Verification method for fault ride-through characteristic of electromagnetic packaging model of wind turbine generator

Abstract

The invention relates to the technical field of new energy power system simulation, in particular to a method for verifying the fault ride-through characteristic of an electromagnetic packaging model of a wind turbine generator. The method comprises the following steps: S1, determining a deviation index for verifying the fault ride-through characteristic of the electromagnetic packaging model of the wind turbine generator; s2, completing identification and boundary calibration of each stage in a wind turbine generator electromagnetic packaging model fault ride-through process in a time domain; and S3, in each divided time interval, performing optimization calculation on the deviation index, and completing verification. The vectorization calculation technology of the SIMD architecture is creatively adopted, arithmetic operation, function calling and other mathematical operation can be applied to the whole simulation data array in a one-time and batch mode in a continuous memory block operation mode, an explicit circulation element-by-element processing mode is abandoned, the calculation efficiency in the fault ride-through characteristic verification process can be remarkably improved, and the fault ride-through characteristic verification efficiency is improved. The technical problem that the timeliness of a tool and a high-level programming language is insufficient in the simulation model verification process can be solved.

Description

Technical Field

[0001] The present application relates to the technical field of new energy power system simulation, and in particular to a method for verifying the fault ride-through characteristics of an electromagnetic packaging model of a wind turbine generator set. Background Art

[0002] Verifying the fault ride-through characteristics of wind turbine electromagnetic packaging models requires high-precision, high-dynamic response electromagnetic transient simulations, requiring the storage and analysis of test data at microsecond resolution. A single model verification condition typically lasts tens of seconds, requiring the storage and analysis of hundreds of millions of model verification test data points. A complete wind turbine electromagnetic packaging model verification process involves the verification of hundreds of model verification conditions, involving the analysis and processing of tens of billions of data points.

[0003] Currently, the programming languages ​​that simulation engineers rely on for data analysis are generally divided into two categories: one is compiled languages, represented by C, which can directly interact with computer hardware and allow precise control of underlying resources such as memory addresses and registers. They have high execution efficiency but low development efficiency, poor readability, and dependence on specific hardware architectures. The other is interpreted languages, represented by Python, which can shield hardware details through natural language syntax and abstract encapsulation, greatly improving development efficiency and maintainability, and supporting cross-platform operation. However, due to the need to be converted into machine code by a compiler or interpreter, there will be a certain performance loss. In the field of new energy power system simulation, simulation engineers need to have a combined background in electrical engineering, computers, and automatic control, so high-level languages ​​with strong ease of use are highly favored.

[0004] However, in the verification of the fault ride-through characteristics of wind turbine electromagnetic packaging models, interpreted languages ​​such as Python are unable to perform computational tasks with strict real-time requirements due to their underlying design characteristics. This is mainly reflected in the following aspects:

[0005] First, the indirect nature of high-level languages. As an interpreted language, Python requires multiple layers of function calls and data type conversions to implement complex functions. It cannot directly send efficient instructions to the CPU, which introduces additional overhead.

[0006] Second, loop efficiency is low. Python's for loop is interpreted, and each instruction must be dynamically parsed by the Python interpreter, resulting in loop speeds tens to hundreds of times slower than compiled languages. Even with just-in-time (JIT) optimizations like numba, it's still difficult to completely eliminate the delays introduced by the interpreter.

[0007] The third limitation is the Global Interpreter Lock (GIL). Python's GIL mechanism hinders multi-threaded parallel computing, further restricting the utilization of multi-core CPUs and making high-concurrency data processing with strict real-time requirements difficult to implement.

[0008] Therefore, in the current process of verifying the fault ride-through characteristics of the electromagnetic packaging model of wind turbines, the inherent defects of interpreted languages ​​make it impossible to meet the computing requirements of microseconds, low latency, and high throughput. Summary of the Invention

[0009] The purpose of this application is to address the problem that the inherent defects of interpreted languages ​​make it impossible to meet the microsecond, low-latency, and high-throughput computing requirements during the fault ride-through characteristics verification of the electromagnetic packaging model of a wind turbine. This application provides a method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine, which can effectively solve the computing efficiency bottleneck problem in the data comparison process between the electromagnetic packaging model of a wind turbine and the control hardware in the loop model in the field of new energy power system simulation.

