Research On Evaluation Method Of Human-Machine Interface Of Power Plant Based On HNN

The objective of human-machine interface study is to design and evaluate the machine's display terminal, the control equipments and environment, in order to make the human-machine system run with high-efficiency, safely and comfortably. Many mathematic models have been applied in the human-machine interface evaluation field, such as integration index method, fuzzy mathematics method, gray connection analysis theory and so on. There is a common defect in these traditional methods, which is a lack of self-learning ability, and random in the evaluation process, the estimators' subjective uncertainty and fuzzy cognition will also affect the evaluation results.To solve such problems, the human neural network (HNN) theory is firstly put forward in the field of human-machine interface evaluation. Based on the HNN' s self-learning process of the training example and its peculiar error-allowing character, it deals with the non-numerical indices of human-machine interface elements, so as to realize subjective and objective evaluation. The HNN method not only avoids the subjective evaluation of the indices weight, but also eliminates a lot of artificial factor, therefore results are guaranteed to be objective and exact.Due to the current research status on the human-machine interface evaluation, several works are carried out in this paper:1. An evaluation indices system on console human-machine interface is brought forward. Based on the previous research work of the group, each kind of equipment's (include vision display equipment, control equipment, console and chair) human machine engineering objective and subjective indices and standards are summarized and established.2. Human-machine interface evaluation model based on HNN is put forward. The evaluation interface which looks friendly and can be used easily is designed using MATLAB GUI.