An Active Injected Location Method for DC Fault of Wind Power Based on Parallel Energy Absorption Module

In recent years, the widespread adoption of wind power has been facilitated by the rapid advancements in new energy generation technology. The precise identification and efficient management of DC faults in onshore wind farms has gained increasing significance. To address issues such as limited functionality, low device utilization, and the inefficient use of fault energy prevalent in current fault location methods applied in wind power flexible grid-connected systems, this study explores the storage and reuse technology of fault energy. The article proposes a topology structure designed to absorb and reuse fault energy under multiple DC faults, aiming to overcome the challenges posed by single-function methods. The structure and control strategy of this topology are clearly delineated, and the design methodology for topology electrical parameters is thoroughly investigated. Simulation studies were conducted to validate the electrical characteristics of the employed devices and assess the operational effectiveness of the proposed topology within wind power grid-connected systems. In comparison with conventional fault location methods, the proposed scheme exhibits multiple energy absorption functions for DC faults and fault location capabilities, especially in the presence of permanent DC faults. As a result, this approach maximizes the utilization of DC fault energy, ensuring higher device efficiency and a broader spectrum of functionalities.


Introduction
Currently, wind power generation stands as a well-established and extensively applied method within the realm of new energy generation [1][2][3].It boasts attributes such as effective adjustability, widespread availability of resources, and an environmentally friendly and clean profile.As wind power technology continues its rapid advancement and sees widespread implementation, the issue of DC transmission line faults in flexible DC transmission systems based on MMC-HVDC is increasingly manifesting [4][5][6].The precise localization of DC faults and the efficient absorption and dissipation of fault energy play pivotal roles in ensuring the stable operation of grid-connected wind power transmission systems and fostering the continued progress of wind power technology.
The distinctive features of a flexible DC grid, characterized by low inertia and weak damping, pose challenges in addressing DC short-circuit faults.These faults lead to a rapid surge in fault current, necessitating a sophisticated control process that encompasses suppression, ride-through, and reclosing procedures.This complexity elevates the demands and intricacies of DC-side fault handling technology, especially in scenarios like long-distance wind power transmission.For such scenarios, an essential aspect is the DC system's capacity for real-time fault spot identification and location under unattended conditions, a critical requirement, particularly during DC faults in wind power 2 transmission over extended distances.Techniques for online detection of DC faults generally fall into two categories: those reliant on DC circuit breakers [7,8] and those dependent on inverters [9][10][11].
An innovative adaptive reclose scheme applied to the MMC-HVDC system is presented in reference [7].This approach proactively introduces voltage pulses by manipulating the solid-state device of the primary circuit breaker through switch control.Simultaneously, it identifies fault characteristics through the application of a traveling wave fault localization algorithm.A comparable active pulse injection strategy was suggested in reference [8].However, it is worth noting that the cited literature introduces adaptive reclosing schemes designed for both high-voltage DC systems and mesh DC power grids.Reference [11] introduces an adaptive automatic reclosing scheme specifically tailored for bipolar high-voltage DC hybrid systems, relying on the Modular Multilevel Converter (MMC).This method capitalizes on the voltage controllability of the healthy pole MMC to introduce DC voltage disturbance.Subsequently, fault identification is accomplished through wavelet transform analysis of the traveling wave induced by the injected signal.
The limitations observed in the aforementioned detection technologies encompass: (1) the need for intricate control and stringent performance standards for converters or the breaking capacity of DC circuit breakers, thereby leading to increased costs; (2) Challenges in disentangling the sequential steps of fault detection and the subsequent recovery operation, impeding the swift restoration of the power grid; (3) Inefficient utilization of energy from short-circuit faults, resulting in wasted resources and a suboptimal device utilization rate.
Diverging from the approaches outlined earlier, reference [12] introduces an adaptive reclosing scheme employing parallel energy absorption module pulse injection.Its principle deviates from the series current limiting method mentioned previously, opting for a parallel circuit topology for fault energy diversion and absorption.This significantly diminishes reliance on the converter or DC circuit breaker.The distinctive topology structure proposed in this scheme additionally leverages the stored fault energy in the absorption module to inject pulse voltage into the fault line, thereby accomplishing online fault localization functions.Nevertheless, it's essential to note that this scheme is specifically designed for bipolar short-circuit faults in DC lines.
Building upon the summarized work above, this paper aims to broaden the applicability of existing schemes and enhance device utilization.To achieve this, a topology is introduced, integrating functions for fault energy absorption and fault location across various DC fault scenarios.The paper provides insights into the topology, the design methodology for key electrical parameters, and details the fault energy absorption and fault location methods.Furthermore, a simulation model is established on the PSCAD/EMTDC simulation platform to validate the proposed approach.

