Jtam.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 50, 1, pp. 251-268, Warsaw 2012 50th anniversary of JTAM MODELING AND NUMERICAL SIMULATION OF UNMANNED AIRCRAFT VEHICLE RESTRICTED BY NON-HOLONOMIC CONSTRAINTS Edyta Ładyżyńska-Kozdraś Warsaw University of Technology, Faculty of Mechatronics, Warsaw, Poland e-mail: e.ladyzynska@mech.pw.edu.pl The paper presents themodeling of flight dynamics of an unmanned air- craft vehicle (UAV) using the Boltzmann-Hamel equations for mechani- cal systemswith non-holonomic constraints. Control laws have been tre- ated as non-holonomic constraints superimposed on dynamic equations of motion of UAV. The mathematical model containing coupling dy- namics of the aircraft with superimposed guidance have been obtained by introducing kinematic relationships as the preset parameters of the motion resulting from the process of guidance.The correctness of the de- velopedmathematicalmodelwas confirmedby the carriedout numerical simulation. Key words: automatically steered aircraft vehicle, non-holonomic con- straints, Boltzmann-Hamel equations 1. Introduction In recent years, unmanned aircraft vehicles (UAV) have been the fastest gro- wing means of recognizing and supporting the armed forces. On the modern battlefield, light small-sized UAVs perform the mission of ground target de- tection, tracking and illumination. Their modified versions, combat UAVs, are supposed not only to autonomously detect the target, but also destroy it with on-deck homing missiles (Fig.1). UAVs usefulness has been confir- med unequivocally both by operations carried out by the Americans in Iraq and Afghanistan, as well as Israeli military operations in Lebanon and Ga- za. No wonder that acquiring a UAV for the Polish army has become one of the priorities of its technical modernization in the years 2009-2018. For the same reasons, the problems associated with the study of dynamics and 252 E. Ładyżyńska-Kozdraś control of the UAV is the focus of many scientific and research centres in Poland and the world. There are also many methods of modelling these issu- es from the equations of classical mechanics using the principle of changing angular momentum (Dogan and Venkataramanan, 2005; Ducard, 2009; Etkin and Reid, 1996; Maryniak, 2005; Rachman and Razali, 2011; Sadraey and Colgren, 2005), to themethods of analytical mechanics for nonholonomic sys- tems. Analytical mechanics provides several methods of generating equations of motion of flying objects. Boltzmann-Hamel equations, as well as Maggi’s, Gibbs-Appel,Keyn equations, and theprojectivemethoddevelopedbyProfes- sorW. Blajer, used in the study, are among them (Blajer, 1998; Blajer et al., 2001; Chelaru et al., 2009; Graffstein et al., 1997; Gutowski, 1971; Koruba and Ładyżyńska-Kozdraś, 2010; Ładyżyńska-Kozdraś, 2008, 2009, 2011;Maryniak, 2005; Sadraey and Colgren, 2005; Ye et al., 2006). Fig. 1. General view of combat UAVmission performance (Koruba and Ładyżyńska-Kozdraś, 2010) Thedynamicalmodeling ofUAVforms theheart of its simulation (Sadraey and Colgren, 2005). The numerical simulation of the aircraft dynamics is the most important tool in the development and verification of the flight control laws and equations ofmotion for a UAV.The ability to test autopilot systems in a virtual (software) environment using a software flight dynamics model for UAVs is significant for development, as shown for example in the work by Blajer et al. (2001), Chelaru et al. (2009), Dogan andVenkataramanan (2005), Jordan et al. (2006), Ładyżyńska-Kozdraś (2011), Ou et al. (2008), Shim et Modeling and numerical simulation of UAV... 253 al. (2003). In many cases, testing newly developed autopilot systems in a virtual environment is the onlyway to guarantee absolute safety. Additionally, the model would