Energy storage low voltage ride through threshold value optimization method considering frequency-voltage coupling

Abstract

This invention discloses an optimization method for the low-voltage ride-through (LVRT) threshold value of energy storage considering frequency-voltage coupling. It utilizes a Beta distribution function to fit the probability density curve of historical voltage fault depth at the PCC point, and iterates through a preset search space to find the threshold value that maximizes the active power performance factor as the optimal LVRT start-up setting. This invention overcomes the problem of excessive active power reduction caused by the "one-size-fits-all" LVRT threshold value setting in existing technologies. It provides an LVRT threshold value optimization configuration method that considers the grid R / X ratio, grid strength, and fault distribution probability, breaking the limitations of traditional fixed threshold values. It can dynamically optimize the control strategy according to the grid strength characteristics and fault distribution patterns, effectively improving the active power support capability and system stability of energy storage units during faults.

Description

Technical Field

[0001] This invention belongs to the field of energy storage control technology, specifically relating to an optimization method for the low voltage ride-through threshold of energy storage considering frequency-voltage coupling. Background Technology

[0002] As fossil fuels are gradually replaced by renewable energy, the penetration rate of energy storage and electrochemical energy storage systems based on voltage source converter (VSC) interfaces has increased significantly. To cope with grid voltage fluctuations, current grid connection guidelines typically configure a uniform low voltage ride-through (LVRT) trigger threshold (usually 0.9 pu) for converters. When the point of common coupling (PCC) voltage falls below this threshold, the energy storage converter must immediately switch from steady-state operation mode to LVRT mode and preferentially inject reactive current to support the voltage according to specific instructions.

[0003] However, due to the thermal limits of the power electronic devices in the converter, its maximum current capacity is limited (typically 1.1 to 1.2 times the rated current). According to the current circle constraint, reactive current is preferentially injected in LVRT mode, which inevitably leads to a significant reduction in active current. For energy storage power stations that undertake the core tasks of grid frequency regulation and grid construction, this rigid mode switching can bring serious systemic risks. For example, in a large energy storage base in Northwest China, an external fault caused a voltage dip, triggering a massive number of units to simultaneously enter LVRT mode, resulting in an instantaneous active power loss of up to 6000MW, and the system frequency once dropped to 49.15Hz, almost causing the system to be disconnected.

[0004] This "one-size-fits-all" fixed threshold triggering mechanism, when applied to energy storage systems, ignores the complex and strong coupling relationship between frequency and voltage, mainly in the following two aspects: 1. Active power reduction exacerbates frequency instability: Energy storage systems are key anchors for maintaining the active power balance of new power systems. When the voltage drops slightly (e.g., 0.85~0.9 pu), if the energy storage is forced to enter LVRT mode and the active power is significantly reduced, the system will lose its valuable frequency support capability. This active power loss caused by voltage disturbance can easily turn into a serious frequency accident.

[0005] 2. Ignoring voltage / active power coupling due to grid impedance characteristics: Existing technologies neglect the differences in grid structure (such as resistance-to-reactance ratio and grid strength) at different access points. In reality, in grids with a high proportion of resistance or in certain weak grid environments, there is a significant positive coupling relationship between active power and node voltage (i.e., active power output can also effectively support voltage). In these scenarios, if a mechanical switch to the "high reactive, low active" LVRT mode is made simply because the voltage is slightly below 0.9 pu, the active power component that is beneficial to voltage recovery is cut off, leading to a double deterioration of both "frequency" and "voltage".

