Using coverage path planning methods for car park exploration

Abstract

With the increasing number of vehicles on the roads, finding a free parking space has become a time-consuming problem. Traditional car parks are not equipped with occupancy sensors, so planning a systematic traversal of a car park can ease and shorten the search. Since car park exploration is similar to coverage path planning (CPP) problems, the core concepts of CPP algorithms can be used. This paper presents a method that divides maps into smaller cells using trapezoidal cell decomposition and then plans the traversal using wavefront algorithm core concepts. This method can be used for multi-storey car parks by planning the traversal of each floor separately and then the path from one floor to the next. Several alternative explorational paths can be generated by taking different personal preferences into account, such as the length of the driven route and the proximity to preferred locations. The planned traversals are compared by step number, the cell visitedness ratio, the number of visits to each cell and the cost function. The comparison of the methods is based on simulation results.