3 It shows Figure 2 The flowchart shown is a schematic diagram of step S11 in the embedded DC transmission-end area new energy power station control method. (See attached diagram.) Figure 2 As shown, in the new energy power station control method provided in this embodiment, step S1 includes:
[0053] S111: According to the PQ decomposition method of power flow calculation, the sub-matrices K and J are obtained from the node voltage and node admittance matrices based on the first formula, which is as follows:
[0054]
[0055] Wherein, ΔP is the change in active power, ΔQ is the change in reactive power, Δθ is the change in the phase angle of the node voltage, and ΔV is the change in the magnitude of the node voltage.
[0056] S112: Convert the first formula into the second formula to obtain the voltage change of the converter bus node of the embedded DC rectifier station caused by the reactive power of the new energy power station.
[0057] Specifically, step S112 includes: converting the first formula into a second formula, the second formula being: ,in, Let J be the element in the i-th row and j-th column of the inverse matrix of submatrix J. This reflects the impact of the reactive power of the renewable energy power plant at node j on the voltage support of node i of the embedded DC rectifier station converter bus, i.e., the degree of reactive power-voltage interaction at the node. In a specific embodiment, the change in reactive power ΔQ at node j can be calculated based on the first formula. j The voltage amplitude change ΔV at node i is measured. i You can get .
[0058] S113: Convert the second formula into the third formula to obtain the reactive power capacity of the new energy power station.
[0059] Specifically, step S113 includes: converting the first formula into a third formula, the third formula being: S j Q represents the reactive power capacity of the renewable energy power station at node j. j_pu This represents the per-unit value of the reactive power capacity of the new energy power station at node j.
[0060] S12: Sort all new energy power stations in descending order of reactive power and voltage support capacity, and select the new energy power stations with the highest rankings according to a preset ratio as reference new energy power stations.
[0061] Specifically, step S12 includes: sorting the reactive power capacity of each renewable energy power station from largest to smallest, and selecting the top 30% of renewable energy power stations as reference renewable energy power stations. For example, if there are n renewable energy power stations, then based on the reactive power capacity S1~S at each node calculated in step S113... n , S1~S n Renewable energy power plants were selected as reference power plants, ranking them from largest to smallest in reactive power capacity. This selection was made because renewable energy power plants with larger reactive power capacity have a greater impact on grid stability.
[0062] S2: Construct the objective function and constraints for optimizing the control parameters of each new energy power station.
[0063] Figure 4 It shows Figure 1The flowchart shown is a schematic diagram of step S2 in the embedded DC transmission-end area new energy power station control method. (See attached diagram.) Figure 4 As shown, in the new energy power station control method provided in this embodiment, step S2 includes:
[0064] S21: Construct the objective function for optimizing the control parameters of each new energy power station:
[0065]
[0066] Where F represents the objective function value, t0 represents the start time of the fault simulation, and V i0 V represents the voltage at node i of the converter bus in the embedded DC rectifier station before a fault. i (t) represents the voltage after a fault occurs at node i of the converter bus in the embedded DC rectifier station.
[0067] S22: Based on the objective function, determine multiple constraints on the voltage at the grid connection point of new energy power plants under various operating conditions.
[0068] Figure 5 It shows Figure 4 The flowchart shown is a schematic diagram of step S22 in the embedded DC transmission-end area new energy power station control method. (See attached diagram.) Figure 5 As shown, in the new energy power station control method provided in this embodiment, step S22 includes:
[0069] S221: Based on the objective function, after fault clearance, to prevent equipment in the new energy power station from disconnecting from the grid, determine the overvoltage constraint condition for the grid connection point of the new energy power station. The overvoltage constraint condition is as follows: , among which, U w For the grid connection voltage of new energy power plants, T Uw>1.1 This indicates the time during which the grid connection point voltage exceeds 1.1 pu.
[0070] Specifically, after the fault is cleared, due to the delay in the voltage detection and control system, the grid connection point of the new energy power station may experience overvoltage. When the grid connection point voltage of the new energy power station exceeds 1.3 pu or exceeds 1.1 pu for 10 seconds, the equipment in the new energy power station is at risk of high voltage disconnection from the grid. Therefore, it is necessary to set overvoltage constraint conditions.
[0071] S222: Based on the objective function, determine the low-voltage constraints for repeated grid connections of renewable energy power plants after fault clearance. n represents the number of times the voltage at the grid connection point of the new energy power station passes through 0.9 pu in a decreasing trend.