[0006] While existing technologies have proposed some methods for optimizing current allocation, such as the literature [Y. Guo, BC Pal, RA Jabr and H. Geng, “Global Optimality of Inverter Dynamic Voltage Support,”[J] IEEE Transactions on Power Systems, vol. 37, no. 5, pp. 3947-3957, Sept.2022] which establishes a current allocation optimization model considering active power limits and derives the optimal power factor angle for maximizing voltage support under different fault conditions; and the literature [A. Arjomandi-Nezhad, Y. Guo, BC Pal and D. Varagnolo, “A Model Predictive Approach for Enhancing Transient Stability of Grid-Forming Converters,” [J] IEEE Transactions on Power Systems, vol. 39, no. 5, pp. 6675-6688, Sept. 2024] which proposes an optimization algorithm based on model predictive control, aiming to achieve active power performance under fault conditions by dynamically adjusting the active power command and power angle magnitude. However, the above technologies are all based on a fixed LVRT threshold value to enter fault ride-through, and then dynamically adjust the current distribution to improve performance under the subsequent control algorithm. They cannot achieve performance improvement immediately, and they are more dependent on fault parameters and cannot cope with multiple fault types.

[0007] This demonstrates that most existing technologies are still based on fixed LVRT triggering standards, lacking adaptive judgment on "when to switch modes." In fact, under certain fault depths and grid impedance conditions, lowering the LVRT triggering threshold (e.g., allowing the energy storage system to maintain steady-state active power control during shallow faults) is more effective in balancing frequency stability and voltage recovery than blindly entering LVRT mode. Therefore, there is an urgent need for an active energy storage support method that can fully consider active power limits and frequency-voltage coupling characteristics, and adaptively adjust the control strategy. Summary of the Invention

[0008] In view of the above, the present invention provides an optimization method for the low voltage ride-through threshold value of energy storage considering frequency-voltage coupling. By considering the grid resistance-inductance ratio, grid strength and fault distribution probability, the LVRT threshold value is optimized and configured to overcome the problem of excessive active power reduction caused by the one-size-fits-all LVRT threshold value setting in the prior art.

[0009] An optimization method for the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling includes the following steps: (1) Construct an equivalent circuit model of the energy storage system connected to the grid via VSC. Based on the VSC current capacity constraint and the LVRT reactive power support guideline, derive the analytical expression of the active power output of VSC during the fault period. (2) Based on the analytical expression, establish the mapping relationship between the power grid parameters and the active power output boundary, and then quantitatively analyze the influence of the power grid short-circuit ratio, resistance-inductance ratio and power grid fault depth on the active power output characteristics; (3) Collect historical voltage sag data of PCC points, use the Beta distribution function to fit the probability density curve of voltage fault depth, and calculate the probability of occurrence of each fault depth after discretization. (4) Define the active power performance factor that includes system transient stability constraints, perform traversal calculations within the preset LVRT threshold value search space, and select the LVRT threshold value that maximizes the performance factor as the optimal LVRT start-up setpoint of the energy storage system.

[0010] Furthermore, the analytical expression for the active power output by the VSC during the fault period in step (1) is as follows: in: P The output active power of VSC. R g and X g These are the grid resistance and grid reactance, respectively. I d and I q These are the output active current and output reactive current of the VSC, respectively.U g The voltage amplitude is the equivalent voltage source of the faulty power grid.

[0011] Furthermore, the output active current I d and output reactive current I q The calculation expression is as follows: in: K This is the reactive power droop factor. U t This is the LVRT threshold value. I max This is the maximum current of VSC. U This is the voltage at point PCC.

[0012] Furthermore, the mapping relationship between the power grid parameters and the active power output boundary in step (2) is expressed as follows: in: U th This is the optimal threshold value for LVRT. R g and X g These are the grid resistance and grid reactance, respectively. I max This is the maximum current of VSC. K This is the reactive power droop factor. U g The voltage amplitude is the equivalent voltage source of the faulty power grid.

[0013] Furthermore, the functional expression of the probability density curve in step (3) is as follows: in: μ ( x Let be the probability density function with respect to the voltage fault depth. x The voltage fault depth refers to the voltage at the PCC point under fault conditions. α and β The shape parameters are fitted based on historical voltage sag data at the PCC point. B ( α, β ) is the Beta function.

