【摘要】 针对日益严峻的城镇供水管网漏损问题，对漏失量与用户水量的变化特征及源信号相互独立特征进行深入分析，并基于Fast ICA算法建立漏失量估算模型，对观测漏失信号进行白化处理、寻优迭代及信号幅值还原；然后，以北方某生活小区为研究区域，构建供水管网实验模型，在实验室条件下验证Fast ICA算法用于管网漏失量估算的可行性；最后，将Fast ICA漏失量估算模型应用于DS山庄工程实例的供水管网漏失分析中，并与小波变换理论在实际供水环境下的适用性进行比较。结果表明，与小波变换算法相比，Fast ICA模型计算出的漏失量与真实漏失量相对误差更小，变化趋势相似性更高。更多还原

【Abstract】 In response to the increasingly severe issue of leakage in urban water supply network, an in?depth analysis was conducted to examine the variation characteristics of leakage and user water consumption, as well as the mutual independence of source signals. A leakage estimation model based on fast independent component analysis(Fast ICA) algorithm was developed. Subsequently, whitening processing, optimization iteration, and signal amplitude restoration were performed on the observed leakage signals. A residential community in the northern region was selected as the case study site, and an experimental model of the water supply network was constructed. The feasibility of applying the Fast ICA algorithm for leakage estimation was validated under controlled laboratory conditions. Furthermore, the Fast ICA leakage estimation model was applied to the leakage analysis of the water supply network in the DS mountain villa project, and its applicability was compared with that of the wavelet transform theory in a real water supply setting. In comparison with the wavelet transform algorithm, the Fast ICA model achieved a smaller relative error between the estimated and actual leakage, as well as a higher degree of similarity in the trend of variation.更多还原