Introduction and effects of common simulation software for robot systems

1 Mainly introduces the following seven simulation platforms (focusing on mobile robot simulation rather than industrial robot simulation such as robot arm):

1.1 USARSim-Unified System for AutomaTIon and Robot SimulaTIon
USARSim is a high-fidelity multi-robot environment simulation platform based on virtual arena engine design. Mainly for ground robots, which can be used for research and teaching. In addition, USARSim is the basic platform for the RoboCup rescue virtual robot competition and virtual manufacturing automation competition. Use Open Dynamics Engine (ODE) to support 3D rendering and physics simulation, high configurability and scalability, compatibility with Player, layered control system, open interface structure simulation function and tool framework module. Robot control can be implemented using virtual script programming or network connections using the UDP protocol. It is widely used in robot simulation, training of military recruits, fire fighting and search and rescue missions. Robots and environments can be generated by third-party software. The software follows the free GPL and multi-platform support can be installed and run on Linux, Windows and MacOS operating systems.
"http://sourceforge.net/projects/usarsim/"

机器人系统常用仿真软件介绍和效果

1.2 Simbad
Simbad is a multi-robot simulation platform for scientific and educational purposes based on Java3D. The main focus is on simple basic questions of artificial intelligence, machine learning and more general-purpose artificial intelligence algorithms in multi-robot systems that researchers and programmers are keen on. It features programmable robot controllers, customizable environments and custom configuration sensor modules, 3D virtual sensing technology, single or multi-robot simulation, and toolboxes such as neural networks and evolutionary algorithms. Software development is easy, open source, based on the GNU protocol, does not support physical computing, and can run on any Java client system that supports the Java3D library.
"http://simbad.sourceforge.net/"

1.3 Webots
Webots is a mobile robot development platform with modeling, programming and simulation, mainly used for ground robot simulation. Users can design a variety of complex heterogeneous robots in a shared environment. The environment can be customized. The properties of all objects in the environment, including shape, color, text, quality, function, etc., can also be freely configured by the user. Using ODE to detect object collisions and simulate the dynamics of rigid structures, you can accurately simulate physical properties such as object velocity, inertia, and friction. Each robot can be equipped with a large number of optional simulation sensors and drivers. The controller of the robot can be programmed through an internal integrated development environment or a third-party development environment. The controller program can be written in C, C++, etc. Can be tested in the real world. Support a large number of robot models such as khepera, pioneer2, aibo, etc., can also import their own defined robots. The simulation software is used by more than 750 universities and research centers around the world, but it costs money to support major operating systems including Linux, Windows and MacOS.
"http://TIcs.com/"

1.4 MRDS-Microsoft RoboTIcs Developer Studio
MRDS is a Windows-based, networked, service-based framework-based robot control simulation platform developed by Microsoft. It uses the PhysX physics engine and is one of the most fidelity simulation engines currently available for academic, enthusiast and commercial development. Support a large number of robot hardware and software. MRDS is based on real-time concurrent coordinated CCR (Concurrency and Coordination Runtime) and distributed software service DSS (Decentralized Software Services), for asynchronous parallel task management and allows multiple services to coordinate management to obtain complex behavior, providing Visual Programming Language (VPL) And Visual Simulation Environment (VSE) [28, 29]. Support for mainstream commercial robots, the main programming language is C#, non-commercial applications are free, but only support development under the Windows operating system.
"http://msdn.microsoft.com/zh-tw/library/bb648760.aspx"
"http://msdn.microsoft.com/library/bb648760"
"http://?id=29081"

1.5 PSG-Player/Stage/Gazebo
PSG is a free platform developed by the University of Southern California (USC) for robotics and sensor systems research, including the network service part of the Player and the robot platform simulation part Stage and Gazebo. Player defines the communication interface between the robot and the sensor and the Stage and Gazebo. The Stage is a 2D environment that provides basic collision detection and distance sensor models but does not support physical simulation. Gazebo is a 3D environment using the ODE physics engine. PSG provides virtual robotics such as sonar, laser scanning range finder, collision detection and actuators to support multi-robot simulation. It is the most popular open source robot simulation software in research and teaching. The developed program can be applied to the control of solid robots with simple modification and even no modification, so the research cost can be greatly reduced and the research cycle can be reduced. A large number of major intelligent robot journals and articles published in conferences use PSG as a real and simulated simulation experiment platform. Free software, based on the GNU protocol, is developed on Unix systems and only supports Linux and Mac OS.
"http://playerstage.sourceforge.net/"
"http://"

1.6 MissionLab
MissionLab is a powerful platform developed by Georgia Tech to develop and test individual or a set of robot behaviors. The code generated by Missionlab can directly control mainstream commercial robots, including ARTV-Jr, iRobot, AmigoBot, Pioneer AT and MRV-2. The main advantage of Missionlab is that it supports simulation and real robot simultaneous experiments. MissionLab is a distributed architecture with six core components: mlab, CfgEdit, cdl, cn, HServer, and CBRServer. Use CMDL and ODL as the development configuration language. It was originally developed for DARPA to study the flexible response control of multi-agent robotic systems in hostile environments. It is now open source and only supports the Linux operating system.
"http://MissionLab/"

