[论文解读] AirPhyNet: Harnessing Physics-Guided Neural Networks for Air Quality Prediction

Kethmi Hirushini Hettige, Jiahao Ji+4 more|arXiv (Cornell University)|Feb 6, 2024

Air Quality Monitoring and Forecasting被引用 5

一句话总结

AirPhyNet 将扩散-对流物理引入神经网络，用于预测 PM2.5，在数据稀疏情况下尤为提升，与先进基线相比表现更好。

ABSTRACT

Air quality prediction and modelling plays a pivotal role in public health and environment management, for individuals and authorities to make informed decisions. Although traditional data-driven models have shown promise in this domain, their long-term prediction accuracy can be limited, especially in scenarios with sparse or incomplete data and they often rely on black-box deep learning structures that lack solid physical foundation leading to reduced transparency and interpretability in predictions. To address these limitations, this paper presents a novel approach named Physics guided Neural Network for Air Quality Prediction (AirPhyNet). Specifically, we leverage two well-established physics principles of air particle movement (diffusion and advection) by representing them as differential equation networks. Then, we utilize a graph structure to integrate physics knowledge into a neural network architecture and exploit latent representations to capture spatio-temporal relationships within the air quality data. Experiments on two real-world benchmark datasets demonstrate that AirPhyNet outperforms state-of-the-art models for different testing scenarios including different lead time (24h, 48h, 72h), sparse data and sudden change prediction, achieving reduction in prediction errors up to 10%. Moreover, a case study further validates that our model captures underlying physical processes of particle movement and generates accurate predictions with real physical meaning.

研究动机与目标

• 通过将物理学与数据驱动学习相结合，激励准确的户外 PM2.5 预测。
• 将基本的空气传输物理（扩散与对流）融入深度学习框架。
• 提高预测准确性和可解释性，尤其在数据稀疏和突变情形下。

提出的方法

• 使用 RNN 编码器（GRU）对历史 PM2.5 与天气数据进行编码。
• 将传感器网络中的扩散与对流建模为带图拉普拉斯的微分方程网络。
• 通过门控机制融合扩散和对流动力学，形成扩散-对流微分方程（DE）。
• 用由GNN启发的神经 ODE 对 DE 网络参数化，以随时间演化潜在状态。
• 将潜在预测解码为未来时 horizon 的 PM2.5 浓度（最多 72 小时）。
• 端到端训练，通过最小化预测值与真实浓度之间的 MAE 来优化。

Figure 1: The overall architecture of AirPhyNet framework consisting of a RNN Encoder, GNN based DE Network and a Decoder. Heatmap indicate the PM2.5 concentrations while the dashed arrows represent the air pollutant transport between nodes due to diffusion and advection.

实验结果

研究问题

• RQ1 物理引导的深度学习（扩散-对流）能否提高户外 PM2.5 预测，相较于纯数据驱动模型？
• RQ2 物理信息组件在数据稀疏和突变情景中是否提升性能？
• RQ3 提出的基于 GNN 的微分方程网络是否能有效捕捉传感器网络上的时空传输动力学？
• RQ4 将预测与潜在扩散和对流过程联系起来的可解释性收益是什么？

主要发现

• AirPhyNet 在北京和深圳数据集的 24h、48h、72h 预测中始终优于基线。
• 在全数据设定下，平均相比第二佳方法在 MAE 提升 3.7%，在 RMSE 提升 6.1%。
• 在数据稀疏设置下，AirPhyNet 相比竞争对手平均在 MAE 提升 8.3%，在 RMSE 提升 8.1%。
• 模型在突变情景下的预测误差减少高达 8%。
• 消融研究显示结合扩散和对流的优势，以及使用 GRU 编码器优于基于 TCN 的编码器。
• 案例研究可视化展示了预测中的物理意义的扩散与对流模式。

(a) Effect of Physical Knowledge in AirPhyNet