To address the need for multi-component gas detection in complex environments, this paper presents the design and implementation of a portable multi-channel gas detection system based on metal-oxide semiconductor (MOS) sensors. The system integrates an eight-channel sensor array, a high-precision signal-acquisition circuit, and a low-power hardware. A multi-branch convolutional neural network combined with a bidirectional long short-term memory network is employed to achieve automatic feature extraction and temporal modeling of multi-channel signals. Experiments using CO, C₂H₅OH, and their interfering gases and mixtures as the research subjects demonstrate that the system exhibits linear responses and high detection accuracy across different ranges of concentration. Comparative tests validate that the sensor-fusion strategy improves classification accuracy, enhances robustness, and increases adaptability to complex environments, with the classification accuracy for mixed gases reaching up to 100%. This study provides a reliable technical basis and practical reference for multi-component gas detection in environmental monitoring, industrial safety, and public health applications.