1.7 MORSE-Modular OpenRobots Simulation Engine
MORSE is a general-purpose multi-robot simulation platform. It is mainly used to control the degree of freedom of actual simulation. It can freely design component models that meet its own needs, use Blender real-time game engine for original rendering, design suitable architecture, and support general-purpose Network Interface. It offers a wide range of configurable sensor and actuator modules, high scalability, interactive simulation of humans and robots, programming in Python, rich documentation and easy installation but no precise dynamics simulation, clock synchronization Poor performance, unsynchronized when multi-robot simulation. Currently used by five schools and research institutions, open source software is limited to Linux and MacOSX operating systems.
"https://"

1.8 Other commonly used robot simulation software
ROS----"http://"
V-Rep----"http://"
MRPT----"http://"

Aerial robot simulation software
Orbiter----"http://orbit.medphys.ucl.ac.uk/"
Flighntgear----"http://"

2 Evaluation indicators

In order to objectively evaluate these simulation software from the perspective of a robotic researcher, a uniform standard needs to be established. Based on the previous work of Alexander, Craighead, Michael and other scholars in this paper, the usual criteria for evaluating robot simulation platforms are determined: fidelity, scalability, ease of development, and cost. These four criteria can be used to judge any virtual robot simulation. Software, but for multi-robot systems, due to its distributed characteristics, the requirements for networking are getting higher and higher, so network functions are also an indispensable important indicator. The simulation software is divided into three levels of high, medium and low according to these standards.

2.1 Fidelity
2.1.1 Physical fidelity Physical fidelity refers to the extent to which the physical environment's appearance, sound, and feel approach the real operating environment. When a simulation software cannot simulate the feeling of the operating environment, it can simulate the visual and auditory parts of the environment. Alexander proves that research results can be enhanced by higher-fidelity simulation software. Taking robot operation as an example, the operator is generally far away from the operating environment, and a high-fidelity visual and auditory simulation software is comparable to the real robot. The level of physical fidelity is determined by the visual rendering and audio performance of each simulation software. A high physics fidelity simulation software can render high resolution textures, materials, lights, reflections, and bump maps. The model uses a large number of polygons instead of using texture maps as the main part of geometric modeling. The motion characteristics of the robot must be displayed correctly. A high physics fidelity simulation software also includes a high degree of integration of robots and environmental sensations. A simulation software with physical fidelity can pass through the 3D rendering environment, and there is no requirement for object details. A low-physical simulation software can only be used in a 2D rendering environment with no sound.

2.1.2 Functional fidelity Functional fidelity refers to the degree to which the robot's behavior in the simulation approximates the real robot's mission operating environment and equipment response. The primary purpose of functional fidelity is to mean that the simulation software user expects behavior similar to that of real equipment. The functional fidelity level is determined by comparing the physical property simulation performance. High-fidelity fidelity is positioned to simulate most of the mechanical properties of the robot and the drive, including gravity, traction, motor or acceleration generated by collisions with other object robots, and can accurately display and record status information for analysis. The mid-function fidelity includes the overall dynamics of the robot, but it is not possible to simulate a single joint. Low-function fidelity means that the software cannot simulate the mechanical behavior of the robot and can only simulate kinematics such as speed and position.

2.2 Scalability Scalability refers to the ability of simulation software to adapt to multiple applications. The main performance of the simulation experiment algorithm development can not be limited by the simulation software function. Including whether it is easier to increase and decrease the number of robots, custom sensors and drive modules, whether to provide a modular standard interface to support third-party software development, the emulation program application and the portability of real robots.

2.3 Development Simplicity Development simplicity refers to the difficulty of developing a suitable robot using simulation software. Including the complexity of environment and robot configuration, the complexity of software configuration new equipment, the type of programming language supported, the richness of software development documentation, and the versatility of communication interface configuration.

2.4 Cost costs include cost and time spent. Mainly the time that must be consumed during the initial installation and development, the price of the software, etc. If the software is open source free and easy to install and use, it is low cost, the software is free but the installation is complicated or the software is paid, but the installation and use are simple. The software purchase is expensive and the installation needs to be positioned to be high.

2.5 Network function Network function is an important part of distributed multi-robot system, including whether to support remote operation of real robots, whether there are complete simulation robot network interfaces and multi-terminal joint debugging simulation of large complex scenes.

3 simulation effects can refer to video recording
"http://i.youku.com/u/UMTg1NDE4MDM2"

Cable Trunking Accessories

Cable Trunking Fittings,Cable Trunking Accessories,Electric Cable Trunking,Pvc Trunking Accessories

FOSHAN SHUNDE LANGLI HARDWARE ELECTRICAL CO.LTD , https://www.langliplastic.com