Model based development of Adaptive Cruise Control (ACC) with CarMaker
Driving on the motorway with heavy traffic is very stressful because the driver has to adapt his own speed to the flow of traffic and to the vehicles running ahead. To assist the driver with this task the automotive industry has developed Adaptive Cruise Control (ACC). This comfort and security system extends the functionality of a conventional speed control with the ability to keep a constant distance with the vehicle ahead. A sensor detects the area in front of the vehicle and analyses the traffic situation. The ACC system adapts the speed to the actual traffic situation by accelerating and braking independently.
A test environment for the model based development of such a system must not only contain detailed models of the test vehicle but also simulate the traffic around the vehicle. For this purpose CarMaker offers an excellent solution: From offline simulations through to hardware-in-the-loop tests, CarMaker can be used for the development of control systems with object detection.
The simulation environment
CarMaker comprises of a complete vehicle environment for virtual test driving. This environment contains models for the test vehicle with all sub-components like drive train, chassis, tires and brakes as well as sophisticated models for the driver and road. The obstacles feature integrated in CarMaker generates the traffic and manages static obstacles (e.g. guard rails, signs) as well as moving objects (vehicles driving ahead or coming from the opposite direction) in the virtual vehicle environment. The obstacles feature furthermore provides data for the 3D animation with IPGMovie.
In the simplest case, the generated objects are represented by cuboids whose dimensions can be parameterized. It is however possible to extend the configuration to have other geometric bodies. The color of the objects displayed with IPGMovie can be freely chosen. The movements of the objects are defined independently to the other objects. Position, speed and acceleration of the objects can be set as a function of time or path. The user has the possibility to build up all imaginable traffic scenarios including objects with arbitrary dimensions via a special menu in the GUI of CarMaker.
The simulation environment meets all the requests to test control systems with object detection under realistic conditions. Depending on the stage of development, a model (for model and software in the loop) or a real ECU is integrated in the vehicle environment for the virtual driving tests.
Due to clearly defined interfaces the integration can be performed easily and quickly. The following figure shows the integration of an ACC software model within Simulink.
The obstacles feature provides the sensor model with information about position, width and length of the generated objects. On the Simulink level, the ACC controller model is integrated between the simulation blocks for the driver and vehicle, and can thereby access the signals for accelerator and brake. In the example presented in this context, a simple sensor model is used that is sufficient for the test of the basic ACC functions. In this implementation, the area inside a defined angle of beam spread and distance starting from a reference point at the vehicle is scanned for objects. When detecting an object inside, it is defined as relevant for the distance control of the ACC model. In this case the sensor model calculates the distance and the speed relative to the vehicle under test and passes these values to the ACC controller model. The vehicle is controlled by braking and accelerating so that it keeps the distance as defined by the user. If there is no relevant object in the sensor area, the speed control system of the ACC model is active and the vehicle is adapted to the reference speed.
Virtual test driving with Adaptive Cruise Control
The virtual test drives show if all objects are detected correctly and the ACC control works without problems. For this purpose the virtual vehicle performs different maneuvers in user defined traffic situations. The following video shows the scenario of a typical motorway situation.
- Click on the image to play or download the movie (1:46 Minutes; 20 MB) - To get an optimal quality please save the movie
The test vehicle approaches a vehicle driving ahead with lower speed from behind. It decelerates in order not to transgress the defined distance. Afterwards it follows the vehicle driving ahead by keeping the reference distance.
At a special point in time the virtual vehicle changes into the left lane and accelerates up to the reference speed. It overtakes two vehicles, then another vehicle changes into the lane in front of the test vehicle. The test vehicle decelerates until it keeps the defined distance. The vehicle in front changes back into the right lane and the test vehicle accelerates up to the reference speed.
Two modes of animation are possible to visualize the virtual test drives of the vehicle equipped with ACC. IPGMovie provides a 3D representation of the driving situation as it is shown in the video. Another animation tool shows the sensor area and the objects present. Relevant and non-relevant objects are labeled with different colors.
The ACC system can be handled via buttons of the Instruments GUI of CarMaker during the virtual test drives. The ACC system can thus be activated and deactivated and the reference speed increased and decreased. Another possibility to control the reference speed and the reference distance of the ACC system is provided by the user interface ACC Control. Both GUIs, the ACC Control and the Instruments, as well as the presentation of the sensor area can be customized.
Summary
Due to the integrated obstacles feature, CarMaker is a practical tool for the model based development of control systems with object detection. ECUs and controller models can be easily integrated into the simulation environment and tested in different traffic scenarios. The animation with IPGMovie reflects exactly the virtual test drives. This makes it easy to validate the driving behavior of the virtual vehicle equipped with sensors and controllers. Thanks to the open structure, the simulation environment of CarMaker can be customized for user specific tasks which results in an individual solution for sophisticated MIL, SIL and HIL projects.
