MMAV - A MINIATURE UNMANNED AERIAL VEHICLE
6.7MMAV - A MINIATURE UNMANNED AERIAL VEHICLE (MINI-UAV)FOR METEOROLOGICAL PURPOSESMarco Buschmann*, Jens Bange, Peter VörsmannTechnische Universität Braunschweig, Germany1INTRODUCTIONMicro Aerial Vehicles (MAV) form acomparably new area of aeronautical research.This type of aircraft is defined by take-off weightstypically less than 500 g and very smalldimensions, e.g. wingspans under 50 cm.Intense research is conducted worldwide tofurther reduce MAV mass and size (e.g.Grasmeyer, 2001; Wu, 2004).Within this paper, this term is expanded toalso include so-called miniature unmanned aerialvehicles (Mini-UAV), which comprise typically ofaircraft with a wingspan of a few meters andseveral kg of take-off mass.Most current MAV projects concentrate onsize and mass but neglect autonomousoperation which means a pilot on the ground hasto control the aircraft manually. The mainresearch activity at the Institute of AerospaceSystems (ILR) of the Technische UniversitätBraunschweig, Germany, is the development ofa fully autonomous MAV, operating without anyintervention from ground.Fig. 1: Carolo P50During the last three years, differentconcepts of aircraft were investigated and thenecessary subsystems were developed. Thisresulted in the “Carolo” family of autonomousaircraft with small dimensions and masses with“Carolo P50” being the smallest and lightestautonomous Micro Aerial Vehicle (MAV) with a* Marco Buschmann, Technische Universität Braunschweig,Institute of Aerospace Systems, 38106 Braunschweig,Germany; e-mail: m.buschmann@tu-bs.dewingspan of 50 cm and a mass of 550 g, asshown in figure 1. This paper gives an overviewabout the project and suggests the use of anautonomous MAV for meteorological purposes.2TOWARDS AUTONOMOUS MAV FLIGHTSeveral steps had to be taken to develop anautonomous MAV: The development of asimulation environment, the experimentalderivation of an aerodynamic data set of theaircraft and the design of an appropriatecontroller structure. The following subsectionsdescribe these steps briefly for the work on thesmallest prototype Carolo P50 to show thefunctional principles.2.1 Flightmechanical SimulationSpecial mathematical tools were developedto allow the simulation of the highly dynamicbehavior of a MAV. Basis is a non-linear flightmechanical simulation tool as presented byKordes et al. (2003). It is based on thecommonly used Matlab/Simulink software andconsiders especially MAV-relevant effects likethe motor gyro effect.Besides this, the dynamic behavior ofsensors and actuators were modeled to allow forrealistic simulation of aircraft behavior. Inaddition to several mathematical wind models,real wind field data can be used, which weremeasured by the helicopter-borne turbulenceprobe “Helipod”. This measurement system isfurther described e.g. by Bange and Roth (1999)and Muschinski et al. (2000). This turbulenceprobe is operated by the ILR for meteorologicalmeasurements and allows for wind vectordetermination with high spatial and temporalresolution.2.2 Wind Tunnel TestsOne foundation for MAV simulation is thedetermination of the aircraft’s aerodynamicproperties. For this reason, wind tunnel testswere conducted at the Institute of FluidMechanics (ISM) of the Technsiche UniversitätBraunschweig. By varying the angle of attack,sideslip, elevator, ailerons and flaps, a 5-
ParameterRangeAngle of Attack-10 α 10Sideslip-32 β 32 Elevator-15 η 15 Aileron-15 ξ 15 Flaps-8 κ 12 Tab. 1: Parameters for 5-dimensionalaerodynamic Parameter FieldFigure 2 shows the lift versus drag diagramof Carolo P50. From this diagram, thedimensionless coefficients for the ideal ratio of liftCL* to drag CD* and the minimum glide angle ε*can be derived. These parameters as well as thecorresponding speed V* are:C L * 0.477C D * 0.065ε * 7.8 V * 15.5ms 1Lift versus Drag1,41,2Lift Coefficient CLdimensional parameter field was derived whichformed the basis for realistic flightmechanicalsimulation. Table 1 shows the variation of theseparameters. Especially the sideslip variationtowards comparably high angles is necessary,since MAV operate at flight speeds which canhave the same magnitude as gusts.10,80,6CL* 0.477CD* 0.0650,40,20-0,2-0,4-0,6-0,800,10,20,30,40,5Drag Coefficient CD0,60,7Fig. 2: Lift vs. Drag of Carolo P502.3 Controller StructureFigure 3 shows the overall controller systemstructure including the models of actuators,sensors and wind. The dynamic model of theaircraft provides a basis for the design of theflight control system (FCS) and the developmentof navigation filters using GPS and INS as statede.g. by Winkler et al. (2003). The flight controllerhas a conventional cascaded structure (figure 3).The advantage is that the aircraft is stillcontrollable on a lower level if a higher levelmalfunction occurs. The flight controller consistsof a damping system, a basic controller tostabilize the aircraft’s attitude and an autopilot fortrack, altitude and airspeed control. The highestlevel is the navigation unit, responsible forwaypoint navigation and mission fulfillment.Since the controller was designed for highly agileaircraft, it is comparably simple to adjust thecontroller parameters for larger aircraft withlarger inertia.Fig. 3. Overall Controller System Structure