[0014] Furthermore, in step (3), the voltage range of 0~0.9 pu is discretized at fixed intervals to obtain multiple voltage fault depths. For any voltage fault depth... x iSubstituting this into the probability density function yields the corresponding probability of fault depth occurrence. μ i .

[0015] Furthermore, the expression for the active power performance factor in step (4) is as follows: in: E AP The active power performance factor, P i To the depth of voltage fault x i The output active power of the VSC is... μ i To the depth of voltage fault x i The probability of failure occurring at the following depth x i The first discretized result i Each voltage fault depth, λ i As a stability penalty factor, at voltage fault depth x i When system transient instability occurs at the current LVRT threshold value λ i =0, otherwise λ i =1.

[0016] Furthermore, the LVRT threshold search space in step (4) is 0.7~0.9pu.

[0017] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the above-described method for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling.

[0018] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described method for optimizing the low-voltage ride-through threshold of energy storage, taking into account frequency-voltage coupling.

[0019] This invention overcomes the problem of excessive active power reduction caused by the "one-size-fits-all" LVRT threshold setting in the prior art. It provides an LVRT threshold value optimization configuration method that takes into account the grid R / X ratio, grid strength and fault distribution probability. It breaks the limitations of the traditional fixed threshold value and can dynamically optimize the control strategy according to the grid strength characteristics and fault distribution law, effectively improving the active power support capability and system stability of the energy storage unit during faults. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the process for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling, as described in this invention.

[0021] Figure 2 This is a schematic diagram of the equivalent circuit of an energy storage system connected to the grid via VSC under fault conditions.

[0022] Figure 3 This is a schematic diagram showing the relationship between the optimal threshold value of LVRT and the grid resistance reactance ratio under different voltage drop depths.

[0023] Figure 4 The graph shows the relationship between the optimal threshold value of LVRT and the grid reactance under different voltage drop depths. Figure 5 This diagram illustrates the transient instability phenomenon in an energy storage system caused by improper setting of the LVRT threshold value.

[0024] Figure 6 This is a schematic diagram of the probability distribution density and discretization of voltage drop faults. Detailed Implementation

[0025] To describe the present invention in more detail, the technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] like Figure 1 As shown, this embodiment provides a method for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling, including the following steps: Step S1: Construct an equivalent circuit model of energy storage connected to the grid via VSC. Based on the converter current capacity constraint and the LVRT reactive power support guideline, derive the analytical expression for the converter active power output during a fault.

[0027] Figure 2 This paper presents an equivalent circuit model of an energy storage system connected to the grid via a VSC under fault conditions. In this model, the VSC front-end stage, responsible for regulating the DC bus voltage, is simplified and represented as an ideal DC voltage source. The variables... U abc and I abc Representing the three-phase voltage and three-phase output current at PCC, respectively, the power grid is characterized by its Thevenin equivalent circuit, where... U g Represents the grid voltage after the fault. R g and X g These are the grid resistance and reactance, respectively.

[0028] Assuming the LVRT threshold value is variable, the active and reactive current outputs of the energy storage inverter under fault conditions are given by the following formula: in: K This represents the reactive power droop coefficient. U t This indicates the default LVRT threshold value. I max This indicates the maximum current of VSC. I d express I q These represent the active and reactive currents output by the VSC, respectively.

[0029] Therefore, the active power output of the energy storage unit during a fault P It can be represented as: The presence of nonlinearity complicates the analysis of the impact of the LVRT threshold on active power output capability. Therefore, by studying the current relationship that maximizes active power output, the optimal reactive current required to achieve this maximization under fault conditions can be derived as follows: The threshold value for LVRT that produces optimal active power performance can be determined. U th It can be represented as: Step S2: Based on the analytical expression, quantitatively analyze the impact of grid short-circuit ratio (SCR), resistance-to-inductance ratio (R / X) and grid fault depth on active power output characteristics, and establish the mapping relationship between grid parameters and active power output boundary.

