VIRTUAL PROTOTYPING of ROBOTS DYNAMICS E. Tarabanov INTRODUCTION Modern computer technology achievements allow: пЃ¬ пЃ¬ пЃ¬ to simulate systemsвЂ™ separate unit functioning; to present systemsвЂ™ 3D shape on a monitor screen; to investigate its behavior under conditions that are close to real ones. With simulation, one can пЃ¬ пЃ¬ пЃ¬ пЃ¬ пЃ¬ gain better understanding how a system works identify problems prior to their implementation test potential effects of changes identify areas for resource deployment design efficient and cost-effective systems VIRTUAL PACKAGES пЃ¬ пЃ¬ пЃ¬ пЃ¬ пЃ¬ пЃ¬ пЃ¬ пЃ¬ MSC.Adams Universal Mechanism (UM) MatLab DyMoLa Webots Simbad ThreeDimWorks RoboWorks MATLAB MatLab is an interactive software system for numerical computations and graphics. MatLab is designed especially for matrix computations: solution solving of linear equation systems, computing eigenvalues, eigenvectors and matrices, and so on. MATLAB scopes: пЃ¬ пЃ¬ пЃ¬ пЃ¬ пЃ¬ пЃ¬ mathematics and calculations; algorithm development; computing experiment, imitating modeling, prototyping; data analysis and result researches; scientific and engineering visualizations; application development, including the graphic interface of the user MATLAB/Simulink Simulink is an interactive tool for dynamics system modeling and analysis. User creates device model by means of standard blocks and carries out calculations. There are additional block libraries for different scopes as SimPowerSystems вЂ“ electrotechnical devices modeling, SimMechanics вЂ“ mechanical devices modeling, Digital Signal Processing Blockset вЂ“ a digital device development, etc. MatLab: Bus suspension Designing an automatic suspension system for a bus MatLab/Simulink: r3 Dynamics model structure DyMoLa DyMoLa is a complete tool for modeling and simulation of integrated and complex systems for use within automotive, aerospace, robotics processes and other applications. DyMoLa is based on Modelica, which is an object-oriented language for physical modeling. DyMoLa: Examples Vehicle model animation including transmission and engine Robot model animation built with the MultiBody library and the Modelica Standard library DyMoLa: Air-conditioning system UNIVERSAL MECHANISM UM was designed to automate the analysis of mechanical objects connected by means of kinematical and force elements. UM widely applies computer graphics methods to display systemвЂ™s 3D motion in equation process and to analyze the obtained results. UM: Subsystem technique Train as a subsystem set UM: Solving direct and inverse kinematic problems Motion animation UM: Examples Spider robot Robot вЂњPumaвЂќ Bipedal walking robot MSC.ADAMS It allows to test virtual prototypes and optimize designs for performance, safety and comfort, without having to build and test numerous physical prototypes. Software package structure ADAMS benefits: пЃ¬ пЃ¬ пЃ¬ пЃ¬ пЃ¬ work in a secure virtual environment, without the fear of losing critical data; reduce risk by getting better design information at every stage of the development process; analyze design changes much faster; improve product quality by exploring numerous design variations in order to optimize full-system performance; vary the kinds of analyses being performed without having to modify physical instrumentation, test fixtures, and test procedures. ADAMS extensions: пЃ¬ пЃ¬ пЃ¬ ADAMS/Control - to analyze control systems; ADAMS/Flex - to examine the impact of flexible parts; ADAMS/Linear - to calculate natural frequencies and mode shapes of large systems. ADAMS: Examples Automobile engine Chassis work Orbit operations Racing car movement ADAMS: Multilink robot General view of a segment ZMEELOC ZMEELOC: Virtual model ZMEELOC: Modeling results Direct course ZMEELOC: Modeling results Lateral course SEPTOPOD: SolidWorks model SEPTOPOD: ADAMS model Rotary paw view SEPTOPOD: Modeling results SEPTOPOD: Modeling results The top view Humanoid robot: Modeling results Movement model Humanoid robot: Modeling results Movement model CONCLUSION пЃ¬ пЃ¬ пЃ¬ Modeling is the first step to design, identify and control robots and itвЂ™s a powerful technique to improve quality and productivity. Most efficient algorithms proposed for these applications are based on good study of robot model parameters. These operations need to be carefully planned. Modern virtual prototyping tools have possibility to project dynamic and mechanic systems and allow to reach greater success in designing. The successful coordination and interaction of robots and software will require new system designs, communication protocols and interfaces.