3THE MAV FAMILY “CAROLO”The goal of the “Carolo” project is thedevelopment of the smallest autonomous MAVpossible. The smallest prototype has a payloadof just 50 g which is not suitable for mostresearch applications. During development of theCarolo P50, several larger aircraft weredeveloped to serve as test bed for controllerdevelopment and subsystem integration. AllCarolo aircraft share the same hardware whichconsists of sensors for attitude and positiondetermination, model plane actuators for drivingthe rudders, and an electrical propulsion system.Figure 4 shows the different subsystems andtheir interconnection. Central element is an onboard computer, which is small and lightweight,yet powerful enough to host the sophisticatedcontrol ControllerModule(SCM)Flight ControllerModule (FCM)R/C receivertelemetryFig. 4: Subsystems of a Carolo aircraft3.1 On-Board Computer StructureThe demands on the on-board computer arevery versatile: On one hand, different types ofelectricalinterfacesareneededforcommunication with the other subsystems. Onthe other hand, sufficient calculating power hasto be provided to host the flight control andnavigation algorithms. In addition, strongconstraints regarding size, mass, and powerconsumption apply for the use within a MAV.For this reason, the on-board computer issplit into two modules: The System ler with versatile analog and digitalinterfaces for the aircraft’s subsystems, and theFlight Controller Module (FCM), a powerful 32 bitRISC microprocessor with 64 MB of RAM tohost even demanding flight control algorithms.Communication between subsystems anddata processing is organized in time slices with afrequency of 100 Hertz. The sensors areconnected to the SCM. The sensor data is sentfrom the SCM to the FCM via a high speed seriallink. The FCM solves the control and navigationalgorithms (state determination, flight controletc.) and the computation results are then routedvia the MSC to the actuators and propulsionsystem.Compared to a single module solution, thissplit computer design results in an increasedlatency of one time slice for the data flow fromthe sensor to the actuators. But hardwaredevelopment and maintenance as well assoftware development and debugging is greatlysimplified. In addition, a modified receiver for amodel plane remote control can be connected tothe SCM, allowing to directly remote control theaircraft. This is an important safety feature duringflight tests when new control algorithms withinthe FCM software are to be tested: Even in theworst case, a malfunction of the FCM cannotblock the remote control signals from the safetypilot on ground and the aircraft can be landedmanually without danger for man or material.While the on-board computer was designedto process the basic flight data needed forattitude and position determination, it offers freecapacities for integrating additional sensors,e.g. for meteorological measurements.3.2 The AutopilotThe main Sensors and the on-boardcomputer were integrated into a single blockserving as autopilot. Since it was designed forminimum weight and size and to control highlyagile aircraft, it is comparably easy to adapt it forthe use in larger aircraft. The current autopilotprototype consists of two printed circuit boardsand has the following characteristics:-6 degree-of-freedom IMU-static and dynamic pressure sensor-16 channel GPS receiver-32 bit computer with 64 MB RAM forflight control algorithms-control of up to 6 servo actuators-input for remote control receiver asbackup for flight tests-capabilities of interfacing to additionaldata sources (sensors)-power consumption: 1.5 W-overall mass: 85 grams-overall size: 40 x 40 x 80 mm³
-50% reduction in weight and sizeexpected by higher integration (prototypescheduled for 09/2004)3.3 Telemetry Link and Data StoragePrincipally, a telemetry link is not needed foran autonomous aircraft. But of course, formission control and adjustment and for receivingpayload data, a telemetry link between theaircraft and Ground Control is necessary.As can be seen in figure 4, the telemetrymodule is connected directly to the Flight ControlModule. This is done by means of a standardasynchronous serial interface, which allows foreasy exchange of the telemetry module itselfaccording to mission demands. The currentprototype incorporates a frequency hoppingspread spectrum radio modem with an effectivedata rate of approximately 20 kbps and a rangeexceeding 1,000 meters. The high data rateallows for detailed sensor and controller stateinformation for in-flight testing, but for practicalapplications, high range is probably much moreimportant than high data range.As alternative for real-time transmission,data can be stored on-board on a Multi MediaCard (MMC), a small solid state storage cardwhich originates from electronic appliances likedigital cameras or Personal Digital Assistants(PDA). This type of storage card is available withcapacities up to 512 Mbytes and can store flightdata for several hours of flight. After landing, thecard is removed and the data is read by acommon personal computer by means of asmall, dedicated card reader hardware.3.4 Ground ControlGround Control consists of a commonpersonal computer and a specially developedsoftware package. The Ground Control softwareis based on a server-client structure:A central server module hosts the data flowbetween different client modules. These modulesprovide e.g. data logging