关键词:
高速公路
神经网络
电能质量预测
摘要:
随着高速公路的大力发展,特别是智慧公路的快速发展,高速公路基础设施建设逐步向数字化、信息化、智慧化方向转变。然而,数字化电力电子设备的广泛使用也带来了日益严重的电能质量问题,影响高速公路系统的正常运行和安全,增加了运营成本和维护难度。本文针对高速公路供电电能质量存在的问题,首先对现有电能质量评估指标的系统进行了归纳和总结,在此基础上,提出了一种基于时域卷积网络(TCN)和长短期记忆网络(LSTM)相结合的电能质量预测模型方法,并通过算例验证了该方法在提高预测准确性方面的优势。With the rapid development of highways, particularly smart highways, highway infrastructure construction is gradually transitioning towards digitalization, informatization, and intelligence. However, the widespread use of digital power electronic devices has led to increasingly severe power quality issues, which affect the normal operation and safety of highway systems, while also increasing operational costs and maintenance complexity. In response to the power quality problems in highway power supply systems, this paper first summarizes and organizes the existing power quality evaluation indices systematically. On this basis, a power quality prediction model is proposed, combining Temporal Convolutional Networks (TCN) and Long Short-Term Memory (LSTM) networks. The advantages of this method in improving prediction accuracy are demonstrated through case studies.