Optimal design methods for a digital human–computer interface based on human reliability in a nuclear power plant part 1: Optimization method for monitoring unit layout
Introduction
Currently, most nuclear power plants (NPPs) use digital control systems (DCSs). In DCSs, the operators obtain information about the state of the plant from the human–computer interface and respond accordingly. Operating experience shows that human–computer interface design influences the acquisition of information; the time of feedback input and the results are also influenced, leading to possible human factor accidents. For example, 60–90% of all NPP accidents involve human actions (He and Huang, 2007).
There are two ways to decrease human errors, namely, improve operator adaptive capabilities in addressing accidents or optimize the digital human–computer interface based on human reliability. The authors of this paper propose an optimization method to decrease human factor accidents caused by the defects of the human–computer interface in a NPP.
Studies have been completed on the design of the human–machine interface, but there are few method for optimizing this interface at NPPs. In 1964, (Bonney et al., 1997) first proposed a human–machine interface design that considered the frequency, importance, operation order and spatial layout of components. Further, Auxiliary control panel design programs were subsequently designed in 1973, 1974 and 1977. In 1981, human, machine and context factors about the design of a human–machine interface were involved; human–machine interface layouts and human fuzzy control model were built according to visual tracing theory (Long et al., 1995). Pulat and Ayoub successfully developed computer auxiliary software that could automatically design and modify a layer of human–machine interface (Pault, 1985). In 1989, an international standard was enacted regarding how to define control room design at NPPs. The standard proposed a new concept for control room systems and defined control room systems to include the human–machine interface, the control room operator, the operational regulations, the training outline and the relative facilities. In 1990, Zhang proposed a design method for the human–machine interface. The method divided the human–machine interface into many parts, such as session management, system management, operation support, and error correction. A human–machine interface design model of mechanical systems was proposed by Song et al., in 2006. The research considered each component feature, the design rule of the human–machine interface, and the object optimal function in terms of optimization theory. Furthermore, it established an optimal mathematics model of the human–machine interface according to the geometric parameters of every component; it considered the component-matching degree of the human–machine interface as a single object function. In 2008, Jin proposed an optimal layout of the human–machine interface. The research used the knowledge notation and oriented-object method, established an optimization inference model and the inference system of fuzzy method, and simulated the human–machine interface.
This research has the following three distinctive features compared with the aforementioned studies: (1) this paper focuses on an optimization method of the human–computer interface in a NPP rather than a design method; (2) the object of this research is the digital human–computer interface; and (3) the optimal criteria are based on human reliability.
Section snippets
Optimization process and method of monitoring unit layout
A monitoring unit (also known as a function block) is considered a similar component that carries out tasks of the same type. The authors utilize a linear reversal genetic hybridization algorithm and the Bayesian network method to optimize the digital human–computer interface layout. The proposed optimization process is shown in Fig. 1.
Human reliability method
Fig. 1 shows that when a newly evolved sequence is generated for the layout of the digital human–computer interface in a NPP, the sequence must be evaluated. However, the reliability of the monitoring process is influenced by the digital human–computer interface. The monitoring process has two stages, namely, the process of moving from one object to another and the process of obtaining information about the current object. Then, the evaluation criterion of each layout sequence in the monitoring
Purpose of the experiment
The purpose of the experiment is to prove the convergence, stability and sensitivity of the proposed method used to optimize the digital human–computer interface. The experiment of the human–computer interface layout is conducted according to the proposed layout-optimization method.
Background
The spurious safety injection event in a NPP is used for illustration. Many human–computer interfaces will be involved in the process, and this experiment takes the main interface of the spurious safety inject event
Conclusions
The authors propose an algorithm based on linear reverse genetic hybridization and a method to optimize the layout of a digital human–computer interface based on Bayesian network theory.
The following conclusions are obtained by analyzing the experiment results: (1) the optimal layout of the main digital human–computer interface in a spurious safety inject event at a NPP is obtained; (2) the error probability tends to stabilize when the iterations of the main human–computer interface layout
Acknowledgements
This work is supported in part by three different National Nature Science Foundations (70873040, 71071051, 71301069) of China, the construct program of the key discipline in Hunan province of China-management science and engineering discipline, the Scientific Research Fund of the Hunan Provincial Education Department (14C0974) of China, the Hunan Provincial Natural Science Foundation of China (14JJ7046) and the Scientific Research Foundation of the University of South China (2012XQD54).
References (15)
- et al.
A discrete-time bayesian network reliability modeling and analysis framework
Reliab. Eng. Syst. Saf.
(2005) - et al.
Markov reliability model research of monitoring process in digital main control room of nuclear power plant
Saf. Sci.
(2011) - et al.
Association rules analysis of human factor events based on statistics method in digital nuclear power plant
Saf. Sci.
(2011) - et al.
A probabilistic cognitive simulator for HRA studies
Reliab. Eng. Syst. Saf.
(2007) - et al.
CAPABLE.A computer program to layout controls and panels
Ergonomics
(1997) Cognitive Modeling and Dynamic Probabilistic Simulation of Operating Crew Response to Complex System Accident(ADA-IDACrew)
(1999)- et al.
Human Reliability Analysis: Theory, Methods and Applications in Industrial System
(2007)
Cited by (6)
Optimal design method for a digital human–computer interface based on human reliability in a nuclear power plant. Part 3: Optimization method for interface task layout
2016, Annals of Nuclear EnergyCitation Excerpt :Bellet et al. discussed a research programme for virtual design of driving assistances in order to develop an integrative simulation platform that can support a human centered design method, the paper focuses on a description of the driver model implemented on the a virtual platform (2012). To solve the layout of relative positions among different functional blocks in a digital human-computer interface of a NPP, Jiang et al. presented an optimization method involving a liner reversal genetic hybridization method and Bayesian method, and showed the proposed method had good performance (2015). This paper proposes a shortest moving path optimization algorithm with dynamic flag to solve the optimal layout of interface task.
Evaluation of human machine interface (HMI) on a digital and analog control room in nuclear power plants using a fuzzy logic approach
2017, Fuzzy Systems: Concepts, Methodologies, Tools, and ApplicationsDigital human-machine interface design for marine nuclear power plant based on task-oriented
2017, International Conference on Nuclear Engineering, Proceedings, ICONEEvaluation of Human Machine Interface (HMI) in Nuclear Power Plants with Fuzzy Logic method
2016, IISA 2016 - 7th International Conference on Information, Intelligence, Systems and Applications