[0030] As shown in the above equation, the key parameters affecting the LVRT threshold setting are: grid impedance ratio, grid strength, and the depth of grid voltage sag. Based on the derived active power transmission model, the following can be plotted: U th The relationship curves between the grid resistance-inductance ratio and the voltage drop depth are shown below. Figure 3 As shown.

[0031] Lowering the LVRT threshold allows the inverter to inject more active current during the same fault event. Due to the grid resistance, the increased active current creates a voltage drop across the resistor, which in turn helps support the grid voltage. Therefore, for grids with a high R / X ratio, lowering the LVRT threshold provides a more significant improvement in active power support compared to grids with a low R / X ratio; and for shallow voltage drops, lowering the LVRT threshold provides a more significant boost in active power than for deep voltage drops.

[0032] Based on the derived active power transmission model, the following can be drawn:U th The relationship curves between voltage drop depths and grid impedance are shown below. Figure 4 As shown.

[0033] When the power grid is strong, the impact of reactive current on voltage support is negligible, and active current becomes the dominant factor affecting active power output; therefore, in this case, lowering the LVRT threshold will enhance active power. However, for weak power grids, reactive current can affect active power output by supporting the grid voltage; therefore, a certain level of reactive power support is beneficial to active power output. It is also important to note that if active power exceeds the power transmission limit in a weak power grid, transient instability may occur, such as… Figure 5 As shown.

[0034] Step S3: Collect historical voltage sag data of PCC points, fit the probability density curve of voltage fault depth using the Beta distribution function, and calculate the occurrence probability of each fault depth after discretization.

[0035] The probability of a voltage sag varies by location, and the probability distribution function of the voltage fault depth is described by a discretized Beta distribution: in: x Represents the depth of voltage sag. α and β For shape parameters, B ( α, β () is a Beta function; the voltage range from 0 to 0.9 pu is divided into several intervals, and the fault probability quality of each interval is calculated. μ i .

[0036] To facilitate traversal, the probability density function is discretized, such as... Figure 6 As shown, the voltage range of 0 to 0.9 pu is divided into intervals with a step size of 0.1 pu.

[0037] Step S4: Define the active power performance factor that includes system transient stability constraints, perform traversal calculations within the preset LVRT threshold value search space, and select the threshold value that maximizes the value as the optimal LVRT start-up setpoint for the energy storage station.

[0038] Define active power performance factor E AP The active power performance factor, used as an indicator to evaluate the performance of different LVRT threshold values ​​under various fault probability distributions, is calculated using the following formula: in: xi Indicates the first i Each voltage fault depth, P i For the active power output at this voltage fault depth, μ i This represents the probability of a fault occurring at this voltage fault depth. λ i As a stability penalty factor, when the system experiences transient instability under a specific combination of voltage fault depth and LVRT threshold value. λ i =0, otherwise λ i =1.

[0039] The preset LVRT threshold value search space of 0.7~0.9 pu is traversed and calculated to select the value that makes it feasible. E AP The maximized LVRT threshold value is taken as the optimal LVRT start-up setpoint for the energy storage station. Through the selection results, we found that the selection of the optimal LVRT threshold value follows the following rules: For power grids with a high R / X ratio, priority should be given to lowering the LVRT threshold value to utilize the resistance voltage drop effect to improve active power output; For strong power grids, i.e., high SCR power grids, priority should be given to lowering the LVRT threshold value to reduce the crowding out of active current capacity by reactive current. For weak power grids, i.e. low SCR power grids, a trade-off is made based on the fault probability distribution characteristics: if the probability of shallow faults is high, a higher LVRT threshold value is maintained; if the probability of deep faults is high, the LVRT threshold value is lowered to prevent excessive active power reduction under deep faults.