[0072] After the fault is cleared, the grid strength decreases, and the reactive power support of the system is insufficient. Equipment at renewable energy power plants will repeatedly experience low-voltage ride-throughs. Furthermore, improper control parameter settings can also cause control system instability, leading to repeated low-voltage ride-throughs by renewable energy sources. Specifically, if the grid connection voltage of a renewable energy power plant decreases over a period of 0.9 pu, the following method can be used to determine this:
[0073]
[0074] in, This represents the voltage at the new energy grid connection point during the k-th data read after the fault. This indicates the voltage at the new energy grid connection point during the (k+1)th data reading after the fault, with a data reading interval of 0.01s.
[0075] S3: Using a deep reinforcement learning algorithm, the control parameter values are determined by optimizing the objective function and constraints for a reference new energy power station.
[0076] Specifically, in the new energy power station control method provided in this embodiment, step S3 includes: optimizing based on the deep reinforcement learning TD3 algorithm to adjust the reactive current proportional coefficient k in the low voltage ride-through control of the new energy power station. q The correction quantity is mapped to an action, which involves setting the voltage status of the converter bus node of the embedded DC rectifier station after the fault and the current reactive current ratio coefficient k of the new energy source. q The values of are mapped to states, and the reward function is constructed using the following formula:
[0077]
[0078] Where F0 is the objective function value under any given initial value of the reactive current ratio coefficient of the new energy power station.
[0079] The new energy power station control method provided in this embodiment can improve the problem of insufficient reactive voltage support in the embedded DC sending end area by optimizing the reactive current ratio coefficient in the low voltage ride-through control of the new energy power station. At the same time, it takes into account the overvoltage constraints and repeated low voltage ride-through constraints to ensure the stable operation of the new energy units, and can effectively support the stability of the converter bus node voltage of the embedded DC rectifier station.
[0080] Figure 6 This is a schematic diagram of an embedded DC power supply and its sending-end renewable energy provided in a specific embodiment of this application. Figure 7 This is a comparison of the simulation curves of the load node voltage in the embedded DC sending-end area after a fault, as shown in a specific embodiment of this application. Figure 8 This is a comparison of the simulation curves of the converter bus node voltage of the embedded DC rectifier station after a fault in a specific embodiment of this application. Figure 9This is a comparison of the simulated voltage curves of the new energy grid connection point before and after taking into account constraints in a specific embodiment of this application.
[0081] This section describes a simple power system with embedded DC transmission based on the PSD-BPA simulation platform. The embedded DC transmission system transmits surplus power from the northern part of the same AC grid to the southern load center. Numerous renewable energy sources are distributed at the northern embedded DC transmission end. A schematic diagram is shown below. Figure 6 As shown.
[0082] As a control group that did not use the control method for new energy power stations provided in this application, after setting an AC fault, due to the insufficient reactive voltage support capacity of the new energy power station, the voltage of the loaded node in the embedded DC sending area did not recover to 0.8pu within 10s, resulting in voltage instability. The voltage recovery rate of the converter bus node of the rectifier station was slow.
[0083] After optimizing the reactive current ratio coefficient of the new energy power plant in the sending-end area using the control method provided in this application, no voltage instability occurred in the embedded DC sending-end area, and the voltage recovery rate of the rectifier station converter bus node was improved. Relevant simulation comparison curves are shown below. Figure 7 and Figure 8 As shown. Furthermore, Figure 9 Further comparison of the grid connection point voltage of new energy power plants under unconstrained and constrained conditions shows that by adopting the new energy power plant control method and overvoltage constraints and repeated low voltage ride-through constraints provided in this application, the voltage support capability of the sending-end area can be improved while ensuring the stable operation of new energy units.
[0084] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.
[0085] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented using software plus a general-purpose hardware platform, or of course, using hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.
[0086] The above embodiments are merely illustrative examples and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations. However, obvious variations or modifications derived therefrom are still within the scope of protection of this application.
Claims
1. An embedded direct current sending end regional new energy plant station control method, characterized in that, The method comprises the steps of: determining a reference new energy station according to the voltage support capability of each new energy station; constructing an objective function and a constraint condition for optimization of a control parameter of each new energy station; determining the value of the control parameter by optimization according to the objective function and the constraint condition for the reference new energy station by using a deep reinforcement learning algorithm.
