
水质指标组合与河流溶解氧预测精度关系的研究
宦娟, 陈波, 徐宪根, 李明宝, 杨贝尔, 史兵, 张勤兰
水质指标组合与河流溶解氧预测精度关系的研究
Research on the Relationship between the Combination of Water Quality Indicators and the Accuracy of River Dissolved Oxygen Prediction
提出了一种探究水质指标组合与河流溶解氧预测精度关系的方法。首先运用XGBoost模型计算水质指标特征重要性分值,然后基于贪心规则和水质指标特征重要性分值,排列出8种水质指标组合,最后使用BP神经网络对8种水质指标组合进行溶解氧预测。实验结果表明,pH、水温、电导率、氨氮是影响溶解氧预测的4个关键指标;在排列出的8种水质指标组合中,pH、水温、电导率、氨氮、浊度、高锰酸盐指数是溶解氧预测精度最高的输入指标组合。通过穷举所有水质指标组合进行实验分析,证明该方法有效可行且时间复杂度更低,可用于选取溶解氧预测精度高的输入指标组合,提升溶解氧的预测精度。
This paper proposes a way to explore the relationship between the combination of water quality indicators and the accuracy of river dissolved oxygen prediction. First, the XGBoost model is used to calculate the water quality index feature importance score, and then based on the greedy rule and the water quality index feature importance score, 8 water quality index combinations are arranged. Finally, the BP neural network is used to predict dissolved oxygen for the 8 water quality index combinations. Experimental results show that pH, water temperature, conductivity, and ammonia nitrogen are the four key indicators that affect the prediction of dissolved oxygen. Among the 8 combinations of water quality indicators arranged, pH, water temperature, conductivity, ammonia nitrogen, turbidity, and CODmn are the most accurate combinations of input indicators for the prediction of dissolved oxygen. Experimental analysis by exhaustively enumerating all water quality indicator combinations proves that the method is effective and feasible with lower time complexity, and can be used to select a combination of input indicators with high accuracy of dissolved oxygen prediction to improve the accuracy of dissolved oxygen prediction.
水质管理 / 水质指标组合 / 溶解氧预测 / 贪心规则 / 特征重要性 {{custom_keyword}} /
water quality management / water quality factor combination / dissolved oxygen prediction / greedy rules / feature importance {{custom_keyword}} /
表1 样本数据Tab.1 Sample data |
时间 | 氨氮/(mg·L-1) | 总磷/(mg·L-1) | 高锰酸盐指数/ (mg·L-1) | 总氮/(mg·L-1) | pH | 水温/℃ | 溶解氧/(mg·L-1) | 电导率/ (μS·cm-1) | 浊度/ntu |
---|---|---|---|---|---|---|---|---|---|
01-01 00∶00 | 1.48 | 0.186 | 3.38 | 4.41 | 7.69 | 16.5 | 8.36 | 492.6 | 84.7 |
01-01 04∶00 | 1.48 | 0.186 | 3.07 | 7.41 | 7.80 | 16.7 | 8.91 | 483.1 | 84.7 |
01-01 08∶00 | 1.29 | 0.329 | 3.20 | 4.79 | 7.71 | 16.5 | 9.49 | 478.4 | 84.7 |
01-01 12∶00 | 1.31 | 0.205 | 3.01 | 3.18 | 7.72 | 16.5 | 9.21 | 473.8 | 84.7 |
01-01 16∶00 | 1.51 | 0.206 | 2.73 | 3.73 | 7.68 | 16.5 | 8.93 | 463.0 | 84.7 |
01-01 20∶00 | 1.52 | 0.192 | 3.16 | 1.63 | 7.68 | 16.5 | 8.79 | 459.6 | 84.7 |
01-02 00∶00 | 1.34 | 0.189 | 2.83 | 1.63 | 7.71 | 16.5 | 9.02 | 467.6 | 84.7 |
01-02 04∶00 | 1.22 | 0.208 | 2.91 | 3.67 | 7.88 | 16.9 | 9.21 | 469.0 | 84.7 |
表2 水质指标特征重要性分值Tab.2 Water quality index feature importance score |
pH | 水温 | 电导率 | 氨氮 | 浊度 | 高锰酸盐指数 | 总磷 | 总氮 |
---|---|---|---|---|---|---|---|
