Sou Hu Cai Jing·2026-01-10 03:58Group 1 - The core point of the news is that China National Petroleum Corporation (CNPC) has applied for a patent related to model training methods and VSP wavefield separation methods, indicating a focus on enhancing oil and gas exploration technologies [1] - The patent application, published as CN121303232A, was filed on July 2024 and aims to utilize deep learning techniques for high-precision VSP wavefield separation [1] - The method involves constructing a deep sparse autoencoder network to process VSP sample data, enabling the separation of different wavefields based on gradient features [1] Group 2 - CNPC was established in 1999 and is primarily engaged in oil and gas extraction, with a registered capital of approximately 18.3 billion RMB [2] - The company has made investments in 1,296 enterprises and participated in 443 bidding projects, showcasing its extensive involvement in the industry [2] - CNPC holds 5000 patent records and 38 trademark information entries, along with 168 administrative licenses, reflecting its active innovation and regulatory compliance [2]