Design And Key Techniques Of Virtual Reality Simulation Platform For Photoelectric Theodolite

Photoelectric theodolites are widely used in the field of photoelectric trackingmeasurement. With the continuous improvement of theodolite performance, theconventional methods for operator training and theodolite testing, which based on theoutdoor experiments or indoor simulation, have been unable to satisfy the requirement.In view of this situation, this thesis is devoted to the research for the combination ofvirtual reality technology and system simulation technology. And a virtual realitysimulation platform has been designed and set up. The platform, which can be used forboth operator training and theodolite testing, has three function modes: manual tracking,automatic tracking and semi-physical simulation. The research in this thesis, which hasimproved the existing methods for operator training and theodolite testing, is quitepractical in engineering application.The basic working process of the photoelectric theodolite is briefly introduced andthe limitations of the conventional indoor methods for simulation testing are analyzed.According to the desired objectives and function requirements, a design of the platform,which includes system structure, composition of the hardware and software, workingmodes and key technical problems, has been proposed. The performance of theapplication shows that the platform is economic, practical, efficient and reliable.In order to meet the requirements of the trajectory of the virtual target for trainingand testing, two methods for target trajectory simulating are studied. One of themethods uses the actual outdoor experiments result according to the single station orintersection measuring model; the other takes advantage of the target movementcharacteristics according to the relationship of the target trajectory, velocity and attitude.The former method is suitable for targeted training and testing for outdoor experiments,and the latter one is convenient in simulating the target trajectory whose angularvelocity and acceleration meets the requirement of the theodolite testing.Rendering the original virtual scene in real time, which is the foundation of thesimulation platform, is carried out. The3D modeling technology based on Creator andvisual simulation technology based on Vega are studied. The theodolite imagingcharacteristic of the virtual scene is simulated based on the principle of OpenGLperspective projection and the imaging characteristic of the theodolite camera. The simulation of the cloud background of the virtual scene, including Vega cloud model,particle system, Perlin noise and Simplex noise, is researched. The scene simulationprogram is developed with Vega, OpenGL and VC. In the instance given by this thesis,the frame rate keeps around270FPS, which completely satisfies the real-timerequirement.The research on adding desired image effects to original virtual scene in real timeis made for the sake of making the virtual scene closer to the real scene. Based onVega’s program structure and channel callback function, a method for the real-timeprocessing of the scene image is proposed. It makes use of modern GPU’s powerfulgraphics capabilities and programmability, and uses GLSL shader to replace OpenGL’sfixed texture mapping function to perform image processing on the scene texture. Theimage effects, including luminance, noise and blur, are added to original scene. Thenoise effect is realized based on Perlin and Simplex noise; the blur effect is realized byconvolution filtering. In order to improve the efficiency of convolution filtering, anoptimization method is proposed based on GPU’s linear texture filtering. For theconvolution filter in this thesis, the efficiency is improved by more than60%. For thefinal scene to which the image effects are added, the frame rate keeps around240FPS.Finally, the realization of the three working modes of the simulation platform isstudied. The manual and automatic tracking modes are realized by rendering the virtualscene in real time and generating tracking state solver module with RTW. These twoworking modes can be conveniently and flexibly used for operator training. Thesemi-physical simulation mode is set up based on the combination of the simulationplatform and real theodolite, and the theodolite tracking control algorithm is fastrealized with RTW/xPC. This working mode can be applied to the fast testing of thetheodolite tracking control system.