The advancement of industrial automation technology has made safety monitoring and efficiency optimization of chemical plant environments a research hotspot. This study proposed a real-time path planning and dynamic obstacle avoidance strategy for inspection robots based on Simultaneous Localization and Mapping (SLAM) technology in response to the complex and ever-changing production environment in chemical plants. This method utilized an improved SLAM algorithm to achieve high-precision map construction of the chemical plant environment, and combined deep learning technology to optimize path planning algorithms, enabling robots to quickly perform path planning and re-planning in diverse environments, and achieve efficient dynamic obstacle avoidance by integrating multiple sensor data. In addition, a cloud computing-based data processing framework was developed to achieve real-time data processing and decision support, significantly improving the operational efficiency and security of robots. From the perspective of obstacle avoidance reaction time and obstacle avoidance success rate, robots showed fast reaction speed and high obstacle avoidance success rate in all experiments, with an obstacle avoidance success rate of 100%. The research results of this article have an important practical significance for improving the automation level of chemical enterprises, reducing production costs, and reducing accident risks. It also provides new research ideas and methods for the development of robotics technology in complex industrial environments in China. Through the research of this project, new ideas are provided for the safety production of the chemical industry, while also promoting technological progress in this field and promoting its widespread application in the chemical industry.