中文摘要：

针对水安全评价的特点，将人工蜂群算法（ABC）引入到BP神经网络中，对其网络结构、初始权值和阈值进行优化，通过将BP神经网络参数的优化过程转化为蜜蜂寻找最佳蜜源的过程，建立了基于人工蜂群算法优化的BP网络。选取11个水安全评价指标，以阿什河流域2011年阿城观测断面实测数据为例，对其2011年的水安全状况进行了全面的评价。研究结果表明：阿什河流域2011年水安全等级普遍较差，均在Ⅲ级以下，与实际情况相符，且评价结果具有很强的稳定性和鲁棒性。因此，采用ABC-BP模型对流域水安全进行评价是一种行之有效的方法，研究成果为流域水安全评价提供了新的思路和方法。

英文摘要：

Aiming at the character of water safety evaluation, network structure, initial weights and threshold of BP neutral network can be optimized by Artificial Bee Colony （ABC for short）. Through transferring parameter optimization process of BP neutral network to finding nectar source process of bee, the BP neutral network based on ABC can be built. Selecting 11 indexes and taking Acheng observation section measured data of Ash river basin in 2011 as an example, the water safety can be evaluated. The results show that the water safety degree is very low of Ash river basin in 2011 and all below Ⅲ, which is coincided with the fact and the results have good stability and robustness. Therefore, it is an effective way of adopting ABC-BP model to evaluate river basin water safety and the results can provide a new way and method for basin water safety evaluation.