Intelligent Scheduling Of Battery Energy Storage Station In A Micro-grid

In modern society, with the rapid economic development and increasingly seriousenergy problems, distributed energy like wind or solar has great foreground in powersystem. In order to realize large-scale use of distributed power, the concept ofmicro-grid appeared. Energy storage station, as one part of the micro-grid, has played animportant role on stable output and economic operation of micro-grid. In order to findoptimal scheduling for energy storage station, research work about this problem iscarried out in the dissertation.The dissertation models for each micro-source in the micro-grid, and summarizesthe research status of output power forecasting technique of wind and photovoltaicsystem, which provides theoretical basis for energy storage station scheduling. Since theenergy storage station has the function of improving power output fluctuations andregulating system economic benefit caused by time-of-use pricing, the optimizationmodel of battery energy storage station is proposed to stabilize the power output and tomaximize profit of the micro-grid under proper assumptions, and necessary constraintsare added according to requirements for the remaining power and ratedcharge-discharge power of battery energy storage unit.When solving the optimization problem with dynamic programming method, it isfound that dynamic programming methods can only deal with discrete state of batteryenergy storage unit, which makes the resulting scheduling policy of energy storage unitnot accurate, and conflict occurs between accuracy of the strategy with hugecomputation. Based on these two shortcomings of dynamic programming method forsolving the optimization problem, convex optimization method is adopted to getsolution for the optimization model after some transformations for the objective andrelated constraints. Take a specified case as example, and the simulation result showsthe comparison diagram on power output with and without the use of energy storagestation, which proves superiority of using convex optimization.