Applications Of Improved Particle Swarm Optimization Algorithm On Parameters Identification Of Photovoltaic Models And Power System Economic Dispatch

With the emerging of the intelligent era,the world has witnessed the climax of the development of artificial intelligence technology.Artificial intelligence technology,as an important driving force,is continuously affecting modern industries.As a significant branch of artificial intelligence,intelligent optimization algorithms have been successfully used in a variety of engineering fields.Based on the particle swarm optimization algorithm,this thesis proposes two improved algorithms to provide effective alternatives for the problem of parameters identification of solar photovoltaic models(modeling as a single objective unconstrained optimization problem)and the problem of economic dispatch of the power system(modeling as a single objective constrained optimization problem).The problem of parameters identification of solar photovoltaic models is to identify the important parameters in the photovoltaic models,so as to obtain accurate photovoltaic models,which is of great significance to the efficient use of solar energy.In fact,this problem can be modeled as a single objective unconstrained optimization problem.For this problem,this thesis proposes a classified perturbation mutation based particle swarm optimization.In the proposed algorithm,the classification perturbation mutation strategy is used to balance the global search ability and local search ability of the population.The damped boundary processing method is used to solve the situation that the basic particle swarm optimization algorithm is easy to fall into local optimal near the boundary when dealing with the problem of parameters identification of photovoltaic models.In order to verify the performance of the proposed algorithm in dealing with the parameters identification of photovoltaic models,experiments are performed on single diode model,double diode model,photovoltaic module model,and photovoltaic module models at different irradiance or temperature conditions.Experimental results show that the proposed algorithm can get better results than other compared algorithms on the problem of parameters identification of photovoltaic models.The above photovoltaic model parameters identification problem is attributed to a single objective unconstrained optimization problem,but the optimization problems in practical engineering applications often have some constraints,such as the power system economic dispatch problem,this is a single objective constraint optimization problem,which refers to the power system can reasonably adjust the output power of the unit under certain constraints to achieve the best economic benefits.Compared with the photovoltaic model parameters identification problem,it adds some constraints on the basis of optimization objective,which makes it more difficult to solve the power system economic dispatch problem.For the economic dispatch problem of power system,this thesis proposes a constrained particle swarm optimization algorithm based on segmentation point.In this algorithm,the generalized opposition-based learning strategy and the improved mutation cross strategy are combined into the basic particle swarm optimization algorithm by setting the segmentation point.At the same time,the feasibility rule is used as the constrained handling technique to compare the solutions.In the experimental study,comprehensive results on 22 test functions from CEC2006 verified the effectiveness of the proposed algorithm in dealing with single-objective constraint optimization problems.In addition,the proposed algorithm is employed to solve the 6 generator units and 15 generator units in the power system.Experimental results show that the proposed algorithm can effectively solve the economic dispatch problem of power system with multiple constraints such as power balance constraints,ramp rate constraints,and prohibited operating zones.