The clothoid, also known as the Cornu spiral, is a curve generated by linearly increasing or decreasing curvature as a function of arc length. The clothoid has been widely accepted as a logical curve for transitioning from straight segments to circle arcs in roads and railways, because a vehicle following the curve at constant speed will have a constant change of centripetal acceleration. Clothoids have also been widely adopted in planning potential paths for autonomous vehicle navigation. They have been viewed as useful representations of possible trajectories that are dynamically feasible. Surprisingly, the assumptions that underlie this choice appear to be lightly treated or ignored in past literature. This thesis will examine three key assumptions that are implicitly made when assuming that a vehicle will follow a clothoid path. The first assumption is that the vehicle's steering mechanism will produce a linear change in turning radius for a constant rate input. This assumption is loosely referred to as the "bicycle model" and it relates directly to the kinematic parameters of the steering mechanism. The second assumption is that the steering actuator can provide a constant steering velocity. In other words, the actuator controlling the steering motion can instantaneously change from one rate to another. The third assumption is that the vehicle is traveling at a constant velocity. By definition, the clothoid is a perfect representation of a vehicle traveling at constant velocity with a constant rate of change in steering curvature. The goal of this research was to examine the accuracy of these assumptions for a typical Ackermann-steered ground vehicle. Both theoretical and experimental results are presented. The vehicle that was used as an example in this study was a modified Club Car Pioneer XRT 1500. This Ackermann-steered vehicle was modified for autonomous navigation and was one of Virginia Tech's entries in the DARPA 2005 Grand Challenge. As in typical operation, path planning was conducted using the classic clothoid curve model. The vehicle was then commanded to drive a selected path, but with variations in speed and steering rate that are inherent to the real system. The validity of the three assumptions discussed above were examined by comparing the actual vehicle response to the planned clothoid. This study determined that the actual paths driven by the vehicle were generally a close match to the originally planned theoretical clothoid path. In this study, the actual kinematics of the Ackermann vehicle steering system had only a small effect on the driven path. This indicates that the bicycle model is a reasonable simplification, at least for the case studied. The assumption of constant velocity actuation of the steering system also proved to be reasonably accurate. The greatest deviation from the planned clothoid path resulted from the nonlinear velocity of the vehicle along the path, especially when accelerating from a stop. Nevertheless, the clothoid path plan generally seems to be a good representation of actual vehicle motion, especially when the planned path is updated frequently. / Master of Science

<p>Accurate measurements from sensors measuring the vehicle's lateral behavior are vital in todays vehicle dynamic control systems such as the Electronic Stability Program (ESP). This thesis concerns accurate plausibilisation of two of these sensors, namely the yaw rate sensor and the lateral acceleration sensor. The estimation is based on Kalman filtering and culminates in the use of a 2 degree-of-freedom nonlinear two-track model describing the vehicle lateral dynamics. The unknown and time-varying cornering stiffnesses are adapted while the unknown yaw moment of inertia is estimated. The Kalman filter transforms the measured signals into a sequence of residuals that are then investigated with the aid of various change detection methods such as the CuSum algorithm. An investigation into the area of adaptive thresholding has also been made.</p><p>The change detection methods investigated successfully detects faults in both the yaw rate and the lateral acceleration sensor. It it also shown that adaptive thresholding can be used to improve the diagnosis system. All of the results have been evaluated on-line in a prototype vehicle with real-time fault injection.</p>

Accurate measurements from sensors measuring the vehicle's lateral behavior are vital in todays vehicle dynamic control systems such as the Electronic Stability Program (ESP). This thesis concerns accurate plausibilisation of two of these sensors, namely the yaw rate sensor and the lateral acceleration sensor. The estimation is based on Kalman filtering and culminates in the use of a 2 degree-of-freedom nonlinear two-track model describing the vehicle lateral dynamics. The unknown and time-varying cornering stiffnesses are adapted while the unknown yaw moment of inertia is estimated. The Kalman filter transforms the measured signals into a sequence of residuals that are then investigated with the aid of various change detection methods such as the CuSum algorithm. An investigation into the area of adaptive thresholding has also been made. The change detection methods investigated successfully detects faults in both the yaw rate and the lateral acceleration sensor. It it also shown that adaptive thresholding can be used to improve the diagnosis system. All of the results have been evaluated on-line in a prototype vehicle with real-time fault injection.

