LVRT Based Motion ControlSadia MalikRam RajagopalMay 2, 2000Motion Control TechnologiesRT TrajectorySpliningPID ControlShared Memory InterfaceHost (PC)Motion BoardMechanical DeviceAnalyze/StoreDesiredTrajectoryProblem Statement◆Hard Coded Control Loops◆Simulation environment very different fromactual environment◆Limited trajectories◆Custom designed solution often neededPossible Solutions◆Provide common simulation andimplementation environment– Adjust input parameters dynamically– Update control algorithm with system changes◆Auto identify and present the most suitablemodel for the system– Allow user defined algorithms– Select optimal control designLabVIEW Real Time◆Graphical Programming Environment– Virtual Instruments as components– Built-in debug environment◆Real time execution environment◆Code portable to any LVRT serieshardwareOur Solution◆Used LVRT platform as the developmentand execution environment◆Identified system and control design priorto generating trajectory◆Adjusted control design according tosystem needs◆Combined G Dataflow with FSM modelNext Generation MotionTrajectoryLearn SystemQualitative optimal designcommandsysidcontrolhomeget datasend dataSimulated PlantEmbedded ControllerHost ProgramPCRT(PXI)RealPlantControl Design CycleSpecify systemand interfacePhysical modelControl modelAlgorithmDesignAlgorithmDesignPrototypeEmbeddedSystemFinalImplementationHardwarein the loop testsDownloadEmbedded VIACTUAL HARDWARESIMULATIONLABVIEW RTAccomplishments◆Proved that– G is strictly bounded in memory– G schedule guarantees complete execution– Graph determinism can be verified (O(|actors|3))◆Implemented– Proposed structure (PID, MIMO)– Transparent Hardware in the Loop simulation– Novel LQR based qualitative design strategy– Oversampled simulationConclusion◆G is a well behaved language that shares severalfeatures with SDF, PN, and BDF models◆Current motion control products can be mademore flexible and responsive◆RT based design adds flexibility and reducesdevelopment time◆Future developments– Include on-the-fly host based system identification– Incorporate other control and optimal
View Full Document