allow better repeatability in testing, with controlled flying environments. This study is a continuation and generalization of the article published in the Journal of Theoretical and Applied Mechanics (Koruba and Ładyżyńska- -Kozdraś, 2010), where was proposed an algorithm of guiding combat UAV, which on having autonomously detected targets, attack them (e.g. radar sta- tions, combat vehicles or tanks) or illuminate them with a laser (Fig.1). The simplified equations of UAVmotion developed there were now generalized for the object with nonholonomic constraints. In the article, a novel method was proposed based on treating the nonholonomic constraints as laws of control- ling and conjugating themwithnonlinear equations ofmotion of an object and with kinematic guidance relations (Ładyżyńska-Kozdraś, 2008, 2009, 2011). Control, that is override aimed at ensuring that themoving object behaves in a desired manner, was brought to testing of differences, that is deviations between the required and actual value of the realized coordinate. The value of this difference, after appropriate strengthening and transforming, resets the deviation. Developed in this way control laws were treated as non-holonomic constraints superimposed on themotion of aUAV. Linkage of these equations with the dynamic equations of motion of the object made it possible, thanks to the use of Boltzmann-Hamel equations for mechanical systems with non- -holonomic constraints, to control the flight of a UAV in an effective manner. 2. Physical model of UAV and adopted reference systems The article presents the process of modeling and numerical simulation of the flight of unmanned aircraft vehicle during the mission. Appropriate formula- tion of the physical model is an important element here, which will constitute the basis for building a mathematical model of motion of the tested object. The following assumptions of the physical model have been adopted: • UAV is treated as a non-deformable object, with six degrees of freedom resulting from themovement of rudders; • Motion of a UAV is examined in calm weather; • UAVweight varies during the flight; • UAV motion control may be carried out in four channels: in the pitch channel θ – by means of elevator deflections δH, in the yaw channel ψ 254 E. Ładyżyńska-Kozdraś – by means of rudder deflections δV , in the roll channel φ – by means of aileron deflection δL, and in the speed channel V0 – by changing the engine thrust δT ; • Control units are moving, but they are non-deformable; • Deflection of surfaces of rudders affects the forces andmoments of aero- dynamic forces; • Constraints superimposed on a UAV resulting from the adopted control laws have been treated as non-holonomic constraints; • Forces andmoments of aerodynamic forces, from the propulsion system and gravitation, act on the UAV; • The impact of the curvature of the Earth was ignored; • The environment has an impact on the dynamic properties of the drive through changes in air temperature tH, air density ρH, air pressure pH, kinematic viscosity νH, sound speed aH depending on the altitude. Fig. 2. Adopted reference systems, linear and angular speed of a UAV The motion of a unmanned aircraft vehicle is examined in the reference system Oxyz, rigidly connectedwith themoving object, with the beginning in the center ofmass of theUAVafter burnupof fuel (Fig.2).Theother reference system used in the paper includes: system O1x1y1z1, rigidly connected with theEarth, as well as systems connectedwith the object, namely: gravitational Oxgygzg parallel to O1x1y1z1 system and velocity Oxayaza connected with the air flow direction (Fig.3). Modeling and numerical simulation of UAV... 255 3. Kinematic correlations and guidance correlations Themotion of a unmanned aircraft vehicle during