Hybrid renewable energy system for moving platforms

Renewable energy system can harvest energy from sustainable renewable resources and supply clean energy to the power system. Hybrid renewable energy system (HRES) that use multiple resources together can potentially provide long-term power supply to moving platforms. Major challenges in the research are estimate power generation of HRES that installed on the moving platform, optimise HRES configuration under size constraints, and manage the use of power to maximise the realibity and performance of the system.

Spatial-temporal power generation simulation

Compare to conventioal systems, the HRES power generation simulation need to be done in a spatial-temporal manner on a moving platform. The moving platfrom as the base of HRES undergoes a global and a local motion while operating. The wind power simulation result shows how the global and the local motion affect the wind power generation.

Use the 'Box Select’ tool, you may able to find where and when the wind power genertaion reach the peak. The wind power generation is determined by both time and location of the platform. This research, for the first time, established a way to estimate the power generation of HRES on a moving platform subject to both global and local motion.

Data driven constrained optimisation

  • Data-driven approach is used to collect and prepare renewable resources data for the simulation. Weather were first subset from real weather database and then reconstruct a spatial-temporal data.

  • Non-stationary HRES model is developed for an accurate power output estimation. As an example, when platform undergoes a local motion, solar angle changes constantly that affect its power output. Novel non-stationary HRES model takes the dynamics of platforms into consideration.

Solar angle

  • Constraints in the design is handle by bio-inspired multi-objectives optimisation. It supports multi-level method (PSO), simultaneous method (GA) and non-linear (NLOPT) method.

Flow chart

Featured with capabilities such as intelligent data handle, nonlinear non-concave optimise, and automatic report generation, software used in this research are open sourced to the public.

Learning based power management

Under construction