2. The new energy plant control method according to claim 1, wherein, The reference new energy station is determined according to the voltage support capability of each new energy station for the embedded DC rectifier station commutation bus node, comprising: calculating the voltage support capability of each new energy station for the embedded DC rectifier station commutation bus node; sorting each new energy station according to the reactive voltage support capability from large to small, and selecting a preset proportion of the new energy stations ranked at the front as the reference new energy stations.
3. The new energy plant control method according to claim 2, characterized in that, The voltage support capability of each new energy station for the embedded DC rectifier station commutation bus node is calculated, comprising: According to the PQ decomposition method of power flow calculation, the sub-matrix K and J are obtained based on the first formula from the node voltage and the node admittance matrix, and the first formula is as follows: ； Where ΔP is the change of active power, ΔQ is the change of reactive power, Δθ is the change of the phase angle of the node voltage, and ΔV is the change of the amplitude of the node voltage. The first formula is converted into a second formula to obtain the voltage change of the embedded DC rectifier station commutation bus node caused by the reactive power of the new energy station. The second formula is converted into a third formula to obtain the reactive capacity of the new energy station.
4. The new energy plant control method according to claim 3, characterized in that, The first formula is converted into a second formula to obtain the voltage change of the embedded DC rectifier station commutation bus node caused by the reactive power of the new energy station, comprising: converting the first formula to a second formula, the second formula being: wherein, is an element of the inverse of the submatrix J in the i-th row and j-th column.
5. The new energy station control method according to claim 4, characterized in that, The second formula is converted into a third formula, which is: where S j represents the reactive power capacity of the new energy station at node j, Q j_pu represents the reactive power capacity unit value of the new energy station at node j.
6. The new energy plant control method according to claim 4 or 5, characterized in that, each new energy station is sorted according to the reactive voltage support capability from large to small, and a preset proportion of the new energy stations ranked at the front is selected as the reference new energy stations, comprising: The reactive capacity of each new energy station is sorted from large to small, and the new energy stations with a reactive capacity ranking in the top 30% are selected as the reference new energy stations.
7. The new energy plant control method according to claim 6, characterized in that, The objective function and the constraint condition for optimization of the control parameter of each new energy station are constructed, comprising: The objective function for optimization of the control parameter of each new energy station is constructed: ； wherein F represents a target function value, t0 represents a time when a fault simulation starts, V i0 represents a voltage before a fault of a node i of a converter bus of an embedded direct current rectifier station i (t) represents a voltage after a fault of a node i of a converter bus of an embedded direct current rectifier station Based on the objective function, a plurality of constraint conditions of the new energy station grid-connected point voltage under a plurality of working conditions are determined.
8. The new energy plant control method according to claim 7, characterized in that, Based on the objective function, a plurality of constraint conditions of the new energy station grid-connected point voltage are determined, comprising: Based on the target function, after fault clearing, the overvoltage constraint condition of the new energy station grid-connected point is determined by avoiding the equipment off-grid in the new energy station, and the overvoltage constraint condition is: Wherein, U w is the voltage of the new energy station grid-connected point, T Uw>1.1 represents the time when the grid-connected point voltage exceeds 1.1 p.u. Based on the target function, after fault clearing, the low-voltage constraint condition of repeated access to the power grid of the new energy station is determined: n represents the number of times the voltage at the new energy station grid connection point decreases by 0.9 p.u.
9. The new energy station control method according to claim 8, characterized in that, The new energy station grid-connected point voltage has a downward trend through 0.9 p.u., and the following method is used to determine: ； wherein, Vgrid(k) represents the voltage of the grid-connected point of the new energy source when reading data for the kth time after the fault, Vgrid(k+1) represents the voltage of the grid-connected point of the new energy source when reading data for the k+1th time after the fault, and the time interval for reading data is 0.01 s.
10. The new energy plant control method according to claim 8, characterized in that, The value of the control parameter is determined by optimization according to the objective function and the constraint condition for the reference new energy station by using a deep reinforcement learning algorithm, comprising: The TD3 algorithm based on deep reinforcement learning is used for optimization, and the proportional coefficient k of the reactive current in the low-voltage ride-through control of the new energy station q The correction amount is mapped to the action, the post-fault embedded DC rectifier station converter bus node voltage state and the current new energy reactive current proportional coefficient k q The value of the proportional coefficient k is mapped to the state, and a reward function is constructed in combination with the following formula: ； Where F0 is the objective function value under the initial value of the reactive current proportion coefficient of the new energy station.