Design and Simulation Verification of Intelligent Perception and Adaptive Control System for Control Rod Drive Mechanism

抄録

Nuclear power plant status monitoring and equipment fault diagnosis are important guarantees for the safe operation of nuclear power plants. With the development of artificial intelligence technology, fault diagnosis of nuclear reactor equipment has entered the stage of intelligent diagnosis and prediction. The control rod drive mechanism (CRDM) is the actuator of the reactor control and nuclear safety protection system. In the normal state, CRDM realizes the stepping movement of the control rod through the alternating action of the three sets of coil currents. When an abnormality in the coil current is detected, it means that the CRDM movement fails, and in more serious cases, it may cause the control rod to accidentally drop. Therefore, in order to prevent the occurrence of control rod stuck, sliding rod and driving failure in the reactor and to ensure the safe and economic operation of the reactor, a necessary measure is to effectively monitor the operation status of the CRDM. In this paper, the overall architecture of an intelligent perception and adaptive control system for CRDM is studied. By analyzing the multi-condition and multi-mode coupling relationship in the time dimension and spatial correlation between the coil current and vibration signals, a condition discrimination and fault diagnosis model for CRDM based on hybrid network of CNN+BiGRU is established. Furthermore, a digital adaptive control model of control rod drive system is built based on the real-time simulation system dSPACE. Through functional verification, the system can perform real-time perception and identification of normal operation conditions, fault diagnosis and adaptive control based on the coil current and vibration signals. The recognition accuracy rate for various operation conditions reaches more than 99.35%, and the control accuracy of the sequential current is improved around ±4% of the rated value under various external environment changes, providing a supportive reference for the subsequent nuclear reactor intelligent equipment development.