装备参数测试模拟训练软件开发平台设计与实现

发布时间：2018-05-06 07:14

本文选题：参数测试 + 开发平台　；参考：《湖北工业大学》2017年硕士论文

【摘要】：随着电子装备固态化、集成化、模块化水平的提高,在大型电子装备上直接开展性能参数测试实习已越来越困难,采用虚拟现实技术来开展模拟训练成为发展趋势,然而,电子装备的种类很多、需要测试的参数也很多,如果对每一种设备的参数测试开发一个模拟训练软件,将使软件的开发工作量非常大。针对以上问题,论文从模拟训练软件开发平台的设计需求分析出发,研究了该软件平台总体设计,操作模型建立及虚拟构件装配,以及交互式模拟训练实现的相关理论与方法,论文所做的主要工作与成果如下:1)提出了基于虚拟构件开发平台与仪器控件库、测试流程库相结合的装备参数测试模拟训练软件开发平台设计方案。论文从模拟训练软件开发平台的需求分析出发,讨论了开发平台研制过程中涉及的关键技术、可能利用现有的软件平台及方法,提出了基于PE平台,融合仪器仪表构件库、测试电缆附件库和测试步骤二次开发描述语言的开发平台总体设计方案,该方案具有软件结构简单,可扩展性强,工作量适当,模拟训练软件开发简单快捷的特点。2)三维操作模型的建立和模型装配直接影响虚拟场景的效果及软件功能的实现,是模拟训练软件开发的关键问题。论文从建模方法、运动方式、模型渲染出发,通过研究自动扫描、手工建模、图像建模、碰撞检测、LOD技术、参数计算模型建立等问题,系统研究了装备参数测试模型的建立方法,形成了LOD在场景中渲染规则,提出了手工建模和图像建模融合的策略。最后通过实际操作对提出的方法进行了验证。3)针对模拟训练需要对训练人员进行考核评估的问题,论文研究了测试参数生成、组卷系统建模、学习曲线的建立方法。首先通过伪随机算法的研究,构建出伪随机算法的装备参数测试生成方法;然后研究了考核难度计算模型和组卷目标函数;最后,研究了学员学习曲线的构建方法,从客观反映学习趋势的角度出发,分别研究了相同复杂度的高斯拟合算法和最小二乘拟合算法,探讨了一般性曲线和震荡型曲线的拟合。然后通过算例对比,对实验结果进行了总结,结果表明提出的最小二乘算法更能接近真实结果,为之后的学习进程规划了合理的计划。
[Abstract]:With the improvement of solid state, integration and modularization of electronic equipment, it is becoming more and more difficult to carry out performance parameter testing practice directly in large electronic equipment. It is becoming a trend to use virtual reality technology to carry out simulation training. There are many kinds of electronic equipment and a lot of parameters need to be tested. If a simulation training software is developed for each kind of equipment parameter testing, the development of software will be very heavy. In order to solve the above problems, this paper analyzes the design requirements of the simulation training software development platform, and studies the related theories and methods of the overall design of the software platform, the establishment of the operation model and the assembly of virtual components, as well as the realization of interactive simulation training. The main work and results of this paper are as follows: (1) A design scheme of equipment parameter test simulation training software development platform based on virtual component development platform, instrument control library and test flow library is proposed. Based on the requirement analysis of the simulation training software development platform, this paper discusses the key technologies involved in the development process of the platform. Using the existing software platform and methods, the paper puts forward the integration of instrument and instrument component library based on PE platform. The overall design scheme of the development platform of the test cable attachment library and the second development description language of the test step has the advantages of simple software structure, strong expansibility and proper workload. The establishment and assembly of 3D operation model directly affect the effect of virtual scene and the realization of software function, which is the key problem in the development of simulation training software. Starting from modeling method, motion mode and model rendering, this paper studies automatic scanning, manual modeling, image modeling, collision detection and LOD technology, parameter calculation model establishment and so on. The modeling method of equipment parameter test model is studied systematically, and the rendering rules of LOD in the scene are formed, and the strategy of combining manual modeling and image modeling is put forward. Finally, the proposed method is verified by practical operation. 3) in view of the need of training simulation to evaluate the training personnel, the paper studies the test parameter generation, test paper system modeling, learning curve building method. First of all, through the research of pseudorandom algorithm, the method of equipment parameter test generation of pseudorandom algorithm is constructed. Then, the evaluation difficulty calculation model and test paper objective function are studied. Finally, the method of constructing learning curve is studied. From the point of view of reflecting the learning trend objectively, Gao Si fitting algorithm and least square fitting algorithm with the same complexity are studied, and the fitting of general curve and concussion curve is discussed. Then the experimental results are summarized by the comparison of numerical examples. The results show that the proposed least squares algorithm is closer to the real results and a reasonable plan for the later learning process is proposed.
【学位授予单位】：湖北工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.52

【参考文献】