针对质子交换膜燃料电池(proton-exchange membrane fuel cells, PEMFC)剩余使用寿命的预测问题,本文提出了一种基于白鲸优化算法(beluga whale optimization, BWO)优化极限学习机(extreme learning machine, ELM)的预测方法。该方法首...针对质子交换膜燃料电池(proton-exchange membrane fuel cells, PEMFC)剩余使用寿命的预测问题,本文提出了一种基于白鲸优化算法(beluga whale optimization, BWO)优化极限学习机(extreme learning machine, ELM)的预测方法。该方法首先应用局部加权回归散点平滑法进行数据重构和平滑,以保留原始数据的主要趋势,同时有效去除噪声和尖峰。然后,通过相关性分析探讨电压与其他参数之间的关系。最后,利用BWO优化算法优化ELM模型的参数,以获取最优参数,从而实现PEMFC剩余使用寿命的精准预测。结果表明,该方法的决定系数接近于1,平均绝对百分比误差最小可达到2.7309e−10,显示了该方法在剩余使用寿命预测方面的优良准确性。For the prediction problem of the remaining useful life of proton-exchange membrane fuel cells (PEMFCs), this paper proposes a prediction method based on beluga whale optimization (BWO) optimized extreme learning machine (ELM). The method first applies a locally weighted regression scatter smoothing method for data reconstruction and smoothing to retain the main trends of the original data while effectively removing noise and spikes. Then, the relationship between voltage and other parameters is explored through correlation analysis. Finally, the parameters of the ELM model are optimized using the BWO optimization algorithm to obtain the optimal parameters for accurate prediction of the remaining useful life of the PEMFC. The results show that the coefficient of determination of the method is close to 1, and the mean average percentage error can be minimized to 2.7309e−10, which demonstrates the excellent accuracy of the method in remaining useful life prediction.展开更多
本文针对风电当并网时风电功率不稳会影响整个地区总网内的电压稳定的问题,建立以优化长短期记忆神经网络(Long-Short Term Memory,LSTM)超参数、提升风电功率的预测精度为目标的优化模型,提出基于多策略改进的白鲸优化算法(Multi-Strat...本文针对风电当并网时风电功率不稳会影响整个地区总网内的电压稳定的问题,建立以优化长短期记忆神经网络(Long-Short Term Memory,LSTM)超参数、提升风电功率的预测精度为目标的优化模型,提出基于多策略改进的白鲸优化算法(Multi-Strategy Beluga Whale Optimization,MSBWO)的求解方法。利用电力集团真实风力发电数据预测数据集1号风机数据来验证所提模型的可行性。试验结果表明,本文提出的方法能够获得更优解,收敛速度更快,预测效果更好。展开更多
文摘针对质子交换膜燃料电池(proton-exchange membrane fuel cells, PEMFC)剩余使用寿命的预测问题,本文提出了一种基于白鲸优化算法(beluga whale optimization, BWO)优化极限学习机(extreme learning machine, ELM)的预测方法。该方法首先应用局部加权回归散点平滑法进行数据重构和平滑,以保留原始数据的主要趋势,同时有效去除噪声和尖峰。然后,通过相关性分析探讨电压与其他参数之间的关系。最后,利用BWO优化算法优化ELM模型的参数,以获取最优参数,从而实现PEMFC剩余使用寿命的精准预测。结果表明,该方法的决定系数接近于1,平均绝对百分比误差最小可达到2.7309e−10,显示了该方法在剩余使用寿命预测方面的优良准确性。For the prediction problem of the remaining useful life of proton-exchange membrane fuel cells (PEMFCs), this paper proposes a prediction method based on beluga whale optimization (BWO) optimized extreme learning machine (ELM). The method first applies a locally weighted regression scatter smoothing method for data reconstruction and smoothing to retain the main trends of the original data while effectively removing noise and spikes. Then, the relationship between voltage and other parameters is explored through correlation analysis. Finally, the parameters of the ELM model are optimized using the BWO optimization algorithm to obtain the optimal parameters for accurate prediction of the remaining useful life of the PEMFC. The results show that the coefficient of determination of the method is close to 1, and the mean average percentage error can be minimized to 2.7309e−10, which demonstrates the excellent accuracy of the method in remaining useful life prediction.
