Roman Schmied

Cooperative adaptive cruise control applying stochastic linear model predictive control strategies

In this paper a cooperative adaptive cruise control approach using stochastic, linear model predictive control strategies is presented. The presented approach deals with an urban traffic environment where vehicle to vehicle and vehicle to infrastructure communication systems are available. The goal is the minimization of a vehicles fuel consumption in a vehicle-following scenario. This is achieved by minimizing a piecewise linear approximation of the vehicles fuel consumption map. By means of a conditional Gaussian model the probability distribution of the upcoming velocity of the preceding vehicle is estimated based on current measurements and upcoming traffic light signals. The predicted distribution function of the predecessors velocity is used in two ways for stochastic model predictive control. On the one hand, individual chance constraints are introduced and subsequently reformulated to obtain an equivalent deterministic model predictive control problem. On the other hand, samples are drawn from the prediction model and used for a randomized optimization approach. Finally, the two developed stochastic control strategies are evaluated and compared against a deterministic model predictive control approach by means of a virtual traffic simulation. The evaluation of the controllers show a significant reduction of the fuel consumption compared to the predecessor while increasing safety and driving comfort.

Extension and experimental validation of fuel efficient predictive adaptive cruise control

Advanced driver assistant systems (ADAS) like adaptive cruise control (ACC) are primarily developed to increase safety and driving comfort and nowadays applied to upper class production vehicles. Additional benefits like improvement of fuel economy is a widespread field in research. In this paper a fuel efficient predictive adaptive cruise control (PACC) approach is performed experimentally with a test vehicle. To this end, a model to predict the predecessors prospective velocity is introduced which allows anticipatory driving. An online model predictive controller (MPC) calculates the desired acceleration of the following vehicle such that fuel consumption is minimized while keeping constraints to the inter-vehicle distance as well as minimum and maximum vehicle speed and acceleration. Experimental results on a road and in HIL tests show a significant benefit in fuel economy as well as in reduction of NOx and particulate matter emissions of the controlled vehicle compared to its predecessor.

Predictive Cooperative Adaptive Cruise Control: Fuel Consumption Benefits and Implementability

Impressive improvements of efficiency and safety of vehicles have been achieved over the last decade, but increasing traffic density and drivers’ age accentuate the need of further improvements. The contributions summarized in this chapter argue that a substantial additional fuel benefit can be achieved by extending the well introduced Adaptive Cruise Control in a predictive sense, e.g. taking into account a predicted behavior of other traffic components. This chapter starts by discussing results on the potential benefits in the ideal case (full information, no limits on computing power) and then examines how much of the potential benefits is retained if approximate solutions are used to cope with a realistic situation, with limited information and computing power. Two setups are considered: vehicles exchanging a small set of simple data over a V2V link and the case of mixed traffic, in which some vehicles will not provide any information, but the information must be obtained by a probabilistic estimator. The outcome of these considerations is that the approach is able to provide—statistically—a substantial fuel consumption benefit without affecting negatively the driveability or the driver comfort like other methods, e.g. platooning, would.

Flexible Spacing Adaptive Cruise Control Using Stochastic Model Predictive Control

This paper proposes a stochastic model predictive control (MPC) approach to optimize the fuel consumption in a vehicle following context. The practical solution of that problem requires solving a constrained moving horizon optimal control problem using a short-term prediction of the preceding vehicle’s velocity. In a deterministic framework, the prediction errors lead to constraint violations and to harsh control reactions. Instead, the suggested method considers errors, and limits the probability of a constraint violation. A conditional linear Gauss model is developed and trained with real measurements to estimate the probability distribution of the future velocity of the preceding vehicle. The prediction model is used to evaluate two different stochastic MPC approaches. On the one hand, an MPC with individual chance constraints is applied. On the other hand, samples are drawn from the conditional Gaussian model and used for a scenario-based optimization approach. Finally, both developed control strategies are evaluated and compared against a standard deterministic MPC. The evaluation of the controllers shows a significant reduction of the fuel consumption compared with standard adaptive cruise control algorithms.

Scenario model predictive control for robust adaptive cruise control in multi-vehicle traffic situations

Considering multi-lane and multi-vehicle scenarios common adaptive cruise control (ACC) systems often face the problem of sudden and uncomfortable control actions when surrounding vehicles change the lane leading to a switch in the target vehicle of the ACC. Probabilistic modeling of the lane change behavior of surrounding traffic participants allows to predict such lane changes. This enables anticipatory control actions to avoid hard braking maneuvers and hence increases driving comfort and economy. This paper presents a scenario model predictive control (SCMPC) which estimates the lane change tendency of surrounding drivers by drawing a number of scenarios from a stochastic lane change prediction model. The model itself is identified based on real driving data. Simulation results show the advantages of the proposed control strategy by means of comparison to a common PI controlled ACC system.

