The Multi-objective Model And Algorithm Of Generation Expansion Planning Including Wind Farms

As one of the most clean and renewable energy, the wind generation is becoming more andmore prevalent due to the contradiction between the rapid growth in energy demand and theincreasing depletion of fossil fuel resources. However, the wind power is with natures ofintermittent, stochastic and anti-peak-shaving, which would affect the system reliability andpeak-shaving ability when including it in the power system. This dissertation adopts NSGA-Ⅱalgorithm to solve the generation expansion planning model through the power systemprobabilistic production simulation based on the available capacity distribution, and the windabandonment capacity and peak-shaving capacity are also considered.This paper mainly contributes to the following aspects:(1) Due to the shortage of long simulation time during the traditional probabilisticproduction simulation based on the available capacity distribution, a method of looking-up tableis proposed to cut down the simulation time. Also, the dissertation provides the calculationmethods of wind abandonment capacity and peak-shaving insufficient capacity in wind systems.(2) In order to review the investment and operation cost and the outage cost simultaneously,a multi-objective generation expansion planning model is proposed. And the methods of VMPand AHP are utilized to solve this multi model. The VMP method can on the one hand decreasesthe dimensions of decision variables, and on the other hand increases the diversity of thepopulation. Meanwhile, the AHP method makes it very conveniently to choose the optimalsolution in the Pareto optimal curve.(3) The paper also analyzes the impact of wind farms in thermal power units generationexpansion planning. Through the comparison of including wind farms or not, the paper analyzesthe impact of wind farms for wind abandonment capacity and peak-shaving insufficient capacity,and it shows that: when the judgment net-load is less than the minimum output, the windabandonment capacity appears, and the size of it is related to the total wind farms output and themaximum absorbing wind energy; and when the peak-shaving capacity is less than the load increase, the peak-shaving insufficient capacity appears, the size of it is related to the differencebetween them.The innovation of this paper is mainly reflected in the following three points:(1) This paper proposes a method of looking-up table firstly, which decreases the simulationtime during the probabilistic production simulation based on the available capacity distribution.(2) This dissertation firstly provides a calculation method for wind abandonment capacityand peak-shaving insufficient capacity in wind systems.(3) For generation expansion planning, the traditional methods are usually single-objectiveoptimization methods, occasionally, multi-objective methods are adopted, but thesemulti-objective methods never have wind farms included, and this is the first time to solvegeneration expansion planning with wind farms included by using the multi-objective method.