[0010] To achieve the above objectives, this application adopts the following technical solutions:

[0011] A method for verifying the fault ride-through characteristics of an electromagnetic packaging model of a wind turbine generator system is designed based on an ARM architecture including a SIMD instruction set and includes the following steps:

[0012] S1, determine the deviation index for verifying the fault ride-through characteristics of the electromagnetic packaging model of the wind turbine;

[0013] S2, in the time domain, completes the identification and boundary calibration of the five stages of the fault ride-through process of the electromagnetic packaging model of the wind turbine generator set: the pre-fault steady state a, the fault transient b1, the fault steady state b2, the post-fault transient c1 and the post-fault steady state c2;

[0014] S3, within the five divided time intervals of pre-fault steady state a, fault transient state b1, fault steady state b2, post-fault transient state c1 and post-fault steady state c2, optimize and calculate the deviation index and complete the verification.

[0015] Furthermore, in step S2, the identification and boundary calibration of the five stages, namely, the steady state before the fault a, the transient state during the fault b1, the steady state during the fault b2, the transient state after the fault c1 and the steady state after the fault c2, are achieved by monitoring the change rate of the time domain physical quantity.

[0016] Furthermore, step S2 includes the following contents:

[0017] S21, align the fault occurrence time of the wind turbine electromagnetic packaging model and the wind turbine control hardware-in-the-loop model test data, and define it as time 0;

[0018] S22, according to the preset fault duration t fault , determine the fault clearing moment c1_begin;

[0019] S23, starting from the fault clearing moment c1_begin, perform reverse search and take the moment of the first deviation from the steady-state value β1p.u. during the fault duration as the boundary between stages b1 and b2, where β1 is between 0 and 0.2;

[0020] S24, starting from the end, search in reverse order. The moment of the first deviation from the steady-state value β2p.u. after the fault is cleared is the boundary between stages c1 and c2, where β2 is between 0 and 0.2.

[0021] Furthermore, step S21 includes the following contents:

[0022] S211, construct the time series physical quantity of the wind turbine control hardware-in-the-loop model to be analyzed into a one-dimensional vector and name it the original vector Determine the original vector The sampling period is Δt; the original vector Push back αΔt and form a lag vector α is a natural number;

[0023] S212, calculate the difference vector And based on the difference vector The preset threshold θ is used to determine the fault occurrence time of the wind turbine control hardware in-loop model test data;

[0024] S213, construct the time series physical quantity of the electromagnetic packaging model of the wind turbine to be analyzed into a one-dimensional vector and name it the original vector Determine the original vector The sampling period is Δt; the original vector Push back αΔt and form a lag vector α is a natural number;

[0025] S214, calculate the difference vector And based on the difference vector The preset threshold θ is used to determine the fault occurrence time of the wind turbine electromagnetic packaging model in-loop model test data;

[0026] S215 , defining the time when a fault occurs in the wind turbine generator electromagnetic packaging model-in-the-loop model test data as time 0, and defining the time when a fault occurs in the wind turbine generator control hardware-in-the-loop model test data as time 0.

[0027] Furthermore, in step S22, c1_begin=0+t fault , t fault is the fault duration.

[0028] Furthermore, in step S23, the original vector The average value during α sampling periods before the fault is cleared is taken as the steady-state value during the fault duration;

[0029] In the difference vector In the , whether the fault has been cleared and reached a steady state is judged according to the absolute value of the previous element of the element corresponding to c1_begin at the time of fault clearing: if it exceeds the preset threshold θ, it indicates that, under this working condition, the electromagnetic packaging model of the wind turbine generator set has not reached a steady state until the fault is cleared, and is directly judged as unqualified; if it does not exceed the preset threshold θ, it indicates that the electromagnetic packaging model of the wind turbine generator set has reached a steady state before the fault is cleared.

