Abstract: In view of the current evaluation status, this paper proposes a method for evaluating human-machine interfaces based on human-machine interaction process analysis and human-machine interaction operation analysis, and establishes an evaluation index system suitable for armored vehicle seats accordingly. Finally, the fuzzy comprehensive evaluation method is used to evaluate the seats. Keywords: human-machine interaction; human-machine interface; armored vehicle human-machine interface evaluation is to judge whether the design of the matching relationship between the operator and related displays, controllers and auxiliary components (such as seats) in the human-machine interface meets the requirements of ergonomics based on relevant design principles and the operator's own experience. Human-machine interface evaluation can not only help to judge the overall status of the existing human-machine interface and its constituent elements, but also to discover problems in the design so as to provide timely feedback to the design department for modification. The factors in common evaluation methods are often selected based on experience or collected relevant data, which leads to problems such as low credibility in the selection process of evaluation factors, difficulty in listing all three major elements of human-machine interface, and some listed factors being irrelevant to human-machine interface evaluation. This paper proposes a method for selecting evaluation factors based on driving process and human-machine interaction operation analysis, and studies the evaluation of armored vehicle driver seats. 1. Human-Machine Interface Evaluation Methods The ultimate goal of human-machine interface evaluation is to guide the design of human-machine interfaces, making them both "machine-friendly" and "human-suitable." To achieve these two requirements, the human-centered design must be implemented within the human-machine system, analyzing and researching the three major elements affecting the human-machine interface: human, machine, and environment, and resolving the two issues of human control of the machine and human information reception. The primary task in researching human-machine interface evaluation is to clarify the human-machine-environment factors and corresponding tasks within the human-machine interface. The selection of evaluation factors essentially involves studying the interaction process of human-machine-environment factors and the human-machine interface properties exhibited during this process. Analysis of the human-machine interaction process ensures a rigorous and reasonable determination of the research object. Analyzing human-machine interaction tasks clarifies the interaction process of human-machine-environment factors, revealing that the process of reflecting the human-machine interface properties aligns with human cognitive habits. The human-computer interface evaluation method proposed in this paper first analyzes the human-computer interaction process to determine the research object, that is, to determine the human-computer environment factors and corresponding tasks involved in the human-computer interaction process; secondly, it analyzes the tasks involved in the interaction process, studies the manifestation of the characteristics of each human-computer environment factor in the interaction process, and selects evaluation factors; then, it analyzes the evaluation factors, determines the evaluation indicators, and establishes an evaluation indicator system; finally, it selects an appropriate evaluation method to conduct human-computer interface evaluation (Figure 1). [img=667,260]http://www.chuandong.com/uploadpic/THESIS/2009/5/20090527133929585629.jpg[/img] As can be seen from Figure 1, the specific steps of human-computer interface evaluation are: 1) Analyze the human-computer interaction process. Taking a typical task as the process analysis object, study the task content, list all relevant human-computer environment factors and corresponding tasks in the interaction process, and perform statistics and classification. 2) Analyze the human-computer interaction tasks. Taking the task obtained in step 1) as the analysis object, the interaction process between human characteristics, equipment attributes and environmental conditions is analyzed using a three-dimensional matrix. All factors reflected at the interaction point are the evaluation factors that need to be analyzed in the human-computer interaction process. At the same time, these factors can reflect the interaction between human, machine and environmental factors and their respective characteristics. 3) Apply the four major techniques of cancellation, merging, rearrangement and simplification to analyze and compare the relationship between evaluation factors and determine evaluation factors and evaluation indicators. 4) Determine the hierarchy and structure between indicators and construct an evaluation indicator system. 5) Select an appropriate evaluation method. 6) Evaluate the human-computer interface. 2 Establishment of the evaluation indicator system for armored vehicle driver's seat 2.1 Analysis of the driving process of armored vehicle driver The typical tasks undertaken by the driver are analyzed, and the environmental factors considered are the physical environment [sup][3][/sup], including noise, temperature, lighting and vibration environment. According to statistics, 62 tasks need to be completed, which can be divided into driving tasks, observation tasks and information receiving tasks. 2.2 Human-computer interaction task analysis In the human-computer interaction task analysis, the human characteristics are mainly considered as physical characteristics, physiological characteristics and psychological characteristics. Human physical characteristics are studied from three aspects: geometry (including static and dynamic geometric characteristics), mechanics, and thermodynamics; human physiological characteristics are studied from two aspects: sensory characteristics and physiological adaptability; and human psychological characteristics are studied from the perspective of human psychological processes. A three-dimensional matrix is ​​used to represent the interaction between human physical, physiological, and psychological characteristics and the seat under noise, temperature, lighting, and vibration environmental conditions. The factors manifested at the interaction points are the evaluation factors. Considering the evaluation factors manifested when human characteristics affect the seat under noise conditions, these include: energy consumption, back posture, leg posture, seat height, backrest height, backrest tilt angle, seat width, seat depth, seat tilt angle, and fatigue. 