In the emergency scenario of abnormal high speed railway operation, dispatchers need to quickly and correctly handle the emergency response to the emergencies, which plays a decisive role in the safe operation of trains. The cognitive decision-making process of high speed railway dispatchers was took as the research object, using the laboratory simulation experiment platform and facing the emergency scenario of abnormal driving, the structure of the cognitive decision-making mechanism model and the data collection method corresponding to the measurement model were designed. Based on the information processing model, combined with the concept of perceptual load theory and situational awareness, adding elements of mental load, a cognitive decision-making mechanism model for high speed railway dispatchers was established. Aiming at the evaluation of mental load, a particle swarm optimization algorithm for joint optimization was designed and adopted to optimize the training process based on multi-source physiological signals, comprehensively using signal processing theory and mathematical methods to extract the data characteristics. Fitting the structural equation model and studying the relationship between mental load and attention and working memory are of more practical value for the later optimization of dispatcher's workload and job adaptability.