Mapping stationary objects and tracking moving targets are essential for many autonomous functions in vehicles. In order to compute the map and track estimates, sensor measurements from radar, laser and camera are used together with the standard proprioceptive sensors present in a car. By fusing information from different types of sensors, the accuracy and robustness of the estimates can be increased. Different types of maps are discussed and compared in the thesis. In particular, road maps make use of the fact that roads are highly structured, which allows relatively simple and powerful models to be employed. It is shown how the information of the lane markings, obtained by a front looking camera, can be fused with inertial measurement of the vehicle motion and radar measurements of vehicles ahead to compute a more accurate and robust road geometry estimate. Further, it is shown how radar measurements of stationary targets can be used to estimate the road edges, modeled as polynomials and tracked as extended targets. Recent advances in the field of multiple target tracking lead to the use of finite set statistics (FISST) in a set theoretic approach, where the targets and the measurements are treated as random finite sets (RFS). The first order moment of a RFS is called probability hypothesis density (PHD), and it is propagated in time with a PHD filter. In this thesis, the PHD filter is applied to radar data for constructing a parsimonious representation of the map of the stationary objects around the vehicle. Two original contributions, which exploit the inherent structure in the map, are proposed. A data clustering algorithm is suggested to structure the description of the prior and considerably improving the update in the PHD filter. Improvements in the merging step further simplify the map representation. When it comes to tracking moving targets, the focus of this thesis is on extended targets, i.e., targets which potentially may give rise to more than one measurement per time step. An implementation of the PHD filter, which was proposed to handle data obtained from extended targets, is presented. An approximation is proposed in order to limit the number of hypotheses. Further, a framework to track the size and shape of a target is introduced. The method is based on measurement generating points on the surface of the target, which are modeled by an RFS. Finally, an efficient and novel Bayesian method is proposed for approximating the tire radii of a vehicle based on particle filters and the marginalization concept. This is done under the assumption that a change in the tire radius is caused by a change in tire pressure, thus obtaining an indirect tire pressure monitoring system. The approaches presented in this thesis have all been evaluated on real data from both freeways and rural roads in Sweden. / SEFS -- IVSS / VR - ETT

We consider first the single-track bicycle model and state relations between the tires’ lateral forces and the turning radius. From the tire model, a relation between the lateral forces and slip angles is obtained. The extra rolling resistance forces from cornering are by linear approximation obtained as a function of the slip angles. The bicycle model is validated against the Magic-formula tire model from Adams. The bicycle model is then applied on an optimization problem, where the optimal velocity for a track for some given test cases is determined such that the energy loss is as small as possible. Results are presented for how much fuel it is possible to save by driving with optimal velocity compared to fix average velocity. The optimization problem is applied to a specific laden truck. / Vi betraktar först den enspåriga cykelmodellen och ställer upp samband mellan däckens sidokrafter och kurvradien. Genom däcksmodellen fås ett samband för hur sidokrafterna beror av slipvinklarna. De extra rullmotståndskrafterna för kurvor fås via linjär approximation som funktion av slipvinklarna. Cykelmodellen valideras mot en däcksmodell från Adams. Cykelmodellen tillämpas sedan på ett optimeringsproblem där den optimala hastigheten längs en bana för några givna testfall bestäms så att energiförlusten blir så liten som möjligt. Resultat presenteras för hur mycket bränsle det är möjligt att spara genom att köra med optimal hastighet jämfört med fix medelhastighet. Optimeringsproblemet tillämpas på en specifik lastad lastbil.

