The Study Of The Mechanism Of Situational Decision In Human-Machine Interaction

This paper focuses on the decision-making strategy of people in specific human-machine situations.In the era of rapid development of intelligent products,it is very meaningful for us to figure out the human decision strategy in human-machine interaction,which is of great significance for system security and user experience.This article lay emphasis on the law of human decision-making in the intelligent system.The first chapter first demonstrates the research background and significance of the thesis,as well as the existing problems in the related research.The second chapter introduces some theories and knowledge related to situational decision-making.The third chapter conducts an experiment on the interaction between two human-recommendation system,studying the interaction between the human and the system from the two angles of the dynamic and the static.We come to the following conclusions:(1)The system presentation time and system intervening time have a significant influence on the performance of subjects(2)Group decision is more easily obeyed than individual decision rules by recommendation system(3)System refresh rate of alternatives to the selection of subjects have significant effect(4)The refreshing time and the intervening time of the system have a significant influence on the situational awareness,and their interaction has significant influence on the SUS score.The fourth chapter conducts a experiment in the field of driving simulation,simulating intelligent driving,which draw the following conclusions:(1)When the cognitive load is high,the degree of self-confidence and their interaction have significant influence on the switch from the manual to the automatic driving(2)In the experiment of rainfall change,the interaction of gender,degree of confidence for the manual switch to automatic driving has a significant impact,and the degree of confidence has a significant impact on automatic switching to manual.(3)In visibility experiments,the degree of confidence has a significant impact on manual switching to the automatic.In the fifth chapter,we use linear regression model,decision tree,multilayer perceptron and radial kernel function to build intelligent driving decision making models respectively,and validate them with the features inside models.The last chapter puts forward its own improvement opinions and prospects.