【摘要】 针对供热管道微小泄漏状况的预测问题，提出了一种基于自适应噪声完备集合经验模态分解(CEEMDAN)以及改进灰狼优化(IGWO)算法优化反向传播(BP)神经网络的泄漏孔径预测方法。所提方法利用CEEMDAN以及能量矩对泄漏信号进行模态分解与特征提取；为提高预测精度，提出IGWO算法。首先，对灰狼优化(GWO)算法的种群初始化方式以及控制参数与位置更新策略进行改进；然后，建立IGWO-BP预测模型，并利用实验室泄漏信号对预测模型进行验证。结果表明：所提预测模型可有效提高管道微小泄漏孔径的预测精度。更多还原

【Abstract】 Aiming at prediction problem of small leaks in heating supply pipeline, a leak aperture prediction method based on the complete ensemble empirical modal decomposition with adaptive noise(CEEMDAN) and the improved gray wolf optimization(IGWO)algorithm optimized back propagation(BP)neural networks is proposed.The proposed method uses CEEMDAN and energy moments for modal decomposition and feature extraction of the leakage signal.To improve the prediction precision, the IGWO algorithm is proposed.Firstly, the population initialization method and the control parameters and position update strategy of the gray wolf optimization(GWO)algorithm are improved.Then, the IGWO-BP prediction model is established, and the prediction model is verified by using the laboratory leakage signal.The results show that the proposed prediction model can effectively improve the prediction precision of pipeline micro-leakage aperture.更多还原