amission is describedwith the use of coordinates and time in the space of eventswhere the location of the object is uniquely determined with the use of linear and angular coordinates in the configuration space. Pursued flight parameters of the UAV are read automatically by the gu- idance system and depend only on the actual behavior of the guided object on the track, and thus on the changes of its linear and angular position. The vector of the actual linear velocity of the UAV (Fig.2) in the Oxyz system is V 0 = Ui+Vj+Wk (3.1) where: U, V , W are respectively: longitudinal, lateral and vertical velocity; i,j,k – unit vectors of the system associated with the Oxyz object. The angular velocity vector Ω= Pi+Qj+Rk (3.2) where: P,Q,R are the angular velocity of banking, tilt anddeflection (Fig.2). Kinematic correlations between the components of angular velocities and derivatives of the angles have the following form (Blajer, 1998; Etkin andReid, 1996; Ładyżyńska-Kozdraś, 2009, 2011)    φ̇ θ̇ ψ̇    =    1 sinφtanθ cosφtanθ 0 cosφ −sinφ 0 sinφsecθ cosφsecθ       P Q R    (3.3) Kinematic correlations between the components of the linear velocity ẋ1, ẏ1, ż1 measured in the system O1x1y1z1 rigidly connected with the Earth and the components of U,V,W velocity in the reference system Oxyz asso- ciated with the moving objects are as follows (Blajer, 1998; Etkin and Reid, 1996; Ładyżyńska-Kozdraś, 2009, 2011)    ẋ1 ẏ1 ż1    =        cosψcosθ cosψsinθsinφ+ −cosφcosψ cosψ sinθcosφ+ +sinφcosψ sinψcosθ sinψsinθsinφ+ −sinφcosψ sinψ sinθcosφ+ −sinφcosψ −sinθ sinφcosθ cosφcosθ           U V W    (3.4) In calmweather the angle of approach α and glide β are expressed by the following formulas (Blajer, 1998; Etkin and Reid, 1996; Ładyżyńska-Kozdraś, 2009, 2011): 256 E. Ładyżyńska-Kozdraś —the angle of approach α =arctan W U (3.5) — the angle of glide β =arcsin V V0 (3.6) Preset flight parameters of the unmanned aircraft result from the adop- ted guidance method. Assuming that during the observation of the area the aircraft flies along the pre-programmed route, constraints are superimposed on the location and angular and linear velocity of the object, thus creating the generator of its programmed motion (Blajer, 1998; Blajer et al., 2001; Ładyżyńska-Kozdraś, 2011) K̂0Ke = s(x1p,y1p,z1p,φp,θp,ψp) Vz(t)= ṡ(ẋ1p, ẏ1p, ż1p, φ̇p, θ̇p, ψ̇p,φp,θp,ψp) (3.7) where: x1p, y1p, z1p, φp, θp, ψp are the preset parameters of motion of the controlled object. When theUAVdetects the target, it is possible to trace or attack it. In this case, the preset parameters of the controlled object were selected in the paper with the use of one of self-guidancemethods, that is themethod of “along the curve of hunt” (Koruba andŁadyżyńska-Kozdraś, 2010; Ładyżyńska-Kozdraś, 2011). It assumes that the preset angular coordinates of the guided object are consistent with the parameters of the target φp = φt θp = θt ψp = ψt (3.8) The vector of the preset location of UAV in O1x1y1z1 system (Fig.2) is rp = x1pi1+y1pj1+z1pk1 (3.9) where x1p = rotcosψtcosθt y1p =−rot sinψtcosθt z1p =−rot sinθt and rot is the distance of the actual location of the object to the target, φt,θt,ψt – angles of banking, tilt and deflection of the maneuvering target. Modeling and numerical simulation of UAV... 257 The components of the vector of the preset linear velocity of UAV in the own system are    Up Vp Wp    =        cosψtcosθt sinψtcosθt −sinθt sinφtcosψt sinθt+ −sinψtcosφt sinφt sinψt sinθt+ +cosψtcosφt sinφtcosθt cosφtcosψt sinθt+ +sinψt sinφt cosφt sinψt sinθt+ −cosψt sinφt cosφtcosθt           ẋ1p ẏ1p ż1p    (3.10) The components of the vector of the preset angular velocity of UAV in the own system are    Pp Qp Rp    =    1 0 −sinθt 0 cosφt sinφtcosθt 0 −sinφt cosφtcosθt       φ̇t θ̇t ψ̇t    (3.11) During the flight of the unmannedaircraft, the preset parameters resulting from the adopted guidance method are compared with the parameters of the flight of the aircraftwhich are read on the ongoingbasis.Thedifferenceswhich occur between these parameters are then removed by the automatic control system. In this way, the control mechanism affects motion of the object by means of relevant control units, depending on changes in one or, most often, many parameters of its motion. 