functionality orGraphical User Interfaces for visualizing sensordata or providing a digital map for waypointediting. For the server, the aircraft itself is seenas a client module.Communication between server and clientsis based on the UDP network protocol, allowingthe ground control modules to run on a single PCor on different computers connected by localnetwork or the internet. For connecting to theMAV, special telemetry hardware is used, aspreviously described. The choice of using arather sophisticated data protocol increases theworkload of the Flight Controller Module onboard, but greatly increases flexibility. It is alsothe basis for controlling multiple MAV from oneGround Control in the future. This allows forcontrolling whole swarms of MAV.Fig. 5: Mission Control ModuleThe main client module for the user is theMission Control software module, which providesa digital map of the operational area. This mapcan consist of a digitized topographic or citymap, combined with elevation information, ordigital landscape models. Figure 5 shows ascreenshot of the current software version. Inthis case, a digitized air photograph combinedwith a digital elevation model was used. Theuser can set waypoints on the map, thusdetermining the MAV's flight path. With eachpoint, special actions can be associated, e.g.“circling in constant height for 60 seconds'”.The mission is planned before take-off andtransmitted to the MAV via the server and thetelemetry module. During mission, the actualposition of the MAV is displayed within the digitalmap, allowing supervision of the MAV's route.The flight path can be adapted manually at anytime by editing the flight path on the map andsending a waypoint update to the MAV.4THE METEOROLOGICAL MAV (M²AV)Since the ILR has several years ofexperience in the field of airborne ence probe Helipod, it stands to reason touse the present knowledge in the field of realtime meteorological measurement technology todevelop an autonomous meteorological microaerial vehicle, shortly called M²AV.
Fig. 6: Carolo T140As mentioned before, a conventional modelplane is used as test bed for hard- and softwaredevelopment and testing (figure 6). This aircraftis a twin engine design with a wingspan of140 cm (hence the name T140) and a maximumtake-off weight of 2 kg, including 300 g ofpayload. It is hand-launched which makeshandling and operating the aircraft very easy.With an endurance of approximately 30 minutes,the range accounts for 27 km at a cruising speedof 15 m/s. To restore operational readiness afterlanding, the batteries simply have to berecharged or replaced by a fresh battery pack.4.1 Possible M²AV MissionsThe described properties of the Carolo T140makes it an ideal test platform for a M²AV tomeasure the basic atmospheric parameterstemperature, humidity and wind vector.The operating altitude is between 10 m(appropriate elevation model to avoid obstaclesprovided) and 1 km above ground, making itespecially suitable for boundary-layer research.Higher altitudes are possible at the expense ofendurance. By using special lifting aids likeballoons, the M²AV could act as an auto-homingradio probe which returns automatically afterreleasing the balloon at a specified altitude.However, constraints like icing in the uppertroposphere have to be considered carefully.By using multiple M²AV simultaneously, thearea coverage can be drastically increasedcompared to existing meteorological systems.4.2 Measuring Temperature and HumidityThe basic constraint for developing MMAVsensors is the limitation of payload mass. Fortemperature, a split sensor concept as applied inthe Helipod seems advisable: Two temperaturesensors with different characteristics will beused: One sealed Pt100 element with highaccuracy but slow response time in themagnitude of 10 s and one open element with afragile mechanical design and rather poor longterm stability but very fast response time in therange of 10 ms. By complementary filtering, thecharacteristics of both sensors can be combined:Long-term stability with high accuracy and fastresponse time. The possible accuracy dependsstrongly on the data acquisition system, and heremainly the analog sensor supply electronics andthe analog-to-digital conversion. Fortunately,very precise and compact components areavailable today which allow a data samplingprecision of 16 bit and better, resulting in aachievable resolution of several Millikelvin for ameasurement range from –40 to 60 C.For measuring humidity, few sensors areoffered which fulfill the requirements regardingsize and mass. This limits the possibilities torather slow sensors with response times in themagnitude of 10 s and accuracies about 2%.Depending on the required precision, theintegration of temperature and humidity sensorsin the Carolo P50 and T140 type of aircraftseems feasible, with using a simplified, onesensor temperature measurement scheme in theP50. First tests with the T140 are scheduled forthis year (2004).4.3 Measuring the Wind VectorSince one focus of meteorological researchat the ILR is the investigation of turbulent fluxesin the boundary layer, a M²AV will be equippedwith a miniature 5-hole probe. This probe wasdeveloped and manufactured by the Institute ofFluid Mechanics (ISM). Figure 7 shows thecomplete probe and figure 8 shows the probe’stip in comparison with a 1-euro-cent coin and.