The purpose of this master thesis is to study the energy efficiency of a vehicle when it is cornering. To achieve this, a Simulink model was built from a simple basic bicycle model and theoretically validated. This model was then analysed and successively improved by adding velocity and yaw moment control. A study of the vehicle model behaviour by changing parameters such as cornering stiffness and centre of gravity position was the nconducted. The traction force needed for a constant radius was calculated and methods such as torque vectoring have been tested using the model to obtain the lowest traction force. The model was compared with different vehicle types and further validated by comparing the simulation results with experimental data acquired from a field test. The rolling resistance and aerodynamic resistance were taken into account when the model was validated with the experimental data and the result suggest that by distributing the required traction force (using torque vectoring between inner and outer driven wheels) the energy efficiency could be improved by 10%. This report ends with recommendations for future work.

Autonomous vehicles have in recent years grownin popularity. An autonomous car has the potential to safelymaneuver in an efficient manner. This in combination with thefocus on increased road safety has put higher emphasis onimplementing an overtaking controller. Model Predictive Control(MPC) is very useful because it can handle linear constraintsand works for autonomous driving. I implemented the controlsystem in Python and did tests on its overtake capability usingdifferent velocities, car distances and initial speeds. Constraintswere implemented so that the autonomous vehicle did not collidewith another vehicle or drive outside the road when overtaking.The results show that a safe overtake could be performed undercertain conditions. The MPC algorithm is proven useful butdifficult to optimize. / Autonoma fordon har lyckats locka till sig mer populäritet under de senaste åren. En autonom bil har möjligheten att manövrera på ett säkert och effektivt sätt. Detta i kombination med ett fokus att öka vägsäkerheten har lagt större press på att implementera reglersystem för omkörningar. Modell prediktiv reglering (MPC) är användbar för den kan hantera linjära bivillkor och fungerar till autonomon körning. Ett reglersystem är implementerat i Python och testades på sin omkörningförmåga med olika hastigheter, avstånd och begynnelse hastigheter. Implementationen utformades med bivillkor som att det autonoma fordonet inte ska krocka med ett annat fordon eller köra utanför vägen i en omkörning. Resultaten visar att det gick att köra om på ett säkert sätt med vissa förutsättningar. MPC algoritmen har visat sig användbar men svår att optimera. / Kandidatexjobb i elektroteknik 2020, KTH, Stockholm

This thesis reports the identification of linear and nonlinear handling models for road vehicles starting from structural identifiability analysis, continuing with the experiments to acquire data on a vehicle equipped with a sensor set and data acquisition system and ending with the estimation of parameters using the collected data. The 2 degrees of freedom (dof) linear model structure originates from the well known linear bicycle model that is frequently used in handling analysis of road vehicles. Physical parameters of the bicycle model structure are selected as the unknown parameter set that is to be identified. Global identifiability of the model structure is analysed, in detail, and concluded according to various available sensor sets. Physical parameters of the bicycle model structure are estimated using prediction error estimation method. Genetic algorithms are used in the optimization phase of the identification algorithm to overcome the difficulty in the selection of initial values for parameter estimates. Validation analysis of the identified model is also presented. Identified model is shown to track the system response successfully. Following the linear model identification, identification of 3 dof nonlinear models are studied. Local identifiability analysis is done and optimal input is designed using the same procedure for linear model structure identification. Practical identifiability analysis is performed using Fisher Information Matrix. Physical parameters are estimated using the data from simulated experiments. High accuracy estimates are obtained. Methodology for nonlinear handling model identification is presented.