0.533 2 | 0.134 6 | 0.098 8 | 0.083 3 | 0.062 8 | 0.038 9 | 0.029 4 | 0.018 7 |
表3 水质指标组合结果Tab.3 Water quality index combination result |
组合 | pH | 水温 | 电导率 | 氨氮 | 浊度 | 高锰酸盐指数 | 总磷 | 总氮 |
---|---|---|---|---|---|---|---|---|
1 | √ | |||||||
2 | √ | √ | ||||||
3 | √ | √ | √ | |||||
4 | √ | √ | √ | √ | ||||
5 | √ | √ | √ | √ | √ | |||
6 | √ | √ | √ | √ | √ | √ | ||
7 | √ | √ | √ | √ | √ | √ | √ | |
8 | √ | √ | √ | √ | √ | √ | √ | √ |
表4 8个分组的组合信息Tab.4 Combination information of 8 groups |
分组1 | 分组2 | 分组3 | 分组4 | 分组5 | 分组6 | 分组7 | 分组8 | |
---|---|---|---|---|---|---|---|---|
m取值 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
组合个数 | 8 | 28 | 56 | 70 | 56 | 28 | 8 | 1 |
组合号码集合 | 组合1~8 | 组合9~36 | 组合37~92 | 组合93~162 | 组合163~218 | 组合219~246 | 组合247~254 | 组合255 |
我们方法排列的组合号码 | 组合1 | 组合9 | 组合37 | 组合93 | 组合163 | 组合219 | 组合247 | 组合255 |
表5 255种水质指标组合在4种评价指标下的排名结果Tab.5 Ranking results of 255 combinations of water quality indicators under 4 evaluation indicators |
排名 | MSE | MAE | RMSE | MAPE | ||||
---|---|---|---|---|---|---|---|---|
误差值 | 组合 | 误差值 | 组合 | 误差值 | 组合 | 误差值 | 组合 | |
1 | 0.926 4 | 219 | 0.748 5 | 219 | 0.962 5 | 219 | 14.215 2 | 164 |
2 | 0.994 5 | 93 | 0.773 6 | 93 | 0.997 2 | 93 | 14.350 3 | 219 |
3 | 0.994 9 | 220 | 0.779 4 | 164 | 0.997 4 | 220 | 14.484 2 | 224 |
4 | 1.008 2 | 255 | 0.779 6 | 220 | 1.004 1 | 255 | 14.561 2 | 220 |
5 | 1.013 5 | 249 | 0.781 8 | 224 | 1.006 7 | 249 | 14.567 1 | 93 |
1 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
2 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
3 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
4 |
刘双印,徐龙琴,李道亮,等. 基于时间相似数据的支持向量机水质溶解氧在线预测[J]. 农业工程学报, 2014,30(3):155-162.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
5 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
6 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
7 |
施珮,匡亮,袁永明,等.基于改进极限学习机的水体溶解氧预测方法[J].农业工程学报,2020,36(19):225-232.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
8 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
9 |
陈英义,程倩倩,方晓敏,等.主成分分析和长短时记忆神经网络预测水产养殖水体溶解氧[J].农业工程学报,2018,34(17):183-191.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
10 |
李占山,刘兆赓. 基于 XGBoost 的特征选择算法[J]. 通信学报, 2019,40(10):101-108.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
11 |
董寅冬,任福继,李春彬. 基于线性核主成分分析和 XGBoost 的脑电情感识别[J]. 光电工程, 2021,48(2).
{{custom_citation.content}}
{{custom_citation.annotation}}
|
12 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
13 |
孙宝磊,孙暠,张朝能,等. 基于BP神经网络的大气污染物浓度预测[J]. 环境科学学报, 2017,37(5):1 864-1 871.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
14 |
陈英义,程倩倩,成艳君,等. 基于GA-BP神经网络的池塘养殖水温短期预测系统[J]. 农业机械学报, 2017,48(8):172-178.
{{custom_citation.content}}
{{custom_citation.annotation}}
|
15 |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
{{custom_ref.label}} |
{{custom_citation.content}}
{{custom_citation.annotation}}
|
/
〈 |
|
〉 |