Fig. 7: Miniature 5-hole ProbeFig. 9: Pressure Transducer Test Board#SensorRange1dynamic pressure0. 1250 Pa4relativ epressure-250 . 250 Pa1static pressure20 kPa.105 kPaTab. 2: Pressure Transducer ElectronicsFig. 8: Tip of Miniature 5-Hole ProbeThe 5-hole probe has a mass of 22 g and adiameter of 6 mm. It is intended for themeasurement of angles of attack and sideslip inthe range of –45 to 45 each.For deriving the angle of attack and thesideslip, the relative pressures between the5 holes will be measured. For a first test ofpressure transducers and analog front-end, asmall electronic board which incorporatesaltogether6 pressuretransducerswasdeveloped (figure 9). Table 2 shows figurations are scheduled for this year, withintegration into the Carolo T140 taking place in2005.4.4 Data Acquisition and ProcessingThe implemented autopilot hardware asdescribed before offers some free capacities fordata acquisition. The autopilot sensors forattitude determination (accelerometers andgyros) are sampled in time slices with afrequency of 100 Hz.Additional sensors can be connected bymeans of a synchronous serial interface. Theautopilot also includes sensors for deriving thealtitude from the static pressure and the velocityfrom the dynamic pressure, but the quality ofthese data is not sufficient for meteorologicalpurposes. Thus, external sensors have to beconnected for every meteorological measurandof interest.A variety of small analog-to-digital converterswith power consumption in the magnitude ofseveral 10 mW are available. Depending of theconverter’s topology, resolutions of 16 bit andmore are possible with data rates in excess of100 Hz. These electronic components can beinterfaced directly to the synchronous serialinterface of the on-board computer. Externalsensor data rate is limited to approximately40 bytes per timeslice. This is sufficient for e.g.
8 sensor channels with a resolution of 16 bits fortemperature, humidity and the differentialpressures delivered by the 5-hole probe, andone sensor channel with higher resolution formeasuring the static pressure.Institute ofFluid Mechanics (ISM) of theTechnische Universität Braunschweig, Germany,for providing the miniature 5-hole probe.5REFERENCESCONCLUSION AND OUTLOOKAt the Institute of Aerospace Systems, afamily of autonomous Micro Aerial Vehicles(MAV) was developed. The aircraft are controlledby setting waypoints on a digital map by meansof a special software running on a commonlaptop or TabletPC. Telemetry link from GroundControl to the aircraft is currently done with adedicated radio modem with short range(1000 m), but due to the modular design,replacement by other telemetry hardware withmuch higher range is comparably easy. Sincethe aircraft operate autonomously, telemetry linkis only necessary for supervision and missionupdates and is not vital for aircraft operationitself. Payload data can be stored on solid statememory cards, allowing high volumes of datastorage and easy handling of data after landing.Two MAV prototypes, Carolo P50 andCarolo T140, are available for meteorologicalpayload integration: The Carolo P50 has apayload of 50 g and is suitable for roughmeteorological measurements of temperatureand humidity. It could act as a flexible tool fordelivering rough meteorological key data duringfield experiments.The second prototype, Carolo T140, allowsfor payloads up to 300 g and will be equippedwith high precision measurement of temperature,humidity and wind vector. This meteorologicalmicro aerial vehicle (M²AV) can be used toexplore the lower atmosphere, especially theboundary layer. Its main advantage compared toexisting systems is the easy handling and thepossibility to cover big areas simultaneously byusing multiple MAV.The Carolo T140 has an endurance ofapproximately 30 minutes which results in arange of 27 km. A new aircraft “Carolo P200” isunder development with a wingspan of 200 cmand a maximum take-off weight of 4 kg. Thissensor carrier will expand the possibilities of theCarolo family especially for missions whichdemand for higher operational altitudes or longerendurance.6ACKNOWLEDGEMENTSThe authors wish to thank the EuropeanUnion for partially financing the project and theBange J. and Roth R., 1999: Helicopter-BorneFlux Measurements in the NocturnalBoundary Layer Over Land - a Case Study.Boundary Layer Meteorol., 92, 295-325.Grasmeyer, J. M. and M. T. Keennon, 2001:Development of the Black Widow micro airvehicle. AIAA 39th Aerospace SciencesMeeting and Exhibit, Reno, NV, Jan. 8-11,2001.Kordes, T., Buschmann, M., Schulz, H.