Autonomous vehicles have gained much importance over the last decade owing to their promising capabilities like improvement in overall traffic flow, reduction in pollution and elimination of human errors. However, when it comes to long-distance transportation or working in complex isolated environments like mines, various factors such as safety, fuel efficiency, transportation cost, robustness, and accuracy become very critical. This thesis, developed at the Connected and Autonomous Systems department of Scania AB in association with KTH, focuses on addressing the issues related to fuel efficiency, robustness and accuracy of an autonomous heavy-duty truck used for mining applications. First, in order to improve the state prediction capabilities of the simulation model, a comparative analysis of two dynamic bicycle models was performed. The first model used the empirical PAC2002 Magic Formula (MF) tyre model to generate the tyre forces, and the latter used a piece-wise Linear approximation of the former. On top of that, in order to account for the nonlinearities and time delays in the lateral direction, the steering dynamic equations were empirically derived and cascaded to the vehicle model. The fidelity of these models was tested against real experimental logs, and the best vehicle model was selected by striking a balance between accuracy and computational efficiency. The Dynamic bicycle model with piece-wise Linear approximation of tyre forces proved to tick-all-the-boxes by providing accurate state predictions within the acceptable error range and handling lateral accelerations up to 4 m/s2. Also, this model proved to be six times more computationally efficient than the industry-standard PAC2002 tyre model. Furthermore, in order to ensure smooth and accurate driving, several Model Predictive Control (MPC) formulations were tested on clothoid-based Single Lane Change (SLC), Double Lane Change (DLC) and Truncated Slalom trajectories with added disturbances in the initial position, heading and velocities. A linear time-varying Spatial error MPC is proposed, which provides a link between spatial-domain and time-domain analysis. This proposed controller proved to be a perfect balance between fuel efficiency which was achieved by minimising braking and acceleration sequences and offset-free tracking along with ensuring that the truck reached its destination within the stipulated time irrespective of the added disturbances. Lastly, a comparative analysis between various Prediction-Simulation model pairs was made, and the best pair was selected in terms of its robustness to parameter changes, simplicity, computational efficiency and accuracy. / Under det senaste årtiondet har utveckling av autonoma fordon blivit allt viktigare på grund av de stora möjligheterna till förbättringar av trafikflöden, minskade utsläpp av föroreningar och eliminering av mänskliga fel. När det gäller långdistanstransporter eller komplexa isolerade miljöer så som gruvor blir faktorer som bränsleeffektivitet, transportkostnad, robusthet och noggrannhet mycket viktiga. Detta examensarbete utvecklat vid avdelningen Connected and Autonomous Systems på Scania i samarbete med KTH fokuserar på frågor gällande bränsleeffektivitet, robusthet och exakthet hos en autonom tung lastbil i gruvmiljö. För att förbättra simuleringsmodellens tillståndsprediktioner, genomfördes en jämförande analys av två dynamiska fordonsmodeller. Den första modellen använde den empiriska däckmodellen PAC2002 Magic Formula (MF) för att approximera däckkrafterna, och den andra använde en stegvis linjär approximation av samma däckmodell. För att ta hänsyn till ickelinjäriteter och laterala tidsfördröjningar inkluderades empiriskt identifierade styrdynamiksekvationer i fordonsmodellen. Modellerna verifierades mot verkliga mätdata från fordon. Den bästa fordonsmodellen valdes genom att hitta en balans mellan noggrannhet och beräkningseffektivitet. Den Dynamiska fordonsmodellen med stegvis linjär approximation av däckkrafter visade goda resultat genom att ge noggranna tillståndsprediktioner inom det acceptabla felområdet och hantera sidoacceleration upp till 4 m/s2 . Den här modellen visade sig också vara sex gånger effektivare än PAC2002-däckmodellen. v För att säkerställa mjuk och korrekt körning testades flera MPC varianter på klotoidbaserade trajektorier av filbyte SLC, dubbelt filbyte DLC och slalom. Störningar i position, riktining och hastighet lades till startpositionen. En MPC med straff på rumslig avvikelse föreslås, vilket ger en länk mellan rumsdomän och tidsdomän. Den föreslagna regleringen visade sig vara en perfekt balans mellan bränsleeffektivitet, genom att minimering av broms- och accelerationssekvenser, och felminimering samtidigt som lastbilen nådde sin destination inom den föreskrivna tiden oberoende av de extra störningarna. Slutligen gjordes en jämförande analys mellan olika kombinationer av simulerings- och prediktionsmodell och den bästa kombinationen valdes med avseende på dess robusthet mot parameterändringar, enkelhet, beräkningseffektivitet och noggrannhet.