[0030] Furthermore, in step S24, the original vector The average value within α sampling periods before the end of the test recording is taken as the steady-state value after the fault is cleared;

[0031] In the difference vector In the equation , whether the absolute value of the last element exceeds the preset threshold θ is used to determine whether the electromagnetic packaging model of the wind turbine generator set has reached a steady state before the end of the test recording. If it exceeds the preset threshold θ, it indicates that under this working condition, the electromagnetic packaging model of the wind turbine generator set has not reached a steady state until the end of the test recording, and is directly judged as unqualified. If it does not exceed the preset threshold θ, it indicates that the electromagnetic packaging model of the wind turbine generator set has reached a steady state before the end of the test recording.

[0032] Furthermore, in step S1, the deviation indicators include average deviation, maximum deviation and weighted mean absolute deviation.

[0033] Further,

[0034] The average deviation is optimized by calculating the arithmetic mean of the fundamental positive sequence component difference between the test data of the wind turbine control hardware-in-the-loop model and the simulation data of the wind turbine electromagnetic package model-in-the-loop model within a steady-state and transient interval containing N time steps, and taking the absolute value thereof; wherein N is a natural number;

[0035] Optimize the calculation of the maximum deviation and calculate the absolute difference between the physical quantity data of the wind turbine control hardware-in-the-loop model and the simulation data of the wind turbine electromagnetic packaging model-in-the-loop model;

[0036] For the optimization calculation of the weighted mean absolute deviation, the average absolute deviation of active power, reactive power, active current, and reactive current before, during, and after the fault is calculated respectively. Then, the average absolute deviation of each time period of a single physical quantity is weighted averaged to obtain the weighted average absolute deviation during the entire fault process.

[0037] Furthermore, the weighted value before the fault is 10%, the weighted value during the fault is 60%, and the weighted value after the fault is 30%.

[0038] Beneficial effects of the present invention

[0039] SIMD technology is a parallel computing model that processes multiple data elements synchronously through a single instruction, which can significantly improve the computing efficiency of data-intensive tasks. The core principle of SIMD technology is to package and store multiple data with the help of wide vector registers, and perform the same operations on these data through a specific instruction set. In vector addition operations, the CPU needs to perform "load-calculate-store" operations on each element one by one, while SIMD technology can complete the parallel processing of all elements at one time, reducing the number of instruction executions and memory access latency. At the hardware level, modern CPUs implement SIMD by extending the instruction set, and the ARM architecture's scalable vector instruction set SVE also supports dynamic adjustment of register length to further optimize cross-platform compatibility.

[0040] The present application provides a method for verifying the fault ride-through characteristics of an electromagnetic packaging model of a wind turbine. The method innovatively adopts the vectorized computing technology of the SIMD architecture, and can apply mathematical operations such as the four arithmetic operations and function calls to the entire simulation data array in batches at one time in the form of continuous memory block operations, completely abandoning the element-by-element processing mode of explicit loops. The method can significantly improve the computational efficiency of the fault ride-through characteristics verification process, and can solve technical problems such as insufficient timeliness of tools and high-level programming languages ​​in the simulation model verification process.

[0041] That is, this application can provide a practical engineering solution with both computational accuracy and execution efficiency for the field of new energy power system simulation, which can be widely used in scenarios such as wind turbine grid-connected performance evaluation and grid adaptability testing, and has important engineering application value and technology promotion significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0043] Figure 1 This is a flowchart of the application.

[0044] Figure 2 Schematic diagram of voltage analysis results (a certain type of wind power converter with 0.1Pn≤P≤0.3Pn two-phase drop, dropping to 75%Un).

[0045] Figure 3Schematic diagram of active power analysis results (a certain type of wind turbine converter with 0.1Pn≤P≤0.3Pn two-phase drop, dropping to 75%Un).

[0046] Figure 4 Schematic diagram of reactive power analysis results (a certain type of wind turbine converter with 0.1Pn≤P≤0.3Pn two-phase drop, dropping to 75%Un).

