Job-Scheduling for automated Car Parking Systems : A Machine Learning Approach
Abstract: The ever growing amount of cars and their requirement for parking space has led to the development of highly sophisticated public automated car parking systems. The user acceptance criteria for such systems is the car drop-off and retrieval time. In this thesis, a genetic algorithm is developed, that tries to minimize the drop-off and retrieval times compared to established “First-In-First-Out” scheduling techniques. The algorithm is a so called crossbreeding algorithm, that combines machine allocation and job order execution into one chromosome encoding. Allocation and order are evolved by using genetic operators separately. A variety of different operators are tested in a Monte Carlo type simulation and the results are compared to the benchmark algorithm using scheduling strategy as currently in use. On average, the genetic algorithm can improve the job scheduling by 14% for a reasonable job queue.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)