Impact Analysis of the Memristor Failure on Real-Time Control System of Robotic Arm

In this paper, the failure mechanism of memristor is analyzed to study the influence of its malfunction on the real-time control system of the robotic arm. In the real-time control system of robotic arm which is equipped with the memristor RBF neural network as the main controller, the failure state and the number of different memristors are assumed and combined, and the velocity value of the robotic arm node 2 is selected as the research test point. The Matlab simulation technology is used to study the operation process of real-time control system of the robotic arm. The simulation results show that the memristor RBF neural network can reconstruct itself to accommodate the failure of the memristor within the allowable range of memristor resistance, but its potential threat is not eliminated, that is, the failure of the memristor is not repaired. Moreover, when the memristor is in open failure state, the operation of the real-time control system of the robotic arm is irretrievable if there is no good control strategy. This research is of great significance in the promotion and application of memristor neural circuits in real-time control-system, especially in the design of control strategies for real-time control systems of robotic arms.