[0047] Figure 5 Schematic diagram of active current analysis results (a certain type of wind power converter with 0.1Pn≤P≤0.3Pn two-phase drop, dropping to 75%Un).

[0048] Figure 6 Schematic diagram of reactive current analysis results (a certain type of wind turbine converter with 0.1Pn≤P≤0.3Pn two-phase drop, dropping to 75%Un).

[0049] Figure 7 Schematic diagram of current analysis results (a certain type of wind power converter with 0.1Pn≤P≤0.3Pn two-phase drop, dropping to 75%Un).

[0050] Figure 8 Schematic diagram of deviation index calculation results (a certain type of wind power converter with 0.1Pn≤P≤0.3Pn two-phase drop, dropping to 75%Un). DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of this application. At the same time, in the description of the embodiments of the present application, the terms "first", "second", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more features. In the description of the embodiments of the present application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0052] Example 1

[0053] In order to address the problem that the inherent defects of interpreted languages ​​make it impossible to meet the microsecond-level, low-latency, and high-throughput computing requirements during the fault ride-through characteristics verification of the electromagnetic packaging model of a wind turbine, this embodiment provides a method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine, which can effectively solve the computing efficiency bottleneck problem that exists in the data comparison process between the electromagnetic packaging model of a wind turbine and the wind turbine control hardware in the loop model in the field of new energy power system simulation.

[0054] like Figures 1 to 8 As shown, the present embodiment provides a method for verifying the fault ride-through characteristics of an electromagnetic packaging model of a wind turbine generator set, which is designed based on an ARM architecture including a SIMD instruction set and includes the following steps:

[0055] S1, determine the deviation index used to verify the fault ride-through characteristics of the electromagnetic packaging model of the wind turbine.

[0056] In this embodiment, the deviation indicators used to verify the fault ride-through characteristics of the electromagnetic packaging model of the wind turbine generator may include three deviation indicators: average deviation, maximum deviation and weighted average absolute deviation, so as to describe the data deviation between the electromagnetic packaging model of the wind turbine generator and the wind turbine generator control hardware-in-the-loop model.

[0057] S2, in the time domain, completes the identification and boundary calibration of the five stages of the fault-crossing process of the electromagnetic packaging model of the wind turbine generator set: pre-fault steady state a, fault transient b1, fault steady state b2, post-fault transient c1 and post-fault steady state c2.

[0058] Specifically, the identification and boundary calibration of the five stages, namely, the steady state before the fault a, the transient state during the fault b1, the steady state during the fault b2, the transient state after the fault c1, and the steady state after the fault c2, can be achieved by monitoring the change rate of the time domain physical quantity.

[0059] In order to ensure the accuracy and reliability of the division of the five stages, namely, the pre-fault steady state a, the fault transient state b1, the fault steady state b2, the post-fault transient c1, and the post-fault steady state c2, and provide an efficient time-domain segmentation solution for the fault ride-through characteristic analysis of the electromagnetic packaging model of the wind turbine generator, step S2 in this embodiment may include the following contents:

[0060] S21, aligning the wind turbine electromagnetic packaging model with the wind turbine control hardware in the loop model test data at the time of the fault. Specifically, step S21 includes the following contents:

[0061] S211, construct the time series physical quantity of the wind turbine control hardware-in-the-loop model to be analyzed into a one-dimensional vector and name it the original vector Determine the original vector The sampling period is Δt; the original vector Push back αΔt and form a lag vector α is a natural number.

[0062] S212, calculate the difference vector And based on the difference vector The preset threshold θ is used to determine whether the moment is the moment of fault occurrence, that is, the boundary between the steady state a before the fault and the transient state b1 during the fault.

[0063] Specifically, in step S212, when the differential vector When the absolute value of the rate of change of exceeds the preset threshold value θ, the moment is determined to be the moment of fault occurrence. In step S211, the original vector Push back 10Δt and form a lag vector At this time, the difference vector The elements in are represented as the original vector The variation within 10 sampling periods. 10 sampling periods is a common setting to prevent test data jitter and criterion sensitivity, and can be adjusted appropriately in actual use.

