Abstract: To achieve the design and optimization of the human-machine interface (HMI) of a vehicle cab in a virtual environment, a vehicle HMI layout reasoning method is proposed. This method includes research on HMI knowledge representation and layout reasoning architecture, and is evaluated and verified using the HMI layout of an armored vehicle cab as an example. The results show that this method can be effectively applied to the design and evaluation of the HMI layout of a vehicle cab in a virtual environment. Keywords: vehicle; human-machine interface; knowledge representation; spatial layout; CATIA Introduction Currently, the artificial intelligence technologies commonly used in the design, analysis, and evaluation of vehicle cab ergonomics mainly include expert systems, neural networks, and genetic algorithms. Expert systems are one of the most important applications of artificial intelligence, aiming to enable computers to reason and judge based on the knowledge and experience provided by one or more experts in a certain field, simulating the decision-making process of human experts, in order to solve complex problems that require human experts to handle. In domestic and international research, the human-machine consultation expert system based on human-machine dialogue boxes proposed by Gilad et al. in the 1990s is an early attempt in this field. Literature [2-3] studied the construction method of knowledge-based workspace design systems. Feyen R et al. [sup][4][/sup] combined computer-aided design software with ergonomics analysis software to develop a software system for evaluating the impact of workspace design on workers' biomechanical characteristics. Ren Jindong et al. [sup][5][/sup] developed a knowledge-based automotive driver's seat layout system on the CATIA platform for the interior layout of car bodies. These research results all have high academic value and each has its own characteristics. To date, although the application of artificial intelligence technology in computer-aided design and ergonomics analysis has made some progress, it has not yet truly solved the communication problem between designers and ergonomics engineers. The vehicle human-machine interface layout optimization reasoning method proposed in this paper, under the premise of analyzing the task and characteristics of vehicle human-machine interface layout optimization, starts from the knowledge expression and reasoning of vehicle cab ergonomic layout, realizing the process from design, simulation, evaluation to feedback modification and optimization; and realizing the tight integration of the human-machine interface layout reasoning system with the 3D CAD platform. 1. Knowledge Representation of Human-Computer Interface Layout Knowledge representation is the formalization of factual knowledge and expert experience, represented using symbols and methods that computers can accept and process. Currently, commonly used knowledge representation methods include production rule representation, frame representation, semantic network representation, and object-oriented representation. The human-computer interface layout design process is quite complex, involving many factors and conditions; it is a process of multi-factor coordination and continuous iterative feedback. Design knowledge is also diverse, involving not only the spatial arrangement and dimensional parameters of related components but also factors such as ride comfort, operational comfort, and visibility. The design process requires not only extensive professional domain knowledge and expert experience and skills but also a large amount of scientific calculation and analysis. Designers must consider the structural and functional characteristics of the design object, as well as its behavioral and operational characteristics. Therefore, by combining the advantages of various single knowledge representation methods, object-oriented technology is used to combine frames, rules, and procedures to form a hybrid knowledge representation method. In the object-oriented method, classes, subclasses, and concrete objects form a hierarchical structure, and subclasses can inherit data and operations from their parent classes. Class structure is the most basic unit of a knowledge base. It can express not only the attribute data of a design object but also the relationships between design objects. Each object's knowledge representation should include four parts: attributes, constraints, methods, and rules. Their descriptive form is as follows: `Class<object name>:<base class> { Attribute<attribute description> {attribute item, attribute name, inheritance relationship, attribute type, attribute set} Methods<method description> {method item, method name, inheritance relationship, method type, method set} Restraint<constraint> {constraint item, constraint name, inheritance relationship, constraint type, constraint set} Rules<rule description> {rule item, rule name, inheritance relationship, rule type, rule set}` For example, the structure of the knowledge representation of a chair is shown in Figure 1. The human-computer interface layout knowledge base system consists of an instance library, a rule library, a constraint library, and a knowledge base. The instance library stores successful instances and new instances generated by the inference engine; the rule library stores rules for knowledge reasoning, such as instance search and similarity algorithms; the constraint library stores various design constraints, such as geometric constraints, performance parameters, and other regulations and standards related to human-computer interface design, including human body dimensions, visual characteristics, and limb reach range; the knowledge base stores a large amount of expert experience, experimental data, design principles, and formulas, and is managed by the knowledge base management system. The knowledge data in the knowledge base can be adopted by the inference engine and maintained through the system's knowledge acquisition. 