Design and Application of Process Object Intelligent Model

Multi-order inertial system is a common controlled object in industrial field. The conventional modeling method is least squares identification algorithm. However, it is necessary to determine the order of the object model before using the excitation data to identify it. This requires prior knowledge, and the identified model is prone to large deviations. This paper presents an intelligent modeling method for multi-order inertial objects. Firstly, the third-order inertial objects are discretized and the DNN deep learning network is embedded in the discrete structure to establish the intelligent model structure. Secondly, in order to ensure that the third-order inertial system can contain the object’s inertial time, three third-order inertial systems are set up in the intelligent model to estimate the inertial time of the controlled object, and the inertial time of the intermediate inertial system is set to be close to the object’s inertial time. At the same time, a margin of (± 50%) is given and set in the other two inertial systems. Finally, the pseudo-random excitation signal is added to the open-loop or closed-loop system of the object, and the input and output data are sent to the intelligent model for training, so the accurate identification model can be obtained. The validity of the intelligent model is verified by testing the simulation data. This method has great practical significance for the design and application of intelligent model.