This paper proposes a neural network-based approach for detecting network threats in chemical plants. Chemical plants are vulnerable to various types of network attacks, including cyber-physical attacks, insider threats, and malware attacks. The proposed approach utilizes a deep neural network model to analyze network traffic and identify anomalous behavior. The model is trained on a large dataset of normal and malicious network traffic and is able to accurately detect network threats in real-time. The results demonstrate the effectiveness of the proposed approach in detecting various types of network threats.