4. Control laws Themotion of the unmanned aircraft along the preset trajectory results from the superimposed kinematic constraints. These constraints are, however, vio- lated due to various external interferences, construction deficiencies, etc. It is therefore necessary to equip the object with a set of devices that will determi- ne the degree of violation of these constraints and will generate appropriate control signals so that the aircraft will fly along the required track. All these tasks are pursued by the control system. In this way, through the appropriate control units, themechanism of auto- matic guidance of the object affects its movement, depending on one or, most often, many parameters such as acceleration, speed, linear and angular posi- tion, track angle. This affects the dynamics of the controlled object, causing a change of control forces which enforce the flight consistent with the adopted 258 E. Ładyżyńska-Kozdraś control system and guidance algorithm. The automatic control system, basing on the values of deviations which compare information about the state of the controlled object and the vector of the preset state, generates control signals. Kinematic and geometric deviations relationships in the servomechanisms be- come strengthened and then they are transferred to the actuators, such as hydraulic, electrohydraulic or electromechanical cylinders. The delay of the control system has been described by means of an inertial unit of the first order. Automatic control of the unmanned aircraft is carried out in four channels: in the pitch channel θ – by means of elevator deflections δH, in the yaw channel ψ – by means of rudder deflections δV , in the roll channel φ – by means of aileron deflection δL, and in the speed channel V0 – by changing the engine thrust δT . Control laws of the UAV take the following form: — in the pitch channel TH3 δ̇H +T H 2 δH = K H x1 (x1−x1p)+K H z1 (z1−z1p)+K H U (U −Up) +KHW(W −Wp)+K H Q (Q−Qp)+K H θ (θ−θp)+ δH0 (4.1) — in the yaw channel TV1 δ̇v +T V 2 δV = K V y1 (y1−y1p)+K V V (V −V1p)+K V W(W −Wp) +KVR(R−Rp)+K V ψ (ψ−ψp)+δV0 (4.2) — in the roll channel TL1 δ̇L +T L 2 δL = K L φ(φ−φp)+K L P(P −Pp)+K L V (V −Vp) +KLR(R−Rp)+K L ψ(ψ−ψp)+δL0 (4.3) — in the velocity channel TT1 ˙δT +T T 2 δT = K T x1 (x1−x1p)+K T U(U −Up)+K T W(W −Wp) +KTθ (θ−θp)+K T Q(Q−Qp)+K T R(R−Rp)+K T ψ(ψ−ψp)+ δT0 (4.4) where: Tji are time constants, K j i – amplification coefficients. Modeling and numerical simulation of UAV... 259 The designated control laws are non-integrable and impose restrictions on the motion system, and therefore they were considered as kinematic equ- ations of non-holonomic constraints (Bloch, 2003; Ładyżyńska-Kozdraś, 2011; Nejmark andFufajew, 1971). In the paper, the aforementioned equationswere linkedwith the dynamic equations ofmotion of the unmannedaircraft vehicle, derived using analytical equations of mechanics in the form of the Bolzmann- Hamel equations with multipliers. This approach made it possible to remove differences resulting from the dynamics of motion of the controlled object occurring between geometric, ki- nematic and dynamic preset parameters and their pursuance during the flight of the tested object. 