[0064] S213, construct the time series physical quantity of the electromagnetic packaging model of the wind turbine to be analyzed into a one-dimensional vector and name it the original vector Determine the original vector The sampling period is Δt; the original vector Push back αΔt and form a lag vector α is a natural number.

[0065] S214, calculate the difference vector And based on the difference vector The preset threshold θ is used to determine whether the moment is the moment of fault occurrence, that is, the boundary between the steady state a before the fault and the transient state b1 during the fault.

[0066] Specifically, in step S214, when the differential vector When the absolute value of the rate of change of exceeds the preset threshold value θ, the moment is determined to be the moment of fault occurrence. In step S213, the original vector Push back 10Δt and form a lag vector At this time, the difference vector The elements in are represented as the original vector The variation within 10 sampling periods. 10 sampling periods is a common setting to prevent test data jitter and criterion sensitivity, and can be adjusted appropriately in actual use.

[0067] In step S215, the fault occurrence time of the wind turbine electromagnetic packaging model-in-the-loop model test data is defined as time 0, and the fault occurrence time of the wind turbine control hardware-in-the-loop model test data is also defined as time 0, thereby aligning the fault occurrence times of the wind turbine electromagnetic packaging model and wind turbine control hardware-in-the-loop model test data. Subsequent phase division is performed only for the wind turbine control hardware-in-the-loop model test data, because under the current definition, the two sets of test data are already synchronized in the time domain.

[0068] S22, according to the preset fault duration t fault , determine the fault clearing moment c1_begin.

[0069] In step S22, for any given operating condition, the fault duration t fault is a known parameter set in advance. At this time, c1_begin=0+t fault .

[0070] S23, starting from the fault clearing moment c1_begin, the original vector Perform a reverse search, and define the moment of first deviation from the steady-state value β1p.u. during the fault duration as the boundary between stages b1 and b2, where β1 is between 0 and 0.2. Preferably, β1 can be selected as 0.05, meaning that the moment of first deviation from the steady-state value 0.05pu during the fault duration is the boundary between stages b1 and b2. It should be understood that the pu in 0.05pu refers to the per-unit value, a dimensionless unit commonly used in power system analysis to simplify calculations and standardize parameters.

[0071] In step S23, take the original vector The average value within α sampling periods before the fault is cleared is taken as the steady-state value during the fault duration. For example, the original vector The average value during the 10 sampling periods before the fault is cleared is taken as the steady-state value during the fault duration.

[0072] Furthermore, in the differential vector In the calculation, whether the absolute value of the previous element of the element corresponding to c1_begin at the fault clearing moment exceeds the preset threshold θ is used to determine whether the fault has been cleared and reached a steady state. If the value exceeds the preset threshold θ, it indicates that under this operating condition, the electromagnetic packaging model of the wind turbine generator set has not reached a steady state until the fault is cleared, and is directly judged as unqualified. If the value does not exceed the preset threshold θ, it indicates that the electromagnetic packaging model of the wind turbine generator set has reached a steady state before the fault is cleared.

[0073] S24, starting from the end, for the original vector Performing a reverse search, the moment of the first deviation from the post-fault-cleared steady-state value, β2p.u., is the boundary between stages C1 and C2. β2 is between 0 and 0.2. Preferably, β2 can be set to 0.05, meaning the moment of the first deviation from the post-fault-cleared steady-state value by 0.05pu is the boundary between stages C1 and C2. It should be understood that the pu in 0.05pu stands for per-unit value, a dimensionless unit commonly used in power system analysis to simplify calculations and standardize parameters.

[0074] In step S24, take the original vector The average value within α sampling periods before the end of the test recording is taken as the steady-state value after the fault is cleared.