A robustified Newton based extremum seeking for engine optimization

Extremum seeking (ES) is a well known approach for online optimization of control parameters, e.g. in engine control. While the basic idea of ES is rather straightforward, in practice its application suffers from the problems related to determine the optimum numerically using measurements corrupted by noise. In addition, nonlinearities of the system under scrutiny, e.g. engines, can lead to a non convex objective function and thus to numerical problems. The purpose of this paper is to introduce a simply implementable extension to Newton based methods to improve the robustness of the convergence under real world conditions and to test it on a production Diesel engine. The extension is based on the regularization idea, and does not introduce significant additional tuning and setup effort. The results clearly show the improvements with respect to standard gradient and Newton based ES algorithms. The key advantage of this method is to provide convergence properties independently from the operating point and without re-tuning.

Improving the transient emission performance of a Diesel engine by input shaping techniques

Caused by the demanding and more and more stringent legislated emission limits for passenger car engines, the effect of transient emission peaks becomes more and more important for the overall emission limits. In case of Diesel engines, the two main concerned pollutants are NOx and PM and typically the compliance with the legislation is achieved by a suitable control of the fuel and air system of the engine. Especially during transients the coordination of both loops is crucial for the overall performance. Typically the control is separated in the two loops, where coupling effects and different time scales and dynamics can lead to undesired overshoots during transients. Against this background in this work an input shaping technique to reduce the transient emission peaks is proposed. The input shaping is based on an identified response model of the transient emission profile and used to determine suitable correction trajectories which can be applied to an existing calibration and reduce undesired effects during transients while allowing to keep the base calibration and control structure of the engine control unit. In this study input shaping is applied for the rail pressure and injection parameters to reduce the NOx emissions while the PM emissions, efficiency and noise levels should be kept similar to the nominal operation. The proposed strategy is evaluated on a 2l turbocharged common rail passenger car Diesel engine mounted at a dynamic engine testbench and promising results are achieved.

A Virtual Development and Evaluation Framework for ADAS—Case Study of a P-ACC in a Connected Environment

Advanced driver assistance systems (ADAS) or even (partially) automated driving functions (ADF) can lead to substantial improvements in fuel economy, safety, and comfort of passenger cars. Especially, in view of new technologies, such as connected vehicles, additional improvements are feasible. However, testing and validation of ADAS in a connected and interacting environment are a critical and not yet fully solved task. In real-world driving situations in a dense urban traffic environment, constant interactions between the system under test (SUT) and other traffic participants occur. The number of possible scenarios and test cases is huge and renders a case by case approach, even for function prototyping and performance evaluation, almost impossible. In this work, a virtual development framework is proposed which allows performance testing under realistic traffic conditions by taking the interaction between SUT and other participants into account. A combination of a microscopic traffic simulation and a high-detailed vehicle simulation is utilized. To handle the interaction between both tools, a co-simulation framework with an interface layer for synchronization is developed which serves also as input for virtual sensors and prototype functions. The framework is demonstrated by a case study for a predictive adaptive cruise control (P-ACC) in a connected environment. This case study shows both the potential benefits of utilizing available information via new communication channels for ADAS and the applicability of the proposed framework.

A minimum entropy based switching approach for automated driving assistance

This paper suggests a switching control approach for the application of automated driving assistance. Since in multi-lane and multi-vehicle scenarios the traffic situation may change frequently, the use of switched systems theory seems suitable to deal with this problem. A two level control approach is suggested to deal with lane changing and longitudinal vehicle control. In the first step the target lane is chosen according to a robust switched equilibrium and the second step deals with stabilizing this equilibrium by switching between different longitudinal vehicle control strategies. Simulation results show the feasibility and effectiveness of the approach by means of an exemplary test scenario.

Simulating the effect of ultrasonic ranging errors on parallel parking performance

This paper investigates the effect of sensor errors on the parallel parking performance of an intelligent park assist system. A path planning algorithm has been employed for generating a trajectory that maneuvers the car into the empty space in one trial. Different control schemes have been implemented for lateral steering control of the vehicle to track the trajectory generated by the planning algorithm. The control schemes include different combinations of PD and non-linear control. All simulations are carried out within the IPG CarMaker® and Simulink® environments. Positioning and sensor errors are modeled in the simulation environment and are applied to an ideal ultrasonic sensor module from CarMaker. Multiple parallel park simulations are run and the effect of these errors on the parking performance is shown with respect to the different control schemes employed.