5. General equations of motion of the unmanned aircraft vehicle – Boltzmann-Hamel equations for mechanical systems with non-holonomic constraints Description of dynamics of theunmannedaircraft, treatedas anon-deformable mechanical system, was made in the reference system rigidly connected with the Oxyz object. In addition, control laws (4.1)-(4.4) were treated as non- holonomic constraints superimposed on motion of the system. Therefore, in order to determine the equations of motion, the Boltzmann-Hamel equations ofmotion of non-holonomic systems in the generalized coordinates (Graffstein et al., 1997;Gutowski, 1971; Ładyżyńska-Kozdraś, 2011;Maryniak, 2005)were used. From the Boltzmann-Hamel equations, after calculating the value of Bolt- zmann multipliers and determination of kinetic energy in quasi-velocities, a system of ordinary second-order differential equations was obtained: — the equation of longitudinal motion d dt (∂T∗ ∂U ) − ∂T∗ ∂V R+ ∂T∗ ∂W Q− ∂T∗ ∂x1 cosψcosθ− ∂T∗ ∂y1 sinψcosθ + ∂T∗ ∂z1 sinθ+ ∂T∗ ∂δH ( QKHW −K H x1 cosψcosθ+KHz1 sinθ− TH2 TH1 KHU ) + ∂T∗ ∂δV ( QKVW −RK V V +K V y1 sinψcosθ ) − ∂T∗ ∂δL RKLV + ∂T∗ ∂δT ( QKTW −K T x1 cosψcosθ− TT2 TT1 KTU ) = QX (5.1) 260 E. Ładyżyńska-Kozdraś —the equation of lateral motion d dt (∂T∗ ∂V ) − ∂T∗ ∂x1 (sinφcosψ sinθ− sinψsinφ) − ∂T∗ ∂y1 (sinφsinψsinθ+cosψcosφ) − ∂T∗ ∂z1 sinφcosθ− ∂T∗ ∂U R+ ∂T∗ ∂W P + ∂T∗ ∂δL TL2 TL1 KLV (5.2) + ∂T∗ ∂δH [PKHW −K H x1 (sinφcosψ sinθ− sinψ sinφ)−KHz1 sinφcosθ +RKHU ]+ ∂T∗ ∂δV [ −PKVW + TV2 TV1 KVV −K V y1 (sinφsinψsinθ+cosψcosφ) ] + ∂T∗ ∂δT [−PKTW −K T x1 (sinφcosψ sinθ− sinψsinφ)+RKTU ] = QY —the equation of climbing motion d dt (∂T∗ ∂W ) − ∂T∗ ∂x1 (cosφcosψ sinθ+sinψ sinφ) − ∂T∗ ∂y1 (cosφsinψ sinθ− cosψsinφ) − ∂T∗ ∂z1 cosφcosθ+ ∂T∗ ∂V P − ∂T∗ ∂U Q+ ∂T∗ ∂δL PKLV − ∂T∗ ∂δH [(cosφcosψ sinθ+sinψsinφ)KHx1 +(cosφcosθ)KHz1 − TH2 TH1 KHW +QK H U ] + ∂T∗ ∂δV [ −PKVW + TV2 TV1 KVV − (cosφsinψ sinθ− cosψ sinφ)K V y1 ] + (TV2 TV1 KTW +QK T U )∂T∗ ∂δT = QZ (5.3) — the equation of roll motion d dt (∂T∗ ∂P ) − ∂T∗ ∂φ + ∂T∗ ∂U (−RKHQ +V K H W)+ ∂T∗ ∂W V − ∂T∗ ∂V W + ∂T∗ ∂R Q− ∂T∗ ∂Q R+ ∂T∗ ∂δV (V KVW −WK V V +QK V R) (5.4) + ∂T∗ ∂δL ( QKLR −K L φ −WK L V + TL2 TL1 KLP ) + ∂T∗ ∂δT (V KTW −RK T Q)= QL Modeling and numerical simulation of UAV... 261 —the equation of pitch motion d dt (∂T∗ ∂Q ) − ∂T∗ ∂φ sinφtanθ− ∂T∗ ∂θ cosφ− ∂T∗ ∂ψ sinφ cosθ − ∂T∗ ∂W U + ∂T∗ ∂U W − ∂T∗ ∂R P + ∂T∗ ∂P R+ ( WKHU +UK H W +K H θ cosφ+ TH2 TH1 KHQ )∂T∗ ∂δH (5.5) + ( −UKVW −PK V R +K V ψ sinφ cosθ ) ∂∗ ∂δV +(−RKLP +K L φ sinφtanθ) ∂T∗ ∂δL + ( −UKTW +WK T U +K T θ cosφ−K T ψ sinφ cosθ + TT2 TT1 KTQ )∂T∗ ∂δT = QM —the equation of yaw motion d dt (∂T∗ ∂R ) − ∂T∗ ∂φ cosφtanθ− ∂T∗ ∂θ sinφ+ ∂T∗ ∂ψ sinφ cosθ + ∂T∗ ∂V U − ∂T∗ ∂U V − ∂T∗ ∂P Q+ ∂T∗ ∂Q P + ∂T∗ ∂δH (−V KHU +K H θ sinφ) + ∂T∗ ∂δV ( UKVV + TV2 TV1 KVR −K V ψ sinφ cosθ ) + ∂T∗ ∂δL ( QKLP + TV2 TV1 KLR −K L φ cosφtanθ−K L ψ sinφ cosθ ) + ∂T∗ ∂δT ( PKTQ +K T θ sinφ−K T R cosφtanθ−K T ψ sinφ cosθ ) = QN (5.6) — the equation of elevator deflections d dt (∂T∗ ∂δ̇H ) − ∂T∗ ∂δH = QδH (5.7) — the equation of direction rudder d dt (∂T∗ ∂δ̇V ) − ∂T∗ ∂δV = QδV (5.8) — the equation of aileron motion d dt (∂T∗ ∂δ̇L ) − ∂T∗ ∂δL = QδL (5.9) — the equation of drive unit d dt (∂T∗ ∂δ̇T ) − ∂T∗ ∂δT = QδT (5.10) 262 E. Ładyżyńska-Kozdraś This set of equations constitutes, after the determination of kinetic ener- gy and components of forces and moments of generalized forces, the general mathematical model of motion of the unmanned aircraft vehicle. Kinetic energy of the flying object moving in any spatial motion, in the reference system Oxyz associated with it, has been expressed in linear and angular quasi-velocities. Since the UAV is treated as a non-deformable object with movable control systems, its total kinetic energy is the sum of the ki- netic energy of the object without surfaces of rudders and kinetic energy of particular control units. Fig. 3. Vectors of force and the moment