The goal of this thesis work is to evaluate and quantify the advantages and disadvantages of Active Rear Steering (ARS). The evaluation is carried out through Multi-Body System (MBS) simulations. An analytical model has been developed to better understand the basic dynamics of vehicles equipped with rear steering. In parallel, a high fidelity MBS model is developed in Simpack. This model includes suspension kinematics and compliance, allowing for detailed analyses of steering hardware performance. Next, different control strategies aiming at improving manoeuvrability, stability and agility are implemented in Simulink. In order to assess their effectiveness, the high fidelity model is utilised by running co-simulation with Simulink. Manoeuvrability is assessed through constant steer, constant radius and ramp steer manoeuvres. Stability is assessed through transient manoeuvres such as step steer and sine with dwell. Agility is assessed through step steer and frequency response. Ultimately, also a subjective assessment is carried out by means of Volvo Cars' dynamic driving simulator. The conclusion from the assessment is that the drivers feel the all wheel steered vehicle more stable during evasive manoeuvres. It is concluded that for manoeuvrability the minimum turning radius is reduced by 19 % at low velocity; this implies that the steering angle request is reduced at low velocity, while it is increased at high velocity. A slightly higher steering angle request at high velocity might be beneficial since the driver would be able to control the vehicle in a wider range of steering wheel angles. For agility the results are contradicting: on the one hand, according to the step steer rise time difference between lateral acceleration and yaw rate, the controlled vehicles are performing worse than the passive vehicle; on the other hand, according to the frequency response analysis, both the delays between steering input and yaw rate and between lateral acceleration and yaw rate are reduced up to respectively 75 % and 46 % for the considered frequency range. Finally, for stability, the yaw rate overshoot from a step steer can be reduced up to 65 % at high velocity and the sideslip angle can always be reduced. The vehicle equipped with ARS outperforms the passive vehicle in the sine with dwell manoeuvre. / Målet med detta examensarbete är att utvärdera och kvantifiera fördelarna och nackdelarna med Active Rear Steer (ARS) för Volvo Cars. Utvärderingen utförs genom Multi-Body System (MBS) simuleringar. En analytisk modell har utvecklats för att bättre förstå den grundläggande dynamiken i fordon utrustade med bakhjulsstyrning. Parallelt utvecklades en MBS-modell med hög precision i Simpack. Denna modell inkluderar hjulupphängningens kinematik och eftergivlighet, vilket möjliggör detaljerade analyser av styrhårdvarans prestanda. Därefter implementeras olika kontrollstrategier som syftar till att förbättra manövrerbarhet, stabilitet och agilitet i Simulink. För att bedöma deras effektivitet används MBS-modellen för att köra co-simulering med Simulink.Manövrerbarhet bedöms genom konstant styrning, konstant radie och rampstyrning. Stabilitet bedöms genom transienta manövrar som stegstyrning och sinus med fördröjning. Agilitet bedöms genom stegstyrning och frekvensrespons. Slutligen görs också en subjektiv bedömning med hjälp av Volvo Cars dynamiska körsimulator. Slutsatsen från bedömningen är att förarna anser att fordonet upplevs vara mycket stabilare vid undanmanövrar. Vidare är slutsatsen att för manövrerbarhet minskar den minsta svängradien med 19 % vid mycket låg hastighet; detta innebär att styrvinkel reduceras vid låg hastighet, medan den ökar vid hög hastighet. En något högre styrvinkeln kan vara fördelaktig eftersom föraren skulle kunna styra fordonet i ett större rattvinkelområde. För agilitet är resultaten motsägelsefulla: å ena sidan, enligt stegstyrningstidsskillnaden mellan lateral acceleration och girhastighet, fungerar de aktiva fordonen sämre än det passiva fordonet; å andra sidan, enligt frekvensresponsanalysen, reduceras både fördröjningarna mellan girhastighet och styrvinkel och mellan lateral acceleration och girhastighet upp till ungefär 30 %. Slutligen, för stabilitet, kan girhastighetens översläng från en stegstyrning minskas upp till 65 % vid hög hastighet och avdriftsvinkeln kan alltid minskas. Fordonet som är utrustat med ARS överträffar det passiva fordonet i manövern sinus med fördröjning.