Lightweight Detection Algorithm for Substation Equipment Based on Improved YOLOv8s

With the increasing demand for electricity and the growing complexity of power systems, deep learning has been widely applied in power equipment detection. However, existing models often suffer from excessive parameters and high computational complexity, making them difficult to deploy efficiently on mobile devices with limited computing resources. To address this issue, we propose a lightweight detection algorithm for substation equipment based on improved YOLOv8s. Firstly, we replace standard convolution modules with lightweight feature extraction modules to reduce computational overhead while maintaining accuracy. Secondly, we embed lightweight attention modules into each layer of the backbone network to enhance feature extraction capability. Thirdly, we employ depthwise separable convolution and C3Ghost modules to replace corresponding modules in the original feature fusion network, further improving detection speed. Finally, we design a novel lightweight parameter sharing detection head to reduce both parameter count and computational complexity. Experimental results demonstrate that our proposed model not only enhances detection accuracy but also significantly reduces model parameters and computational costs, which makes it more suitable for deployment on mobile devices with limited computing resources.