Short-term Photovoltaic Power Prediction And Its Optimization In Power System

Photovoltaic power plant is a large intermittent power affected by weather changes,when it incorporated into the power system,its volatility and randomness will affect the conventional unit output and the system's economic and environmental performance.It is necessary to predict the output power of the PV system that incorporated into the power system.This paper has carried on the power forecast to the photovoltaic power station which has been incorporated into the electric power system,and has carried on the optimization research to the electric power system.Firstly,this paper introduces the principle and the characteristics of PV system.The model and characteristic curves of the photovoltaic cell are given.The power characteristics of the PV system are introduced,and the power characteristic curves based on the sampled data are given.At the same time,the typical operation modes of the PV system are also introduced according to the block diagrams.Secondly,this paper uses the improved back propagation neural network algorithm to predict the solar irradiance,and then combined with the power model to short-term predict the photovoltaic power.For the improved back propagation algorithm,adds the steepness factor to the bipolar Sigmoid function,and adopts adaptive adjustment of learning rate in the network.At the same time,the network datas are normalized and anti-normalized to enhance the generalization ability of the neural network model.In the case study,different solar irradiance prediction methods and different PV power prediction methods are given.The calculated results are compared with the actual values,and the feasibility of the model used in this paper is verified.Thirdly,this paper presents a two-objective optimization model that minimizes the cost of the system and the least of the daily exhaust emissions.The objective functions are the minimum cost of daily integrated power generation and the minimum emission of daily exhaust gas.The cost of integrated power generation include the operating costs of photovoltaic and the thermal power units and the cost of coal savings due to access to photovoltaic power plants,and the emission of exhaust gas refers to the emission of 2CO and 2SO and NOx of the system.This paper uses the bacterial colony chemotaxis algorithm to optimize the system.The bacterial colony chemotaxis algorithm adds the chaos optimization model and the elite retention strategy to improve the convergence performance of the algorithm.Finally,this paper optimizes the power system based on the short-term photovoltaic power prediction results.This paper analyzes the influence of different conditions on the optimization results,and validates the practicability and validity of the two-objective optimization model and the optimization algorithm.At the same time validates that accurately predict the output power of the photovoltaic systems not only can co-ordinate the units power,but also can reasonably absorb the PV resources.