Structure of Topology
The illustrative structure of a wind power flexible direct grid connection system is depicted in figure 1.Its primary components comprise the wind farm, converter station, DC line (overhead line), and AC main network.DC transmission adopts a symmetrical single-pole connection.From an economic standpoint, converter stations on both sides typically opt for the half-bridge sub-module topology.

Operation Mode under Different Fault Scenarios
The operation mode of the topology under different fault scenarios.1 max Where k is the reliability factor, generally equals 1.25.

Number of Diodes in Series in
Diode Group.To facilitate analysis, the line model in this mode is represented as a lumped parameter with resistance and inductance.Upon the current i(t) decreasing to 0, the termination time tm of the energy absorption and the capacitor's voltage Um can be computed [13][14].
The capacitance C should be constrained by both tm and Um.It is essential to have a small tm value for swift fault handling.Simultaneously, Um must be sufficiently high to amplify the voltage pulse injected during subsequent fault localization, minimizing errors caused by noise interference.Moreover, Um should be maintained within the breakdown voltage of the capacitor, considering the capacitor's withstand voltage limit.
Illustrating with the two-terminal MMC-HVDC system depicted in figure 2, the selection of a suitable capacitance value is exemplified.Consequently, a capacitance value of 15 mF is deemed appropriate.

Control Strategy for Fault Energy Absorption
In figure 5, we briefly illustrate the variations in DC current and capacitor voltage at key nodes during a bipolar short-circuit fault scenario.Idcp represents the current on the positive line, and Uc signifies the voltage of the absorption capacitor.
By t2, the DCCB completes the disconnection process, effectively isolating the fault current.From t2 to t3, the voltage rises as C1 and C2 absorb fault energy through the circuit illustrated in figure 3(b).At t3, the Idcp current decreases to zero, and the capacitor voltage Uc approaches its maximum value.

The Method of DC Fault Location
As an illustration, in the case of a bipolar short-circuit fault, S1 and S2 are closed, and T1 and T2 are controlled to turn on within a short duration Top.The value of Top is constrained by the IGBT's switching frequency and sampling frequency.During this period, the topology forms the loop illustrated in figure 3 This square wave can be regarded as a voltage traveling wave.The calculation of the backward voltage traveling wave and forward voltage traveling wave at a specific point can be determined through voltage u(t), current i(t), and wave impedance Zc: The u1(t) corresponds to the backward propagating voltage traveling wave, and u2(t) signifies the forward propagating voltage traveling wave.
Assuming the initial injection of the forward voltage traveling wave occurs at time tf, and the detection of the first reflected traveling wave takes place at time tr, the fault distance lm can be calculated as follows: Where v represents the speed of traveling wave propagation along the line.For the overhead line, it can be approximated as 3×10^8 m/s.

Simulation
To validate the performance of the proposed topology, as well as the efficacy of the control strategy and positioning method, a simulation model is constructed using the PSCAD/EMTDC simulation platform.The system model, depicted in figure 2, is designed to assess the proposed methodology.The simulation parameters for the system are configured based on the values specified in table I.The MMC utilizes sub-modules with a half bridge structure, while the overhead line is modeled using a phase-frequency model within the simulation environment.

