Path planning, modelling and simulation for energy optimised mobile robotics

Student thesis: Doctoral Thesis


This thesis is concerned with an investigation of a solution for mobile robotic platforms to minimize the usage of scarce energy that is available and is not wasted following traditionally planned paths for complex terrain environments. This therefore addresses the need to reduce the total energy cost during a field task or mission. A path planning algorithm is designed by creating a new approach of artificial potential field method that generates a planned path, utilising terrain map. The new approach has the capability of avoiding the local minimum problems which is one of the major problems of traditional potential field method. By solving such problems gives a reliable solution to establish a required path. Therefore the approach results in an energy efficient path of the terrain identified, instead obvious straight line of the terrain. A literature review is conducted which reviews the mainstream path planning algorithms with the applications in mobile robotic platforms was analysed. These path planning algorithms are compared for the purpose of energy optimized planning, which concludes the method of artificial potential field as the path planning algorithm which has the most potential and will be further investigated and improved in this research. The methodology of designing, modelling and simulating a mobile robotic platform is defined and presented for the purpose of energy optimized path planning requirement. The research is to clarify the needs, requirements, and specifications of the design. A complete set of models which include mechanical and electrical modelling, functional concept modelling, modelling of the system are established. Based on these models, an energy optimized path planning algorithm is designed. The modelling of force and the kinematics is established to validate and evaluate the result of the algorithm through simulations. Moreover a simulation environment is established which is constructed for multi perspective simulation. This also enables collaborative simulation using Simulink and ADAMS to for simulating a path generated by the path planning algorithm and assess the energy consumption of the driven and steering mechanism of an exemplar system called AgriRover. This simulation environment allows the capture of simulated result of the total energy consumption, therefore outlines the energy cost behaviour of the AgriRover. A total of two sets of paths was tested in the fields for validation, one being generated by the energy optimized path planning algorithm and the other following a straight path. During the field tests the total cost of energy was captured . Two sets of results are compared with each other and compared with the simulation. The comparison shows a 21.34% of the energy saving by deploying the path generated with the energy optimized path planning algorithm in the field test. This research made the following contribution to knowledge. A comparison and grading of mainstream path planning algorithms from energy optimisation perspective is undertaken using detailed evaluation criteria, including computational power required, extendibility, flexibility and more criteria that is relevant for the energy optimized planning purpose. These algorithms have not been compared from energy optimisation angle before, and the research for energy optimised planning under complex terrain environments have not been investigated. Addressing these knowledge gaps, a methodology of designing, modelling and simulating a mobile platform system is proposed to facilitate an energy optimized path planning. This , leads to a new approach of path planning algorithm that reduces unnecessary energy spend for climbing of the terrain, using the terrain data available. Such a methodology derives several novel methods: Namely, a method for avoidi
Date of Award9 Nov 2021
Original languageEnglish
Awarding Institution
  • University Of Strathclyde
SupervisorXiu Yan (Supervisor) & Mark Post (Supervisor)

Cite this