of the external force as well as their components in the reference system Oxyz Right sides of the equations create forces F and moments of forces M generated by external loads acting on the object in the smooth configu- ration (Fig.3), composed of forces and gravitational moments Qg, genera- ted by the drive QT , control Qδ and aerodynamic Qa forces (Koruba and Ładyżyńska-Kozdraś, 2010; Ładyżyńska-Kozdraś, 2011;Nejmark andFufajew, 1971).When determining themoments of forces acting onmovable surfaces of rudders, the occurrence of moments from active forces generated by the drive of control elements (electric motor) (MHN, MV N, MLN), hinge moments (MHZ, MV Z, MLZ) and moments of forces generated by the inhibitory re- sponse resulting fromthe inertia of the system (MHR,MV R,MLR) are taken into account (Graffstein et al., 1997; Ładyżyńska-Kozdraś, 2011). The equations of motion of the unmanned aircraft, derived with the use of Boltzmann-Hamel equations (5.1)-(5.10) together with equations of non- holonomic constraints (4.1)-(4.4) and equations of kinematic correlations and guidance correlations (3.1)-(3.11), constitute the set of ordinary differential equations,with theuse ofwhichonemay, at given initial conditions, determine Modeling and numerical simulation of UAV... 263 16 unknowns being a function of time determining the components of linear and angular velocity of the object U, V , W , P, Q, R, its temporary location on the route during the guidance x1, y1, z1, φ, θ, ψ and angles of deflections of aerodynamic rudders δH, δV , δL, δT . The preset flight parameters of UAV resulting from the adopted guidan- ce method, (3.7)-(3.11), were included in equations of dynamics (5.1)-(5.10) through suitable amplification coefficients Kji in particular control channels. This way, they were included in themathematical model of the UAV for con- jugating parameters presetwith the realized state of the flight. This revealed a close relationship of dynamic equations of motion of a controlled flying object with the laws of control and kinematic relations, which as a whole, form a system of nonholonomic constraints. The equations derived in such awaymake it possible to conductmathema- tical calculations for controlled aircrafts: airplanes, unmanned aircrafts, roc- kets and aerial bombs (Koruba and Ładyżyńska-Kozdraś, 2010; Ładyżyńska- Kozdraś, 2008, 2011). A separate issue is the proper selection of amplification coefficients in the control laws, which problem, still remaining in research (Blajer et al., 2001; Graffstein, 2009), has been examined by the author, among others, in Łady- żyńska (2009). When selecting the reinforcement coefficients of the autopilot, an integral, quadratic criterion of quality control was used in the study, which complemented the assessment of transitional processes in all control channels J = 4 ∑ i=1 tk ∫ 0 [yi(t)−yzi(t) yimax ]2 dt (5.11) where: yi(t) – stands for the actual course of the variable, yzi(t) – denotes the predetermined course of the variable, yimax is the maximum preset range of the i-th state variable or the preset value yzi of the i-th state variable, if it takes a non-zero value. 6. Numerical simulation of guided UAV Test numerical simulation of a guidedUAV for newly designedWaleń unman- ned aircraft, which is to be used as an airborne target, was carried out in the Technical Institute of Air Force (Fig.4). According to the diagram shown inFig.1, it has been assumed that aUAV moving along the preset trajectory in the first phase of simulation performs a 264 E. Ładyżyńska-Kozdraś Fig. 4. Reference drawing ofWaleń unmanned aircraft steady flight rectilinear (at the speed V0 = 50m/s at an altitude of 500m), then bypasses the obstacle by climbing U-turn (no slip) with a change of altitudeabout300m,afterwhich it