2. Human-Computer Interface Layout Reasoning Architecture 2.1 Overall Structure The overall structure of the vehicle human-computer interface layout reasoning system is shown in Figure 2. To simplify the design state space, the system adopts a layered structure. The top layer is the scheme design layer, followed by the master parameter design layer, and the bottom layer is the detailed design layer. The task of scheme design is to derive the values ​​of each scheme element. After all scheme elements are determined, master parameter design is performed; based on the results of scheme design and master parameter design, detailed design is performed, generating graphics and technical documents. This system can, according to the user's needs, start from the design data of the upper two layers, reason, match, generate detailed design parameters, and finally output the results. 2.2 Reasoning and Evaluation System The vehicle human-machine interface layout optimization system is built in layers, and each layer has its own characteristics. Therefore, there are different types of reasoning mechanisms, including not only the main reasoning engine, but also the sub-reasoning engines embedded in each object. (1) Reasoning and Evaluation System of Scheme Design In the scheme design stage, since experience design is mainly used, the reasoning method adopts a combination of fuzzy logic reasoning and evaluation. ① Fuzzy matrix method of reasoning. For fuzzy rules: IF x is A THEN y is B, A and B are fuzzy subsets on the domains U and V, respectively, indicating that there is a fuzzy causal relationship between A and B; let R be. If the fuzzy fact A is known, B can be obtained by combining A and R, that is, B = R * A. ② Evaluation strategy in scheme design. The evaluation process is similar to the reasoning process. On the one hand, the evaluation assists the reasoning, and on the other hand, it filters the generated multiple schemes. Evaluating the scheme elements in the scheme design stage can make up for the lack of knowledge or uncertainty in the reasoning process. There are two strategies for evaluation: always evaluate the reasoning results; evaluate only when the membership of the results generated by reasoning is close. (2) Reasoning system of the main parameter design stage The parameter design stage includes the main parameter design stage and the graphic parameter design stage (i.e., detailed design stage). The main parameter design stage refers to the main parameter design stage. It not only uses the results of the scheme design as a premise for reasoning, but also uses the intermediate parameters that have been generated for reasoning. Moreover, it is constantly used in various numerical calculation methods. It is a generalized reasoning system that includes logical reasoning, numerical calculation, optimization design and comprehensive evaluation process. 3 Human-computer interface At present, some three-dimensional CAD systems have secondary development interfaces, which are conducive to the close integration of CAD systems and human-computer engineering. Among them, CATIA/CAA is representative. CATIA is a high-end CAD/CAM software system developed by IBM/DS based on the Windows core. It has a unified user interface, data management and compatible database and application interface. With its powerful design functions, it is widely used in aviation, aerospace, automobile, shipbuilding and electronic equipment industries. CAA is built under Microsoft Visual C++ 6.0, a powerful tool for DS system product extension and customer secondary development. It calls CATIA's core program via API functions, allowing users to integrate their knowledge into specific CATIA application modules, thus achieving tight integration between client programs and the original system. In China, the application of CAA to CATIA for component application architecture secondary development is still in its early stages, with relatively little research. Due to the powerful functions of CATIA software and CAA's secondary development capabilities, exploring and implementing CATIA-based secondary development technology itself has significant application value. 4. Case Study Based on the above analysis, CATIA/CAA secondary development was carried out on the Windows operating platform to implement a vehicle human-machine interface layout optimization inference system. Within the system, considering the characteristics of Chinese users, a reasoning and evaluation study was conducted on the layout of the human-machine interface in the driver's cab of a pair of armored vehicles. 4.1 Establishing a Digital Human Body Model Based on relevant design standards and measurement data of relevant human body data, the driver's limb range of motion, functional dimensions in a sitting posture, activity space dimensions, and joint angle comfort range were determined. A driver sitting posture human body dimension database was designed, and an evaluation digital human body model was established. CATIA's built-in human body model module includes 103 anthropometric measurements, over 100 independent and unconstrained segmental connections, 148 degrees of freedom, and various pose profiles, allowing for constraint and adjustment of joint movements. The software contains anthropometric data from the United States, Canada, South Korea, France, and Japan, but not from China. Comparison reveals inconsistencies between existing anthropometric measurements for Chinese armored vehicle drivers and those in CATIA (only 31 items are identical). Some anthropometric data required for model building in the software are not included in my country's anthropometric measurements, and some domestic measurements are not reflected in the software's model. The reasons for this are twofold: firstly, my country has not conducted measurements for certain common parameters; secondly, the measurement standards for some parameters differ from internationally accepted standards. Therefore, existing anthropometric dimensions for armored vehicle drivers do not