强迫振荡检测通过一个混合网络的尖端反复的神经网络和LSTM的混合网络
Xiaomei Yang1, Jinfei Wang1, Xingrui Huang1
1College of Electrical Engineering, Sichuan University, Chengdu 610065, China.
Sensors (Basel, Switzerland)
|April 26, 2025
概括
本研究介绍了一种高效的混合深度学习模型,用于电力系统的稳定性. 这种新型网络准确地检测到强制振荡,即使有噪声,也超过了现有的方法.
科学领域:
- 电气工程 电气工程
- 计算机科学 计算机科学
- 人工智能的人工智能
背景情况:
- 电力系统稳定性监测在检测强制振荡方面面临挑战.
- 现有的人工神经网络 (ANN) 对边缘设备来说是计算密集型和耗电型.
- 深度学习 (DL) 显示了准确强制振荡检测的前景.
研究的目的:
- 提出一种新的混合网络,以高效准确地检测电力系统中的强制振荡.
- 解决ANN在边缘计算环境中的限制,用于电力系统监控.
- 为了更好地区分强制和自然的振荡.
主要方法:
- 用于计算和能源效率的尖端循环神经网络 (SRNN) 与用于捕获时间依赖的长短期记忆 (LSTM) 网络的集成.
- 使用反向传播通过时间 (BPTT) 训练混合网络,并对SRNN的不连续梯度进行修改.
- 对模拟和现实世界的振荡数据集进行评估.
主要成果:
- 拟议的混合SRNN-LSTM模型在区分被迫和自然振荡方面表现出更高的准确性.
- 该模型与纯LSTM和其他基于SRNN的模型相比,表现出更高的性能.
- 即使存在大量噪音,也可以实现有效的检测.
结论:
- 混合SRNN-LSTM网络为电力系统稳定性监控提供了高效和准确的解决方案.
- 这种方法适用于需要低功耗和及时检测的边缘设备.
- 该模型增强了在具有挑战性的条件下将强制振荡与自然振荡区分开来的能力.
相关概念视频
01:30Oscillations In An LC Circuit
2.1K
An idealized LC circuit of zero resistance can oscillate without any source of emf by shifting the energy stored in the circuit between the electric and magnetic fields. In such an LC circuit, if the capacitor contains a charge q before the switch is closed, then all the energy of the circuit is initially stored in the electric field of the capacitor. This energy is given by
2.1K
01:30RLC Circuit as a Damped Oscillator
791
An RLC circuit combines a resistor, inductor, and capacitor, connected in a series or parallel combination.
Consider a series RLC circuit. Here, the presence of resistance in the circuit leads to energy loss due to joule heating in the resistance. Therefore, the total electromagnetic energy in the circuit is no longer constant and decreases with time. Since the magnitude of charge, current, and potential difference continuously decreases, their oscillations are said to be damped. This is...
Consider a series RLC circuit. Here, the presence of resistance in the circuit leads to energy loss due to joule heating in the resistance. Therefore, the total electromagnetic energy in the circuit is no longer constant and decreases with time. Since the magnitude of charge, current, and potential difference continuously decreases, their oscillations are said to be damped. This is...
791
01:25Propagation of Action Potentials
4.7K
The propagation of an action potential refers to the process by which a nerve impulse, or "action potential," travels along a neuron.
Neurons (nerve cells) have a resting membrane potential, with a slightly negative charge inside compared to outside. This is maintained by ion channels, such as sodium (Na+) and potassium (K+) channels, which control the flow of ions. When a stimulus, like a touch or a signal from another neuron, triggers the neuron, sodium channels open, allowing sodium...
Neurons (nerve cells) have a resting membrane potential, with a slightly negative charge inside compared to outside. This is maintained by ion channels, such as sodium (Na+) and potassium (K+) channels, which control the flow of ions. When a stimulus, like a touch or a signal from another neuron, triggers the neuron, sodium channels open, allowing sodium...
4.7K
01:22Muscle Stimulation Frequency
1.9K
The contraction strength of muscles is regulated by motor neurons, which modulate the frequency of action potentials dispatched to the motor units based on the body's requirements. This process of varying the muscle stimulation frequency allows muscles to contract with a force that is precisely tailored to the needs of the moment, whether lifting a feather or a heavy box.
Wave summation
At low firing rates, motor neurons induce individual twitch contractions in muscle fibers. These twitches...
Wave summation
At low firing rates, motor neurons induce individual twitch contractions in muscle fibers. These twitches...
1.9K
01:21State Space to Transfer Function
145
The conversion of state-space representation to a transfer function is a fundamental process in system analysis. It provides a method for transitioning from a time-domain description to a frequency-domain representation, which is crucial for simplifying the analysis and design of control systems.
The transformation process begins with the state-space representation, characterized by the state equation and the output equation. These equations are typically represented as:
The transformation process begins with the state-space representation, characterized by the state equation and the output equation. These equations are typically represented as:
145