Planning and Decision Making algorithm slides, Grafiken und Mindmaps von Computergestützte Statistik

Nur auf Docsity: Lade Planning and Decision Making algorithm slides und mehr Grafiken und Mindmaps als PDF für Computergestützte Statistik herunter!

Planning and Decision-

Making Algorithms

How can a computer make decisions?

Organisational Aspects

Lecturer

 Andreas Hagerer

Professor für technische Informatik und Ingenieurmathematik

 Erreichbarkeit

Email: [email protected] Sprechstunde: Donnerstag 15:00 – 16: Raum: A

Material

 Moodle “Planning and Decision-Making Algorithms”

 Enrolment key Xsd

Automated Decision-Making (ADM)

 Real-world examples

 Reasons to automate decision-making  ...

What is this course about?

Planning Algorithms

 “Planning is the art and practice of thinking before acting.”  a method used in computer science to generate a sequence of actions that lead to the achievement of a goal by systematically exploring possible states and actions in a logical manner  classical problem "Seven Bridges of Königsberg"

What is this course about?

Hierarchy of Planning Tasks

What is this course about?

Topics

• Mission-Planning / Pathfinding
• Manoeuvre Planning
• Motion Planning
• Trajectory Planning

1

Challenges

Mission Planning

Overview of Approaches

Mission Planning

Examples

Manoeuvre Planning

Information Sources

 route/mission planning output: a sequence of information on lanes and associated start and end positions  vehicle attributes: current position, current lane, current relative position in the lane, speed  history: previous decisions in similar/similar situations  information on the vehicle environment: all objects within a certain radius; each object is attributed, e.g. with lane, speed, direction, predicted trajectories (output of a perception and prediction component)  traffic guidance information and traffic sign information  traffic rules currently valid in the given situation

Manoeuvre Planning

Characteristics

 search for a route that must be safe and practicable  determination of the most feasible route  consideration of vehicle dynamics  physical limits of the control system (e.g. motor power, maximum torque)  vehicle internal conditions (e.g. speed, available friction between tires and road, torque transmission in clutches)  maneuverability in the presence of obstacles  traffic regulations and road boundaries

Motion Planning

Characteristics

 search for a route that must be safe and practicable  determination of the most feasible route  consideration of vehicle dynamics  physical limits of the control system (e.g. motor power, maximum torque)  vehicle internal conditions (e.g. speed, available friction between tires and road, torque transmission in clutches)  maneuverability in the presence of obstacles  traffic regulations and road boundaries

Motion Planning

Characteristics

 determination of a sequence of configurations/states that the vehicle assumes, parameterized by time and often also by speed  formalization  trajectory  =  1 ,  2 , … ,   ∈ ℝ ^ state / configuration  motion model +1 =   ,   ∈ ℝ ^ controller inputs  conditions  ≤ +1 ≤   ≤  ≤   cost function 

 task: determine arg min ,  , 

Trajectory Planning