2019 CSCE Annual Conference - Laval (Greater Montreal) Conference
Dr. SangHyeok Han
In heavy industrial projects, modular based construction is getting popular because of its efficiency and the lower cost. The key success factor of modular industrial projects is lift path planning that influences the productivity and safety through the entire process. Many algorithms such as hill climbing, A *, and genetic algorithms are introduced by researchers for mobile crane lift path planning. However, a comprehensive comparison of these algorithms is not done yet to identify the direction of potential research for the mobile crane lift path. This paper compares the algorithms for the lift path planning of mobile crane in the modular-based heavy industrial project to find the competent method to search the collision-free path with the lower operating cost and less computation time. The lift path planning for the mobile crane is currently in progress with two algorithms in this paper: A* search and Genetic algorithm. In the algorithms, the crane configuration (combination of swing, luffing, and hoisting) is defined to move the modules from pick location to the final location. The algorithms are developed in Python and Matplotlib is used for the visualization. The results of these will be compared with two other algorithms that already developed by Alberta University. The case study is based on the industrial modular project by PCL industrial Management in Alberta, Canada, that includes a considerable number of module liftings by mobile crane (Demag CC 2800). The sample modules to analyze are selected by its difficulty of work (levels of congested site areas) which are classified by three levels based on the scheduled sequence, obstacles, and the installation level. The algorithms’ results are compared by criteria such as computation time, the number of movements, success rate, the linear travelling distance of module, and crane operation time to determine the effectiveness for the path planning of mobile crane in the heavy industrial projects. This comparison will show us which algorithm is more effective for the crane path planning in heavy industrial projects and suggest the direction of further research.