2.3 Establishment of Evaluation Index System for Armored Vehicle Driver's Cabin Seat The relationships between evaluation factors are analyzed and compared, and they are then eliminated, merged, rearranged, and simplified. The selection process of evaluation factors is the process of human characteristics acting on the machine under a certain environment. Combining the selected evaluation factors, the selected indicators are: human geometric characteristics, human mechanical characteristics, and human physiological adaptability. The sub-indicators are: backrest, seat surface, and seat height. The evaluation index system for the driver's seat of an armored vehicle is shown in Table 1. 2.4 Evaluation of the driver's seat of an armored vehicle Fuzzy comprehensive evaluation method is a widely used evaluation method that combines quantitative and qualitative methods and is suitable for multi-level analysis. It has high scientific rationality [sup][4] [/sup]. The following takes the geometric characteristics index of a person in the driver's seat of an armored vehicle as an example. The design parameters are: seat height 310mm, seat length 386mm, seat width 390mm, seat inclination angle 4°, backrest height 180mm, and backrest inclination angle 90°. The comprehensive evaluation steps are as follows: 1) Determine the evaluation index system. The index system in this paper is shown in Table 2. 2) Use the Delphi method and the analytic hierarchy process (AHP) to determine the weight of each index layer (Table 2). 3) The evaluation level is divided into 5 levels, namely V = {"Excellent", "Good", "Average", "Poor", "Inferior"}. The evaluation of the armored compartment is described as a five-level fuzzy evaluation domain {"Excellent", "Good", "Average", "Poor", "Inferior"}, represented by subscripts {5, 4, 3, 2, 1}. The specific standard meanings are as follows: Excellent: The positions of each component are properly arranged, the operation is smooth and unobstructed, only a small movement of the body is needed to complete the operation, there is no discomfort after the operation is completed, the operator does not feel fatigued, the vision is very comfortable, and the operating force is appropriate. Good: The positions of each component are basically properly arranged, there are slight obstacles to the operation, the range of body movement is slightly larger, but there is no obvious discomfort after the operation is completed. Average: The range of body movement is relatively large, the operation feels obstructed, but the operation can basically be completed. Poor: The operation is obviously unreasonable due to the position arrangement, the range of body movement is large, and the operation feels strenuous. Inferior: The operation is very strenuous, the operation is restricted and cannot be completed accurately, the number of misoperations increases, and the correct reaction time increases. 4) Conduct a comprehensive evaluation [sup][5] [/sup]. Step 1: Divide the factor domain (indicator) into s subsets according to a certain attribute. Step 2: Conduct a first-level fuzzy comprehensive evaluation. Perform a single-factor fuzzy comprehensive evaluation for each u. Let the domain of the evaluation rating be V = (V₁, V₂, ..., V_p), and the fuzzy weight vector of each factor in u_i be W_j = (W_j1, W_j2, ..., W_jt). The single-factor evaluation result of u_i is (rows t, columns P). In the formula, r_j represents the membership degree of the t-th factor to the P-th evaluation in the evaluation of the j-th indicator. Then, the single-level evaluation model is: Step 3: Perform a two-level fuzzy comprehensive evaluation. Treating u[sub]i[/sub] as a comprehensive factor and using it as its single-factor evaluation result, we can obtain the membership matrix. If there are still too many ui (i = 1, 2, ..., s) in the first step of partitioning, we can continue to partition to obtain a three-level or higher model. Let the fuzzy weight vector W = (w1, w2, ..., wi) of the comprehensive factors u[sub]i[/sub] (i = 1, 2, ..., s). Then the second-level fuzzy comprehensive evaluation model is: When the seat length is 386mm, the corresponding {u[sub]5[/sub](x), u[sub]4[/sub](x), u[sub]3[/sub](x), u[sub]2[/sub](x), u[sub]1[/sub](x)} = {0, 0, 0.2, 0.8, 0}. Thus, the evaluation matrix R for the seat surface is: Similarly, the evaluation matrix for the seat back is: The corresponding seat height is {u[sub]5[/sub](x), u[sub]4[/sub](x), u[sub]3[/sub](x), u[sub]2[/sub](x), u[sub]1[/sub](x)}={1, 0, 0, 0, 0} B[sub]1[/sub]=A[sub]1[/sub]·B[sub]1[/sub]={0.312,0.3276,0.07208,0.28832,0} B[sub]2[/sub]=A[sub]2[/sub]·R[sub]2[/sub]={1,0,0,0,0} B[sub]3[/sub]=A[sub]3[/sub]·R[sub]3[/sub]=｛0.5312,0,0,0,0｝ We can obtain the membership matrix R′, and then the comprehensive evaluation of the seat is B′=W·R′=（0.5495,0.0959,0.0211,0.2676,0） According to the "maximum membership principle", we can conclude that the evaluation of the seat surface, backrest, and seat height is "good", "excellent", and "excellent", and the evaluation of this seat is "excellent". 3 Conclusion Based on the analysis of driving process and human-computer interaction, evaluation factors and evaluation indicators were selected, and an evaluation index system suitable for armored vehicle driver seats was established. Finally, the fuzzy comprehensive evaluation method was used to evaluate the driver seats. The evaluation method based on human-computer interaction process analysis and human-computer interaction operation analysis proposed in this paper also has a good reference role for the evaluation of other human-computer interfaces. References: [1] Jiang Tao. Research on Comprehensive Evaluation of Human-Machine Interface of DCS System in Thermal Power Plant [D]. Harbin: Harbin Engineering University, 2006. [2] Sun Linyan. Human Factors Engineering [M]. Beijing: China Science and Technology Press, 2001. [3] Long Shengzhao, Huang Duansheng, Chen Daomu, et al. Theory and Application of Human-Machine-Environment Systems Engineering [M]. Beijing: Science Press, 2004. [4] Zhang Xizheng. Virtual Enterprise Partner Selection Based on Trust Review [J]. China Mechanical Engineering, 2005, 16(2): 161-164. [5] Ye Yicheng, Ke Lihua, Huang Deyu. System Comprehensive Evaluation Technology and Its Application [M]. Beijing: Metallurgical Industry Press, 2006. For details, please click: Evaluation of Human-Machine Interface of Armored Vehicle Cockpit