| 地下水硝酸盐污染抽出处理优化方法模拟研究 |
| 摘要点击 2563 全文点击 1326 投稿时间：2013-12-24 修订日期：2014-02-27 |
| 查看HTML全文
查看全文 查看/发表评论 下载PDF阅读器 |
| 中文关键词 地下水 硝酸盐 抽出处理 模拟优化 遗传算法 模拟退火法 |
| 英文关键词 groundwater nitrate pump-and-treat simulation and optimization genetic algorithm simulated annealing |
|
| 中文摘要 |
| 以北京某场地生活垃圾填埋场硝酸盐污染地下水为研究对象，针对抽出处理技术，利用遗传算法（GA）和模拟退火法（SA）对地下水硝酸盐污染羽治理区中布置18口备选井方案进行优化. 优化目标函数最小化治理成本，确定最佳井数及其井的位置和抽水速率，同时达到在100 d内将研究区硝酸盐浓度降低至10 mg·L-1的治理目标. GA优化结果为8号井优化抽水速率为155 m3·d-1，14号井优化抽水速率为10 m3·d-1. SA优化结果为8号井优化抽水速率为82 m3·d-1，14号井优化抽水速率为39 m3·d-1. 基于GA和SA优化抽水速率，硝酸盐总量去除率分别为76.89% 和84.92%. 优化结果说明，最佳井的位置应位于硝酸盐污染羽中游及中下游的中轴线上，且中游抽水速率比下游要大. 两种优化算法对比表明SA优化系统治理成本比GA节省6.8%，且波动性小，更收敛. |
| 英文摘要 |
| To study the groundwater polluted by nitrate in a landfill site in Beijing, with the pump-and-treat technology, genetic algorithms (GA) and simulated annealing (SA) were used to optimize the cases of 18 planned pumping wells in the groundwater nitrate plume. The optimization method was run to determine the minimum total costs, as a result of the optimal number of wells with the optimal locations and pumping rates. The results of optimization showed that, both GA and SA optimization conditions could reach the desired results, which means that the concentration of nitrate could be reduced to 10 mg·L-1 after 100 days pumping. However, under the GA optimization conditions, the pumping rates of well 8 and well 14 were 155 m3·d-1 and 10 m3·d-1, respectively. In contrast, under the SA optimization conditions, the pumping rate of well 8 was 82 m3·d-1, and that of well 14 was 39 m3·d-1. Based on the GA and SA optimized pumping rate, the total mass removal rates of nitrate were 76.89% and 84.92%, respectively. The results showed that the best location of the well was in the central axis of either midstream or downstream of the nitrate contamination plume. The pumping rate of midstream was larger than downstream. Comparison of the two optimization algorithms showed that the optimized system management cost of SA was 6.8%, which was lower than that of GA. Meanwhile, SA was easily convergent with smaller volatility. |
|
|
|