[0040] The above description of the embodiments is provided to enable those skilled in the art to understand and apply the present invention. Those skilled in the art can readily make various modifications to the above embodiments and apply the general principles described herein to other embodiments without creative effort. Therefore, the present invention is not limited to the above embodiments, and any improvements and modifications made to the present invention by those skilled in the art based on the disclosure thereof should be within the scope of protection of the present invention.

Claims

1. A method for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling, characterized in that, Includes the following steps: (1) Construct an equivalent circuit model of the energy storage system connected to the grid via VSC. Based on the VSC current capacity constraint and the LVRT reactive power support guideline, derive the analytical expression of the active power output of VSC during the fault period. (2) Based on the analytical expression, establish the mapping relationship between the power grid parameters and the active power output boundary, and then quantitatively analyze the influence of the power grid short-circuit ratio, resistance-inductance ratio and power grid fault depth on the active power output characteristics; (3) Collect historical voltage sag data of PCC points, use the Beta distribution function to fit the probability density curve of voltage fault depth, and calculate the probability of occurrence of each fault depth after discretization. (4) Define the active power performance factor that includes system transient stability constraints, perform traversal calculations within the preset LVRT threshold value search space, and select the LVRT threshold value that maximizes the performance factor as the optimal LVRT start-up setpoint of the energy storage system.

2. The method for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling according to claim 1, characterized in that, The analytical expression for the active power output of VSC during the fault period in step (1) is as follows: in: P The output active power of VSC. R g and X g These are the grid resistance and grid reactance, respectively. I d and I q These are the output active current and output reactive current of the VSC, respectively. U g The voltage amplitude is the equivalent voltage source of the faulty power grid.

3. The method for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling according to claim 2, characterized in that, The output active current I d and output reactive current I q The calculation expression is as follows: in: K This is the reactive power droop factor. U t This is the LVRT threshold value. I max This is the maximum current of VSC. U This is the voltage at point PCC.

4. The method for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling according to claim 1, characterized in that, The mapping relationship between the power grid parameters and the active power output boundary in step (2) is expressed as follows: in: U th This is the optimal threshold value for LVRT. R g and X g These are the grid resistance and grid reactance, respectively. I max This is the maximum current of VSC. K This is the reactive power droop factor. U g The voltage amplitude is the equivalent voltage source of the faulty power grid.

5. The method for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling according to claim 1, characterized in that, The functional expression of the probability density curve in step (3) is as follows: in: μ ( x Let be the probability density function with respect to the voltage fault depth. x The voltage fault depth refers to the voltage at the PCC point under fault conditions. α and β The shape parameters are fitted based on historical voltage sag data at the PCC point. B ( α, β ) is the Beta function.

6. The method for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling according to claim 5, characterized in that: In step (3), the voltage range of 0~0.9 pu is discretized at fixed intervals to obtain multiple voltage fault depths. For any voltage fault depth... x i Substituting this into the probability density function yields the corresponding probability of fault depth occurrence. μ i .

7. The method for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling according to claim 1, characterized in that, The expression for the power performance factor in step (4) is as follows: in: E AP The active power performance factor, P i To the depth of voltage fault x i The output active power of the VSC is... μ i To the depth of voltage fault x i The probability of failure occurring at the following depth x i The discretized first i Each voltage fault depth, λ i As a stability penalty factor, at voltage fault depth x i When system transient instability occurs at the current LVRT threshold value λ i =0, otherwise λ i =1.

8. The method for optimizing the low-voltage ride-through threshold of energy storage considering frequency-voltage coupling according to claim 1, characterized in that, In step (4), the search space for the LVRT threshold value is 0.7~0.9 pu.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: The processor is used to execute the computer program to implement the energy storage low voltage ride-through threshold optimization method considering frequency-voltage coupling as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by the processor, it implements the energy storage low-voltage ride-through threshold optimization method considering frequency-voltage coupling as described in any one of claims 1 to 8.

Images