[0075] Furthermore, in the differential vector In the equation , whether the absolute value of the last element exceeds the preset threshold θ is used to determine whether the electromagnetic packaging model of the wind turbine generator set has reached a steady state before the end of the test recording. If it exceeds the preset threshold θ, it indicates that under this working condition, the electromagnetic packaging model of the wind turbine generator set has not reached a steady state until the end of the test recording, and is directly judged as unqualified. If it does not exceed the preset threshold θ, it indicates that the electromagnetic packaging model of the wind turbine generator set has reached a steady state before the end of the test recording.

[0076] According to steps S21 to S24, during the fault ride-through process of the electromagnetic packaging model of the wind turbine generator, the boundaries of the five stages of pre-fault steady state a, fault transient state b1, fault steady state b2, post-fault transient c1 and post-fault steady state c2 have all been confirmed.

[0077] Step S2 of this embodiment fully utilizes the instruction-level parallelism and continuous memory access mechanism of modern CPUs, combines the SIMD instruction set to achieve data-level parallel acceleration, and optimizes the fault ride-through characteristic verification process of the electromagnetic packaging model of the wind turbine generator set, so that the verification process can be compiled into a low-level language through underlying optimization, avoiding the efficiency bottleneck of loop calculation; at the same time, this step S2 ensures the accuracy and reliability of the five-stage division of the pre-fault steady state a, the transient state b1 during the fault, the steady state b2 during the fault, the transient c1 after the fault, and the steady state c2 after the fault through quantitative analysis of the differential change rate, providing an efficient time domain segmentation solution for the fault ride-through characteristic analysis of the wind turbine generator set.

[0078] S3, in the five divided time intervals of pre-fault steady state a, fault transient state b1, fault steady state b2, post-fault transient state c1 and post-fault steady state c2, optimize the calculation of the deviation index to complete the verification.

[0079] In this embodiment, step S3 optimizes the calculation and statistics of the deviation algorithms of the average deviation, the maximum deviation, and the weighted average absolute deviation, and specifically includes the following contents:

[0080] Optimized calculation of mean deviation

[0081] In the steady-state and transient intervals containing N time steps, the arithmetic mean of the difference between the fundamental positive sequence components of the test data and the simulation data is calculated, and the absolute value is taken; where N is a natural number.

[0082] The mathematical formula for the mean deviation is as follows:

[0083]

[0084] By using vectorized operations, single instruction multiple data and continuous memory block operations are used to replace explicit loops, and some mathematical operations such as arithmetic operations or function calls are applied to the entire array at one time instead of processing them element by element.

[0085] The optimized average deviation mathematical formula is as follows:

[0086]

[0087] Among them, N1=b1_begin-a_begin=N2=ad_b1_begin-ad_a_begin;

[0088] rt_data[i,x] represents the physical quantity data set of the wind turbine control hardware-in-the-loop model; ad_data[j,x] represents the simulation data set of the wind turbine electromagnetic packaging model-in-the-loop model.

[0089] Optimized calculation of maximum deviation

[0090] In a steady-state interval containing N time steps, the maximum absolute value of the difference between the fundamental positive sequence components of the test data and the simulation data is calculated; where N is a natural number.

[0091] The maximum partial mathematical formula is as follows:

[0092] σ MXE =max(|σ(1)|,|σ(2)|,...,|σ(N)|)

[0093] The optimized maximum deviation formula is as follows:

[0094] Among them, diff a For construction Difference vector, whose elements are the absolute differences of the corresponding elements of the two sets of data sequences; rt a_begin+i,x Represents the physical quantity data of the wind turbine control hardware-in-the-loop model; ad ad_a_begin+i,x Represents the simulation data of the electromagnetic packaging model of the wind turbine.

[0095] Optimized calculation of weighted mean absolute deviation

[0096] The average absolute deviation of active power, reactive power, active current, and reactive current before, during, and after the fault are calculated respectively. Then, the average absolute deviation of each time period of a single physical quantity is weighted averaged to obtain the weighted average absolute deviation during the entire fault process. Specifically, the weighted value before the fault can be 10%, the weighted value during the fault can be 60%, and the weighted value after the fault can be 30%.

