Abstract
Using the bilevel optimization (BIO) scheme, this paper presents a time-optimal path planner for autonomous underwater vehicles (AUVs) operating in grid-based environments with ocean currents. In this scheme, the upper optimization problem is defined as finding a free-collision channel from a starting point to a destination, which consists of connected grids, and the lower optimization problem is defined as finding an energy-optimal path in the channel generated by the upper level algorithm. The proposed scheme is integrated with ant colony algorithm as the upper level and quantum-behaved particle swarm optimization as the lower level and tested to find an energy-optimal path for AUV navigating through an ocean environment in the presence of obstacles. This arrangement prevents discrete state transitions that constrain a vehicle’s motion to a small set of headings and improves efficiency by the usage of evolutionary algorithms. Simulation results show that the proposed BIO scheme has higher computation efficiency with a slightly lower fitness value than sliding wavefront expansion scheme, which is a grid-based path planner with continuous motion directions.
Highlights
Autonomous underwater vehicles (AUVs) are frequently employed to perform environmental monitoring and exploration tasks, such as surveillance of the dynamics of plankton assemblages, temperature, and salinity profiles, and the onset of harmful algal blooms [1,2,3]
The bilevel programming scheme decomposes the task of path planning into two parts: (1) The outer optimization problem or the upper level, which is defined as finding a free-collision channel from a starting point to a destination consisting of connected grids, and (2) the inner optimization problem or the lower level, which is defined as finding the energy-optimal path in the channel generated by the upper level
The bilevel optimization (BIO) scheme is presented to solve the problem of path planning for AUV
Summary
Autonomous underwater vehicles (AUVs) are frequently employed to perform environmental monitoring and exploration tasks, such as surveillance of the dynamics of plankton assemblages, temperature, and salinity profiles, and the onset of harmful algal blooms [1,2,3]. The bilevel programming scheme decomposes the task of path planning into two parts: (1) The outer optimization problem or the upper level, which is defined as finding a free-collision channel from a starting point to a destination consisting of connected grids, and (2) the inner optimization problem or the lower level, which is defined as finding the energy-optimal path in the channel generated by the upper level This scheme uses the ant colony algorithm (ACA) [23] as the upper level algorithm and relies on the quantum-behaved particle swarm optimization (QPSO) [24] at the lower level.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