Fault Energy Absorption
Take the bipolar short-circuit fault as an example.The fault is introduced at a location 300 km away from MMC1 and is triggered at 1.8 s.The DCCB activates at 1.802 s and completes the breaking process at 1.81 s.As per the previous analysis, the value of single capacitor is determined as 15 mF.The simulation results of the current Idcp on the positive line and the voltage Uc of the absorption capacitors are depicted in figure 6.During the bipolar short-circuit fault, once the DC fault current reaches 7.5 kA at 1.813 s, the DCCB completes the disconnection process, allowing capacitors C1 and C2 to commence absorbing the fault energy.After 86 ms, Idcp reduces to zero, and Uc reaches its maximum value of 2.95 kV.

Fault Location of the DC Line
The verification of the fault location function is based on the verification of fault energy absorption in the previous section.After the absorption process is completed, a delay is used to recover the insulation.Then actively inject a traveling wave into the DC line.
The duration of the traveling wave, Top, equals 0.2 ms.The simulation results are depicted in figure 7.According to the results and equation (3), the fault distance of the bipolar short-circuit fault is 307 km, with a slight error of 2.3%.The fault distance of the positive-ground fault is calculated to be 308.35km, with a slight error of 2.8%.

Discussion
The topology proposed in this paper is based on HBSM based MMC.Compared with hybrid structure MMC and FBSM MMC configurations, this scheme distinctly excels in terms of device cost.In relative terms, the control strategy of the HBSM-based MMC for HVDC systems exhibits a higher degree of simplicity.
Meanwhile, the reclosing scheme delineated in the literature [7] is tailored exclusively to circumstances wherein overhead lines serve as DC transmission lines, specifically addressing DC double-pole short-circuit fault conditions.In contrast, the topological configuration introduced in this paper is versatile.Structurally, it adds an absorbing capacitor and sets a grounding point between the two capacitors.Through similar devices and structural designs, it finds application in both double-pole short circuit and single-pole ground fault conditions.
Building upon the foundations laid in this paper, the proposed topology can be synergistically integrated with a chopper circuit, rendering it operationally proficient even in the presence of AC power grid faults.Simultaneously, when employing a combination of overhead lines and submarine cables as DC lines within offshore wind power application scenarios, enhancements to existing fault localization methods become imperative to ensure the efficacy of utilizing traveling waves for localization.These studies and improvements constitute subjects for our future research.

Conclusion
Aiming at the problems of low equipment utilization and single function of existing DC fault location schemes for wind power, this paper proposes a topology with fault energy absorption capability and fault location method based on it.In this paper, the structure of the topology and its control strategy is clarified, the topology electrical parameters design method is studied, and the operation effect of the topology is verified by simulation.Compared with conventional schemes, the method proposed in this paper has the following advantages: 1) With the energy absorption function for multiple DC faults, the function is more comprehensive.
2) With the fault location function in the case of DC permanent fault, it is more conducive to the maintenance and long-term operation of the system.
3) Each functional module is highly integrated, and the devices assume multiple functions, avoiding the problems of low device utilization and high cost caused by scattered arrangement.

Figure 2 .
Figure 2. Topology of the device.

2. 1 . 1 .Figure 3 .Figure 4 .
Figure 3. DC bipolar short-circuit fault.(a) Energy absorption circuits.(b) Fault location circuits.2.1.2.Single Pole Grounding Fault (Using Negative Pole Grounding as an Example).In figure4(a), in the event of a positive grounding fault on the DC line, the system remains in a unipolar state for a certain duration.Following the open circuit action of the DCCB, all IGBT switches, S1, and S2 are

Figure 5 .
Figure 5. Trend of fault current and capacitor voltage under DC faults.
(b), generating a voltage pulse injected into the line.The configuration of switches S1 and S2 ensures that the voltage pulse doesn't need to traverse the current-limiting reactor.Due to the brief conduction time, the capacitor voltage can be considered constant.The voltage of a single capacitor eventually charges to Um, making the voltage pulse theoretically a square wave with an amplitude of Um and a duration of Top.

Figure 6 .
Figure 6.Simulation of DC fault energy absorption.

7 Figure 7 .
Figure 7. Simulation of traveling wave pulse for location of DC Fault.(a) Bipolar short-circuit fault.(b) Negative ground fault.