[0097] Finally, according to the maximum allowable value of the fault ride-through characteristic electromagnetic transient model verification deviation in the "NB / T 131053 Wind Turbine Electromagnetic Packaging Model Verification Procedure", the deviation range of different physical quantities is restricted, that is, different qualified ranges are defined for different physical quantity deviations; different physical quantities include active power, reactive power, active current, reactive current, etc.

[0098] This embodiment provides a method for verifying the fault ride-through characteristics of an electromagnetic packaging model of a wind turbine generator. This method addresses the computational efficiency bottleneck problem that exists in the data comparison process between the electromagnetic packaging model of a wind turbine generator and the control hardware-in-the-loop model in the field of new energy power system simulation. By innovatively using vectorized computing technology based on the SIMD architecture, mathematical operations such as arithmetic operations and function calls are applied to the entire simulation data array in batches at one time as continuous memory block operations, completely abandoning the element-by-element processing mode of explicit loops. Furthermore, this method fully utilizes the instruction-level parallelism and continuous memory access mechanism of modern CPUs, combines the SIMD instruction set to achieve data-level parallel acceleration, and optimizes the fault ride-through characteristics verification process of the electromagnetic packaging model of a wind turbine generator, enabling the verification calculation process to be compiled into a low-level language through underlying optimization. Furthermore, this method ensures the accuracy and reliability of the five-stage division of pre-fault steady state, fault transient state, fault steady state, post-fault transient state, and post-fault steady state through quantitative analysis of the differential change rate, providing an efficient time-domain segmentation solution for wind turbine fault ride-through characteristics analysis.

[0099] The present embodiment provides a method for verifying the fault ride-through characteristics of an electromagnetic packaging model of a wind turbine, which can significantly improve the computational efficiency during the fault ride-through characteristics verification process, and can resolve technical difficulties such as insufficient timeliness of tools and high-level programming languages ​​in the simulation model verification process. It can provide a practical engineering solution that combines computational accuracy and execution efficiency for the field of new energy power system simulation, thereby saving manpower expenses and operating costs of the model verification department of power grid enterprises. It can be widely used in scenarios such as wind turbine grid-connected performance evaluation and power grid adaptability testing, and has important engineering application value and technical promotion significance.

[0100] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences between it and other embodiments.

[0101] The foregoing description of specific embodiments of this specification describes the process. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are possible or may be advantageous.

[0102] The foregoing is merely an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A method for verifying the fault ride-through characteristics of an electromagnetic packaging model of a wind turbine generator, characterized by: Based on the ARM architecture design including SIMD instruction set, the following steps are included: S1, determine the deviation index for verifying the fault ride-through characteristics of the electromagnetic packaging model of the wind turbine; S2, in the time domain, completes the identification and boundary calibration of the five stages of the fault ride-through process of the electromagnetic packaging model of the wind turbine generator set: the pre-fault steady state a, the fault transient b1, the fault steady state b2, the post-fault transient c1 and the post-fault steady state c2; S3, within the five divided time intervals of pre-fault steady state a, fault transient state b1, fault steady state b2, post-fault transient state c1 and post-fault steady state c2, optimize and calculate the deviation index and complete the verification.

2. The method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine generator set according to claim 1, characterized in that: In step S2, the identification and boundary calibration of the five stages, namely, the steady state before the fault a, the transient state during the fault b1, the steady state during the fault b2, the transient state after the fault c1 and the steady state after the fault c2, are achieved by monitoring the change rate of the time domain physical quantity.

3. The method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine generator set according to claim 2, characterized in that: Step S2 includes the following contents, S21, align the fault occurrence time of the wind turbine electromagnetic packaging model and the wind turbine control hardware-in-the-loop model test data, and define it as time 0; S22, according to the preset fault duration t fault , determine the fault clearing moment c1_begin; S23, starting from the fault clearing moment c1_begin, perform reverse search and take the moment of the first deviation from the steady-state value β1p.u. during the fault duration as the boundary between stages b1 and b2, where β1 is between 0 and 0.2; S24, starting from the end, search in reverse order. The moment of the first deviation from the steady-state value β2p.u. after the fault is cleared is the boundary between stages c1 and c2, where β2 is between 0 and 0.

