Regional Electrical Power System Ultra-short-term Load Forecasting And Network Modeling Analysis

As socialist economy develops further,electric power demand in China is increasing.The transmission of electric power distinguishes strongly from that of other industries.Power energy can not be abundantly stocked,thus the electric power generation,power supply,power sale and power consumption must be processed in the meantime.Power network,as a wholesale system,should retain stable and reliable in electric power generation and management.Therefore,completes the regional power system planning,regardless in the economic or in the technology has a very important significance.In power system planning,the single issue's solution is the basis for whole optimization planning.However,the unified plan has many difficulties and shortcomings,such as the Unified Modeling difficulties,solving difficulties,often omitted some important factors.But,if follow a scientific approach to individual planning,we can simplify the overall power grid planning.Not only avoids neglecting some important factors,but also ensure that the entire power grid of the economy, reliability and safety performance.In this dissertation,two important issues was chose as the main contents,namely ultra short-term power load forecasting and modeling of regional electric power network.These two issues are the two key needed to resolve on the regional power system planning.The ultra-short-term load forecasting is the foundation of the electricity price.While the regional power network model for analysis and planning, then relates the regional power system's stable operation and the development. Therefore,this dissertation use complex systems theory,intelligent control theory and complex network theory to resolve the two key issues.And the reasonable proposal for the construction the region electrical power system was put forwardThe power load forecast is the basis of the power system planning.With the actual planning of regional development,uses a scientific and effective load forecasting methods forecast plans the local load electric quantity,is an important work to the regional power system planning.This dissertation uses the BP algorithm, proposed a double hidden layer structure artificial neural network to the forecast ultra-short-term power system load.After the simulation with the actual data,the method which this dissertation proposed compared to the traditional linear extrapolation Prediction method average absolute percent error of 2.5 percentage points,ten minute load forecasts achieve the average absolute percentage error are smaller than 2.5%accuracy.The electric power network planning usually is in the load demand basic clear that the future load growth within a period of time scale known on the basis of the load forecasting,proposed the electric power network planning plan,will meet the load need and the guarantee electrical power system safe operation need.The electric power network planning is a multi-objective,multi-stage,including a large number of discrete factors of uncertainty,bound by the non-linear integrated optimization problem.Considered the above characteristics and difficulties,with the purpose of the study in this dissertation,the decomposition-coordinated planning model and power system source planning model were firstly established.At the same time,the region electric power network model using the W_S microcosm network model has also been established.The model was applied to analyze the regional power system when the accident occurred,different scale emergency precautionary measures under the power transmission system disaster and recovery trend.The ideas and main contents of this dissertation are:(1) First is the analytical study background,reviewed the power load forecasting,digital filtering and the electric power network planning research development process,put this dissertation on the issue;(2) From the power system at the composition of load,load forecasting the characteristics of the main factors and basic principles of electricity,and other aspects of in-depth analysis of the complex characteristics of load forecast;(3) To improve the accuracy of load forecast,based on the optimal design of a Hankel norm approximate linear phase IIR filter used to load forecast data pretreatment;(4) Proposed a modified ultra-short-term load forecasting methods based on artificial neural network and realized the power system load forecast for the modeling,analysis and forecast,the simulation results show that this method of prediction accuracy higher,with better Application of value;(5) Analysis of the complex power system of large-scale systems,application of complex systems theory of regional power system modeling analysis, realizes the electrical power system power source plan using the genetic algorithm;(6) Analyzes electrical power system's complex network characteristic,the complex network theory was applied to establish a regional electricity power network model,a W-S small world network model was used on the power system disaster recovery modeling and analysis;(7) Summarizes this dissertation main research results and the conclusion,and in the fore-mentioned work's foundation,makes the forecast to the region electrical power system plan's further research work.The major innovation of this dissertation in the following areas:(1) As the multi-layer neural network approach to the nonlinear function and the ability to learn of BP algorithm,early neural network forecast model mostly uses the single hidden layer forward network,hidden layer neurons Activation function selects non-linear Sigmoid function,the network training uses the BP algorithm.According to the regional power system load forecast of ultra-short-term characteristics,this dissertation proposed a double hidden layer network architecture.After the simulation with the actual data,the method which this dissertation proposed compared to the traditional linear extrapolation Prediction method average absolute percent error of 2.5 percentage points,ten minute load forecasts achieve the average absolute percentage error are smaller than 2.5%accuracy.(2) For the filter load data,this dissertation proposed an method that is based on optimal Hankel norm approximation of linear phase infinite impulse response(IIR) filter design.Through the use of linear phase finite impulse response(FIR) filters and optimal Hankel norm similar approach to design linear phase IIR filter.Not only this method reduced in the inverse matrix solution process computation load,simultaneously has given the norm approximation error L_x boundary.The simulation result has confirmed this method validity.(3) This dissertation applies complex systems theory and complex networks theory to analysis the regional electric power network,establishes the load actual region electrical power system network model,faces the electric power supply and demand question and the network risk in view of the current our country electrical power system,proposes the corresponding suggestion solution.Through the study,in order to satisfy the growing demand for power load and improve power system reliability,builders should take some effective measures so that the whole power system in the following direction:(1) Increasing voltage.Gradually increase the voltage distribution networks,simplifying voltage division,reducing voltage levels,is conducive to the management and distribution system of economic operations.At present,the domestic network of urban distribution system voltage levels are mostly 10kv,with the load of urban growth,distribution networks need to transfer more power.To improve the transmission capacity of transmission lines through the following two ways:line voltage levels unchanged,increasing current,increasing the line voltage.And improve voltage current economic often than increased.Therefore,in power system planning,increase voltage is better.(2) Simplifies the electrical network structure,uses the new facility,establishes the accident emergency measure,improve the degree of automation.The electric power network is developing to further simplify and higher reliability.Such as wiring Substation Transformer Group promote the use of lines, such as single-bus connection.Electricity network structure to promote the use of multi-line,all kinds of ring network,more and more connected,and other sub-ways to improve the utilization and reliability,the installation of automated devices to reduce the fault recovery time.Power System Planning has a lot of uncertainty and unknown factors.It is difficulty to describe using mathematical models.In many cases can only rely on the experience of the planning staff to complete its work.Therefore utilizes scientific,the advanced theory method and the research results in the electrical network plan,not only may reduce the electric power to plan personnel's work load greatly,but also has guaranteed the plan accuracy. At present the overall power grid optimization plan,or substation and transmission network expansion and other individual plans,the planning of economic research more fully.However,due to the accident continued emergence of a safe and stable operation of power grids more and more attention to the need to strengthen the plan reliability,security considerations.To large-scale electrical network and long-term dynamic programming,but also has the computation speed and the convergence question.In the more popular electrical network flexibility plan research aspect,the uncertainty factor which considered was still incomplete,many uncertainty factors are neglected or the processing method is improper.China's power market reform has been launched and has made certain achievements,the future electricity market environment of power system planning study work has just begun,yet in-depth,the direction of future research should also include:(1) Strengthen the new algorithm research,,seeks for fast,the highly effective practical solution method.(2) Considers reasonably the multi-stage planning in various stages of transition between the plan and mutual restraint,to find a more reasonable and practical model to solve.(3) Considers more reasonably in the coordinated efficiency and the reliable relations,considers the safety factor fully.(4) More comprehensive and reasonable consideration and handle the various uncertain factors,such as economic,environmental and policy so that the plan has a higher flexibility and adaptability.(5) Consider and determine a reasonable grid planning models and algorithms,to meet the needs of the electricity market model.