Magnetic resonance imaging (MRI) is a widely used medical imaging modality. However, due to the limitations in hardware, scan time, and throughput, it is often clinically challenging to obtain high-quality MR images. In this article, we propose a method of using artificial intelligence to expand the channel to achieve the effect of increasing the virtual coil. The main feature of our work is utilizing dummy variable technology to expand the channel in both the image and k-space domains. The high-dimensional information formed by channel expansion is used as the prior information of parallel imaging to improve the reconstruction effect of parallel imaging. Two features are introduced, namely variable enhancement and sum of squares (SOS) objective function. Variable argumentation provides the network with more high-dimensional prior information, which is helpful for the network to extract the deep feature in-formation of the image. The SOS objective function is employed to solve the problem that k-space data is difficult to train while speeding up the convergence speed. Ablation studies and experimental results demonstrate that our method achieves significantly higher image reconstruction performance than current state-of-the-art techniques.