returns to steadyflight onthenewaltitude. The coefficients of amplification appearing in control laws (4.1)-(4.4), took the following values: KHx1 =−0.0029 K H z1 =−0.008 KHU =0.0201 KHW =−1.036 K H Q =−2.3 K H θ =0.3 KVy1 =0.0007 K V V =−0.00054 K V W =0.0231 KVR =1.1 K V ψ =−0.074 KLφ =−0.0009 K L P =0.0007 K L V =0.0011 KLR =−1.5 K L ψ =−3.3 KTx1 =1.06 K T U =−3.004 K T W =0.22 KTθ =4.1 K T Q =−0.06 K T R =0.07 KTψ =−0.003 The results of simulation have been presented in a graphicalway inFigs.5- -Figs.8. The correctness of performing the manoeuvre has been ensured through the coordinated use of deflections and by making a slip angle β impossible to occur. The analysis of graphs shows that the automatic control system ensures themaintenance of the desired flight trajectory, bringing the UAV on the right course ψ. For this purpose, it has proved necessary to increase the speed of flight depending on the angle of roll φ and to increase the angle of attack α. After performing theU-turn, theUAVmaintains the values of flight parameters by returning to steady flight on the new course with the set new height. Modeling and numerical simulation of UAV... 265 Fig. 5. Diagrams of the real and preset UAV flight altitudes (a) and lateral projections (b) Fig. 6. History of the angles of roll, pitch, yaw (a) and the angle of attack (b) Fig. 7. History of the elevator (a) and rudder (b) deflection 266 E. Ładyżyńska-Kozdraś Fig. 8. History of the aileron deflection (a) and the changing engine thrust (b) 7. Conclusions The use of the Boltzmann-Hamel equations for mechanical systems with non- holonomic constraintsmade it possible todevelop themodel of dynamics of the flight of unmanned aircraft. Using the control laws as kinematic correlations of deviations from the preset ideal guidance parameters, the control laws have been linked with dynamic equations of motion of unmanned aircraft vehicle. The obtained simulation results show the correctness of the developedma- thematical model. The flight of the unmanned aircraft takes place in a proper manner. It maintains the preset parameters resulting from the adopted gu- idance method throughout the entire flight. The developedmathematical model and simulation program are universal and can be easily adapted to simulation, guidance and calculations of any unmanned aircraft vehicle (after proper parametric identification of the tested object). A reliable UAV simulation process which can be adapted for different air- crafts would provide a platform for developing autopilot systemswith reduced dependence on expensive field trials. Inmany cases, the testing of newly deve- loped autopilot systems in a virtual environment is the only way to guarantee absolute safety. Additionally, the model would allow better repeatability in testing, with controlled flying environments. References 1. Bloch A.M., 2003,Nonholonomic Mechanics and Control. Systems and Con- trol, Springer, NewYork Modeling and numerical simulation of UAV... 267 2. Blajer W., 1998, Metody dynamiki układów wieloczłonowych, Monografie nr 35,Wydawnictwo Politechniki Radomskiej, Radom 3. BlajerW.,Graffstein J.,KrawczykM., 2001,Predictionof the dynamic characteristics and control of aircraft in prescribed trajectory flight, Journal of Theoretical and Applied Mechanics, 39, 1, 79-103 4. Chelaru T., Pana V., Chelaru A., 2009, Dynamics and flight control of the UAV formations, Journal WSEAS Transactions on Systems and Control, 4, 4 5. Dogan A., Venkataramanan S., 2005, Nonlinear control for reconfigura- tion of unmanned-aerial-vehicle formation, Journal of Guidance, Control, and Dynamics, 28, 4 6. Ducard G.J.J., 2009, Fault-Tolerant Flight Control and Guidance Systems: Practical Methods for Small UnmannedAerial Vehicles, Advances in Industrial Control Series, Springer 