meet the requirements for model building. Determining the missing anthropometric parameters is the primary issue to be addressed in establishing an anthropometric model for Chinese armored vehicle drivers. Currently, the anthropometric dimensions that must be addressed are those included in anthropometric measurements from other countries but not in Chinese anthropometric measurements. The solution is to refer to anthropometric data from South Korea and Japan and select parameters based on experience. Measurement work for the required projects should continue in the future. Based on existing anthropometric data of armored vehicle drivers, this paper compiles a human anthropometric data file for Chinese armored vehicle drivers following the ".sws" file format, embeds it into CATIA, expands the CATIA human anthropometric model parameter library, and provides interfaces for modifying parameters of various parts of the human body. 4.2 Constructing the Task Model and Operation Layout According to the corresponding vehicle model, the human-machine interface layout model is loaded, defining various possible seat states, human operating postures, and the layout of control handles and foot pedals. 4.3 Defining Human-Machine Analysis Factors This includes various evaluation indicators for comfort analysis. Comfort evaluation generally includes riding comfort, operating comfort, and dynamic comfort. Here, the focus is mainly on static riding comfort and operating comfort, that is, the comfort of the driving posture. Based on the relationship between the geometric parameters of the seat and driving device, anthropometric parameters, human sitting joint angles, etc., and user riding comfort, the comfort of the human operating posture in the seat is judged. For the human-machine interface of the armored vehicle driver's cockpit, the installation position of each control element determines the angle of the driver's limb joints during operation, directly affecting the driver's driving behavior. Extensive research has been conducted on the maximum and comfortable range of motion of human joints, resulting in a wealth of research data. Furthermore, the suitable range of motion of human joints does not change with variations in the evaluation environment; evaluating joint angles can accurately reflect the operator's riding and maneuvering comfort. However, considering the need to increase the tank's protection during combat, it is necessary to effectively reduce the vehicle's height, and data such as the seat height during closed-window driving should be minimized within permissible limits. 4.4 Simulation Execution and Design Evaluation Visualization, dynamic analysis, and comprehensive research were conducted on the real-time task, and design improvements were proposed. Figure 3 shows the initial layout model of the human-machine interface (HMI) of an armored vehicle driver's cab, the virtual human's driving posture, and the added HMI layout reasoning and evaluation module. In the original HMI layout, analysis of static riding comfort revealed that the original layout did not conform to ergonomic standards in many aspects. Figure 4 shows the results of the human driving posture comfort evaluation analysis before improvement. It can be seen that the comfort of the armored vehicle driver's legs is low, mainly due to the armored vehicle driver's cab seat height not conforming to ergonomic standards. [align=center] [/align] Using information from system feedback, the layout of the armored vehicle's human-machine interface (HMI) was improved, mainly including optimization of the seat and foot pedal arrangement, followed by a re-evaluation through simulation. In the improved HMI layout, analysis of human seating comfort revealed that the improved layout conforms to ergonomic standards in many aspects compared to the previous layout. The improved layout model, design parameters, and evaluation results are shown in Figure 5. It can be seen that the virtual human's leg comfort has improved. However, due to limitations in driving space, the armored vehicle HMI layout still has some problems. 5 Conclusion A reasoning method for optimizing the vehicle HMI layout was proposed, the key technologies were analyzed, a reasoning and evaluation system was established, and HMI modeling and simulation were performed. During the simulation, ergonomic analysis and evaluation, such as riding comfort, were conducted. Virtual verification results show that this method can be effectively applied to the optimization reasoning and evaluation of the vehicle cab HMI layout in a virtual environment. However, the system itself still has some shortcomings, such as insufficient human model data, imperfect analysis and evaluation algorithms, and application systems limited to certain specific industries. Further improvements are needed in future research. References: 1. Issachar Gilad, Reuven Karmi. Architecture of an expert system for ergonormics analysis and design[J]. International Journal of Industrial Ergonomics, 1999, 23(3): 205～221. 2. Yadav R, Tewari V K. Tractor operator workplace design-a review[J]. Journal of Terramechanics, 1998, 35 (1): 41-53. 3. Shikdar Ashraf, Al-Araimi Saeed, Omurtag Bill. Development of a software package for ergonomic assessment ofmanufacturing Industry [J]. Computers & Industrial Engineering, 2002, 43 (3): 485-493 4. Feyen R, Liu YL, Chaffin D, et. al.Computer-aided ergonomics:a case study of incorporating ergonomic analysis into workplace design[J].Applied Ergonomics, 2000, 31(3): 291-300. 5. Ren Jindong, Ge Anlin, Huang Jinling. Knowledge-based automotive driver seat arrangement system [J]. Automotive Engineering, 2003, 25(3): 275-278. 6. Fang Yonglong, Zou Naiwei, Yang Zhihua, et al. Human-machine interface design and knowledge base establishment of automotive braking expert system [J]. Transactions of the Chinese Society for Agricultural Machinery, 2006, 37(6): 23-26. 7. Zhou Yiming, Mao Enrong. Vehicle Ergonomics [M]. Beijing: Beijing Institute of Technology Press, 1999.