-W.,Vörsmann, P., 2003: Progresses in theDevelopment of the Fully Autonomous MAV"CAROLO", Proceedings of 2nd AIAA"Unmanned Unlimited", San Diego,CaliforniaMuschinski A, Frehlich R, Jensen M, Hugo R,Hoff A, Eaton F, Balsley B, 2000: Fine-scalemeasurements of turbulence in the lowertroposphere: An intercomparison between akite- and balloon-borne, and a helicopterborne measurement system. Bound-LayerMeteor 98, p. 219-250.Winkler, S.; Buschmann, M.; Kordes, T.; Schulz,H.-W.; Vörsmann, P., 2003: MAV StateEstimation Using One and Multiple GPSAntennas. GNSS 2003 – The EuropeanNavigation Conference, Graz, Austria, 22-25AprilWu, H., Sun, D., Zhou, Z., 2004: Micro AirVehicle: Configuration, Analysis, Fabrication,and Test, IEEE/ASME Transactions onMechatronics, vol. 9, no. 1, p. 108-117
6.7 MMAV - A MINIATURE UNMANNED AERIAL VEHICLE (MINI-UAV) FOR METEOROLOGICAL PURPOSES Marco Buschmann*, Jens Bange, Peter Vörsmann Technische Universität Braunschweig, Germany 1 INTRODUCTION Micro Aerial Vehicles (MAV) form a comparably new area of aeronautical resea
awards those who can respond to "JAUS" for Networked unmanned systems SOURCE: : Crouse, J. (2008). The joint architecture for unmanned systems: a subsystem of the wunderbot 4. Elizabethtown College research report. MORE ON THIS in COMPUTING LECTURES Wunderbot 4 Wireless Communication by Jeremy Crouse (advisor: Dr. Wunderlich)
STANAG 4586(NATO Standardization Agreement 4586) is a NATO Standard Interface of the Unmanned Control System (UCS) Unmanned Aerial Vehicle (UAV) interoperability. It defines architectures, interfaces, communication protocols, data elements and message formats. It includes data
Linear Rail System Ball Screw Support Unit Linear Ball Bush Cross Roller Guide Robot Carrer Guide Memo Miniature Linear Rail System Miniature Linear Rail System Miniature through tap hole rail Caution for mounting miniature through tap hole rail Model W1 W3 H1 S G F L0 (Max length) Mass (kg/m) SBM 09-B 9-5.5 M4x0.7 7.5 20 1195 0.32 SBM 12-B 12 .
Thomson as your supplier brings some additional advantages as well. Miniature Lead Screws Miniature Brakes Miniature Linear Motion Systems Customization Thomson Advantages Advantage Benefits Widest variety of miniature linear products on the market Expedited design time Single source of engineering support Consolidated supply base
avian influenza. This study led to the construction of deep-learning-based object-detection models with the aid of aerial photographs collected by an unmanned aerial vehicle (UAV). The dataset containing the aerial photographs includes diverse images of birds in various bird habitats and in the vicinity of lakes and on farmland.
Aerial Platform Lift Operator Safety Training Aerial Platform Mobile Lift Equipment for Construction & Industrial Safety Risk Management Department 6/24/2010 RSK-W505 REV. A JUNE 24, 2010 Page 1 of 6. QUICK CARD TM Aerial Lifts Safety Tips Aerial lifts
Aerial Lift Safety Program. Provide specific operational training for each aerial lift. Observe the operation of aerial lifts, and correct unsafe practices. 6.3 Operators Read the Aerial Lift Safety Program. Complete the Daily Pre-Use Inspection Che
Principles of Animal Nutrition Applied Animal Science Research Techniques for Bioscientists Principles of Animal Health and Disease 1 Optional Physiology of Electrically Excitable Tissues Animal Behaviour Applied Agricultural and Food Marketing Economic Analysis for Agricultural and Environmental Sciences Physiology and Biotechnology option Core Endocrine Control Systems Reproductive .