7. Etkin B., Reid L., 1996, Dynamics of Flight. Stability and Control, John Wiley & Sons Inc., NewYork 8. Graffstein J., 2009,Wpływ parametrycznej niepewności modelu na zmiany współczynników wzmocnień automatycznej stabilizacji samolotu, Prace Insty- tutu Lotnictwa, 201, 65-75 9. Graffstein J., Krawczyk M., Maryniak J., 1997,Ogólnymodel dynami- ki automatycznie sterowanego samolotu bezpilotowego „ĆMA”, Materiały IV Konferencji „Układy mechaniczne – teoria i zastosowania”, Łódź, 211-216 10. Gutowski R., 1971,Mechanika analityczna, PWN,Warszawa 11. Jordan T.L., Foster J.V., et al., 2006, AirSTAR, A UAV Platform for Flight Dynamics and Control System, NASA Langley Research Center, Report Number: AIAAPaper 2006-3307: 8 12. Koruba Z., Ładyżyńska-Kozdraś E., 2010 The dynamic model of combat target homing system of the unmanned aerial vehicle, Journal of Theoretical and Applied Mechanics, 48, 3, 551-566 13. Ładyżyńska-Kozdraś E., 2008, Analiza dynamiki przestrzennego ruchu ra- kiety sterowanej automatycznie, [In:] Mechanika w Lotnictwie ML-XIII 2008, J. Maryniak (Edit.), PTMTS,Warszawa 14. Ładyżyńska-KozdraśE., 2009, The control laws having a form of kinematic relations betweendeviations in the automatic control of a flying object, Journal of Theoretical and Applied Mechanics, 47, 2, 363-381 15. Ładyżyńska-Kozdraś E., 2011, Modelowanie i symulacja numeryczna ru- chomych obiektów mechanicznych skrępowanych więzami nieholonomicznymi w postaci praw sterowania, Prace naukowe:Mechanika, z.237,OficynaWydawni- cza PolitechnikiWarszawskiej,Warszawa 268 E. Ładyżyńska-Kozdraś 16. Maryniak J., 2005, Dynamika lotu, [In:] Mechanika techniczna, Tom II – Dynamika układówmechanicznych, J. Nizioł (Edit.), KomitetMechaniki PAN, IPPTPAN,Warszawa, 363-472 17. Nejmark J.I., Fufajew N.A., 1971,Dynamika układów nieholonomicznych, PaństwoweWydawnictwoNaukowe,Warszawa 18. Ou Q., Chen X.C., Park D., Marburg A., Pinchin J., 2008, Integra- ted flight dynamics modelling for unmanned aerial vehicles,Proceedings of the Fourth IEEE/ASME International Conference on Mechatronic and Embedded Systems andApplications (MESA08), ISBN: 978-1-4244-2368-2,Beijing,China, 570-575 19. Rachman E., Razali R., 2011, A mathematical modeling for design and development of control laws for unmanned aerial vehicle (UAV), International Journal of Applied Science and Technology, 1, 4 20. Sadraey M., Colgren R., 2005, UAV flight simulation: credibility of line- ar decoupled vs. nonlinear coupled equations of motion, AIAA Modeling and Simulation Technologies Conference and Exhibit, San Francisco, California 21. Shim D.H., Kim H.J., Sastry S., 2003, A flight control system for aerial robots: algorithms and experiments, IFAC Control Engineering Practice 22. Ye Z., Bhattacharya P., Mohamadia H., Majlesein H., Ye Y., 2006, Equational dynamicmodeling and adaptive control of UAV,Proceedings of the 2006 IEEE/SMC International Conference on System of Systems Engineering, Los Angeles, CA, USA, 339-343 Modelowanie i symulacja numeryczna automatycznie sterowanego bezzałogowego statku powietrznego skrępowanego więzami nieholonomicznymi Streszczenie W pracy zaprezentowano modelowanie dynamiki lotu automatycznie sterowane- go bezzałogowego statku powietrznego z zastosowaniem równańBoltzmanna-Hamela dla układówmechanicznych o więzach nieholonomicznych. Prawa sterowania potrak- towano jako więzy nieholonomiczne nałożone na dynamiczne równania ruchu BSP. Uzyskano model matematyczny zawierający sprzężenie dynamiki statku powietrzne- go z nałożonym sterowaniem, wprowadzając związki kinematyczne jako parametry zadane ruchuwynikające z procesu naprowadzania.Poprawność opracowanegomode- lu matematycznego potwierdziła przeprowadzona symulacja numeryczna. Manuscript received December 28, 2010; accepted for print June 14, 2011