PV DC Series Arc Simulation Data

Solar photovoltaic (PV) systems have emerged as a pivotal player in the pursuit of clean energy solutions. As the adoption of solar technology continues to increase, it becomes imperative to address the some of the challenges associated with these systems. One such challenge that demands meticulous attention is the occurrence of DC series arcs. DC series arcs, characterized by unintentional electrical discharges across series-connected components, represent a potential threat to the integrity and efficiency of solar energy systems. Identifying and mitigating these arc faults is crucial not only for enhancing the overall performance of solar panels but also for safeguarding against the risk of electrical fires. This dataset was generated using a Simulink model based on the most common topology for PV DC power converters and known properties of arc events. The model consists of 4 boost converters connect to 4 individual PV panels and each can be individually controlled. Each converter is operated with a PWM module commanded by a duty cycle value and operating at a fixed frequency of 64kHz. The currents of each PV string are measured and added into a single signal, simulating existing HW architecture for arc detection in residential PV systems. The arc event is simulated by injecting gaussian noise limited to a specific frequency band with a range of 10 to 50 kHz. This dataset is intended to help in the evaluation of arc detection algorithms based on current measurements. Each file in the dataset contains 6 columns as follows: simulation timestamp, 250kHz sample rate current data, slow DC current measurement for each string in the system (4 strings) The name of the file described the conditions in which that set was obtained and follows this pattern: test_ + A + current target on string A + B + current target on string B + C + current target on string C + D + current target on string D + A + voltage target on string A + B + voltage target on string B + C + voltage target on string C + D + voltage target on string D + A + impedance on string A + B + impedance on string B + C + impedance on string C + D + impedance on string D + + _trans + A, B, C or D + + _arc + 0 or 1 + .csv

太阳能光伏系统（Solar Photovoltaic，PV）已成为清洁能源解决方案探索进程中的关键参与者。随着太阳能技术的应用规模持续扩大，应对此类系统相关的各类挑战已成为当务之急。其中一项需要细致关注的挑战便是直流（Direct Current，DC）串联电弧的发生。直流串联电弧指的是串联连接的组件间出现的非预期性放电现象，会对太阳能发电系统的运行稳定性与整体效率构成潜在威胁。识别并缓解此类电弧故障，不仅有助于提升光伏面板的整体性能，还能有效防范电气火灾风险。
本数据集基于当前光伏直流功率变换器的主流拓扑结构与已知的电弧事件特性，通过Simulink模型生成。该模型包含4台升压变换器，分别连接4块独立的光伏面板，且每一路均可独立控制。每台变换器均通过脉宽调制（Pulse Width Modulation，PWM）模块进行控制，由占空比参数指令，且工作在64kHz的固定频率下。本数据集采集每一路光伏串的电流并整合为单路信号，以此模拟住宅光伏系统中现有的电弧检测硬件（Hardware，HW）架构。电弧事件通过注入限定在10至50kHz频段内的高斯噪声进行模拟。
本数据集旨在助力基于电流采集数据的电弧检测算法的性能评估。数据集中的每个文件均包含以下6列数据：仿真时间戳、250kHz采样率的整合电流数据、系统内4路光伏串各自的低速直流电流测量值。数据文件的名称用于描述该数据集样本的生成条件，遵循以下格式：test_ + A + 路A电流目标值 + B + 路B电流目标值 + C + 路C电流目标值 + D + 路D电流目标值 + A + 路A电压目标值 + B + 路B电压目标值 + C + 路C电压目标值 + D + 路D电压目标值 + A + 路A阻抗值 + B + 路B阻抗值 + C + 路C阻抗值 + D + 路D阻抗值 + _trans + A/B/C/D + _arc + 0或1 + .csv。

搜集汇总

背景与挑战

背景概述

该数据集是一个关于太阳能光伏系统直流串联电弧模拟的数据集，通过Simulink模型生成，旨在帮助评估基于电流测量的电弧检测算法。数据集包含多个CSV文件，记录了不同条件下的电流和电压数据，适用于电力电子和电子安全领域的研究。