Research and Analysis of Optimization Scheduling and Fault Perception Technology for Full Performance Test System

Based on research on the task scheduling control architecture and optimization strategy of the full performance testing system, we propose adaptive hybrid genetic algorithm and reinforcement learning algorithm based on different business requirements. The adaptive hybrid genetic algorithm has good stability and is suitable for demand scheduling at medium scale, while the reinforcement learning algorithm has a decoding time close to real-time and is suitable for demand scheduling at super large scale. Based on research on fault perception and risk assessment techniques based on big data, propose outlier detection algorithms, we use outlier detection algorithms combined with deep learning to help detect and warn of abnormal waveforms in test equipment during fault detection in advance, completing the sorting of risk assessment strategies for full performance test system equipment.