2.

4. The method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine generator set according to claim 3, characterized in that: Step S21 includes the following contents: S211, construct the time series physical quantity of the wind turbine control hardware-in-the-loop model to be analyzed into a one-dimensional vector and name it the original vector Determine the original vector The sampling period is Δt; the original vector Push back αΔt and form a lag vector α is a natural number; S212, calculate the difference vector And based on the difference vector The preset threshold θ is used to determine the fault occurrence time of the wind turbine control hardware in-loop model test data; S213, construct the time series physical quantity of the electromagnetic packaging model of the wind turbine to be analyzed into a one-dimensional vector and name it the original vector Determine the original vector The sampling period is Δt; the original vector Push back αΔt and form a lag vector α is a natural number; S214, calculate the difference vector And based on the difference vector The preset threshold θ is used to determine the fault occurrence time of the wind turbine electromagnetic packaging model in-loop model test data; S215 , defining the time when a fault occurs in the wind turbine generator electromagnetic packaging model-in-the-loop model test data as time 0, and defining the time when a fault occurs in the wind turbine generator control hardware-in-the-loop model test data as time 0.

5. The method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine generator set according to claim 4, characterized in that: In step S22, c1_begin=0+t fault , t fault is the fault duration.

6. The method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine generator set according to claim 5, characterized in that: In step S23, take the original vector The average value during α sampling periods before the fault is cleared is taken as the steady-state value during the fault duration; In the difference vector In the , whether the fault has been cleared and reached a steady state is judged according to the absolute value of the previous element of the element corresponding to c1_begin at the time of fault clearing: if it exceeds the preset threshold θ, it indicates that, under this working condition, the electromagnetic packaging model of the wind turbine generator set has not reached a steady state until the fault is cleared, and is directly judged as unqualified; if it does not exceed the preset threshold θ, it indicates that the electromagnetic packaging model of the wind turbine generator set has reached a steady state before the fault is cleared.

7. The method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine generator set according to claim 6, characterized in that: In step S24, take the original vector The average value within α sampling periods before the end of the test recording is taken as the steady-state value after the fault is cleared; In the difference vector In the equation , whether the absolute value of the last element exceeds the preset threshold θ is used to determine whether the electromagnetic packaging model of the wind turbine generator set has reached a steady state before the end of the test recording. If it exceeds the preset threshold θ, it indicates that under this working condition, the electromagnetic packaging model of the wind turbine generator set has not reached a steady state until the end of the test recording, and is directly judged as unqualified. If it does not exceed the preset threshold θ, it indicates that the electromagnetic packaging model of the wind turbine generator set has reached a steady state before the end of the test recording.

8. The method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine generator set according to any one of claims 1 to 7, characterized in that: In step S1 , the deviation indicators include the average deviation, the maximum deviation, and the weighted mean absolute deviation.

9. The method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine generator set according to claim 8, characterized in that: The average deviation is optimized by calculating the arithmetic mean of the fundamental positive sequence component difference between the test data of the wind turbine control hardware-in-the-loop model and the simulation data of the wind turbine electromagnetic package model-in-the-loop model within a steady-state and transient interval containing N time steps, and taking the absolute value thereof; wherein N is a natural number; Optimize the calculation of the maximum deviation and calculate the absolute difference between the physical quantity data of the wind turbine control hardware-in-the-loop model and the simulation data of the wind turbine electromagnetic packaging model-in-the-loop model; For the optimization calculation of the weighted mean absolute deviation, the average absolute deviation of active power, reactive power, active current, and reactive current before, during, and after the fault is calculated respectively. Then, the average absolute deviation of each time period of a single physical quantity is weighted averaged to obtain the weighted average absolute deviation during the entire fault process.

10. The method for verifying the fault ride-through characteristics of the electromagnetic packaging model of a wind turbine generator set according to claim 9, characterized in that: The weighted value before the fault is 10%, the weighted value during the fault is 60%, and the weighted value after the fault is 30%.

Images