Software Platform Design And Research Of Mobile Robot In Multicore Environment

With the evolution of science and technology, intelligent devices had been graduallyintegrated into our daily lives. In order to solve the bottleneck of power and performanceof single-core processor, multi-core processors emerged. Embedded devices such assmart phones, smart glasses, and small intelligent service robots had begun widespreadadoption of high-performance multi-core processors. And the wheeled mobile robot wasthe most concerned research hotspot in robotics field, it was also got widely applied inindustrial, scientific and lives.This thesis designed and implemented a wheeled mobile robot platform basedhigh-performance multi-core SoC, and the platform was committed to solving the indoorenvironment map modeling, autonomous navigation and other issues. The platformadopted Xilinx’s ZYNQ series SoC, which contained a hardcore (Processing System, PS)called Cortex A9MPCore and programmable logic (PL) module. In order to makeadvantage of multi-core performance and to ensure real-time robot control, the platformintroduced embedded virtualization to make real-time operating system and generaloperating system coexist, and also introduced robot operating system called ROS (RobotOperation System). The mobile robot platform architecture was divided into threedifferent functional layers, which was data accelerated processing layer, real-time controland human-computer interaction layer and remote high-performance processing layer.Data accelerated processing layer was responsible for data acquisition and pre-processingof some sensors and acceleration of complex parallel computing. Real-time control andhuman-computer interaction layer was responsible for comprehensively handling sensordata, distributing control instructions and data exchange with remote high-performanceserver using ROS service. Remote high-performance processing layer was mainlyresponsible for handling the summary information send by the robot platform through wireless network and assisting with map modeling and autonomous navigation and etc.In this thesis, the mobile robot platform was built according to design goal, anddesigned and realized the software platform for it. Then we set up a testing environmentvalidated map modeling and autonomous navigation of the mobile robot in unknownenvironments.