The engine room of a ship is the heart of the cabin. The power plant and electrical facilities located in the engine room provide all kinds of energy for the ship to ensure the normal operation of the ship, ensure the normal life of the crew and help the crew to complete various operations. The engine room is an important place for the generation, transmission and consumption of all kinds of energy on the ship, and is the core of the whole ship. The generalized regression neural network is a kind of radial basis function neural network, which has great advantages in solving nonlinear problems and is not affected by multiple collinearity. The purpose of this paper is to study the intelligent monitoring of cabin temperature based on recurrent neural network. In the experiment, the temperature detection system of the ship engine room is designed, the resistance temperature sensor is calculated, and the data of the ship engine room temperature prediction system design and model training process are investigated.