Research On Key Technologies Of Power Network Planning Considering Wind Generation

The intermittent power sources represented by wind power have gained wide attention and fast development due to their advantages on environment protection and sustainability.With their increasing penetration in power system,new demands to planning and operation of power network emerge.Considering the effects of intermittent power sources in power network planning research is crucial to reliability and economy of power system.In this dissertation,the power network planning method considering intermittent power sources was researched from three respects:planning models,solving algorithms,and evaluation methods considering probabilistic production simulation,probabilistic load flow calculation and scenario set generation.Details are as follows:1.Multi-objective immune optimization algorithm is effective in solving power network planning problems,but its efficiency is constrained by premature elimination of new antigen and lack of pertinence design.In order to improve the effectiveness of solutions to these specific problems,an improved multi-objective immune optimization algorithm for power network planning is proposed.The concept of survival circle is adopted and different operations to antibodies were carried out based on the grade of survival circle.So the new antigen can be moderately protected and the diversity of antigens is enhanced.Search space of the algorithm can be expanded so as to get more complete Pareto frontier.Furthermore,sensitivity related indicator is proposed considering the efficiency of cost variation to relieve the maximum load rate of transmission lines,by which heuristic local search process and feasible adjustments are guided.The Pareto frontier is maintained through an outer memory set in the process of optimization.A more complete Pareto frontier result can be got with low computational cost by the proposed algorithm.2.In probabilistic load flow(PLF)calculation,there is contradiction between computational cost and ability dealing with complex factors.To evaluate the probability distribution of active power flow in the transmission lines,active power flow of transmission lines in DC model is expressed and calculated in three parts according to different influence factors.With the wind speed which obeys correlated Weibull distribution,a decomposed expression of origin moment is adopted to deal with the wind power related items,so the factors correlated with wind power and transmission network can be calculated independently.Efficiency of the PLF method for calculations with large amount of different networks is improved.Failures of wind generators,transmission lines,and traditional generators are considered in the PLF as well as correlation among wind speeds and loads.The PLF method can achieve fast and accurate calculations,and it is applicable for power network planning with intermittent power sources.3.Multi-scenarios method is an effective method to deal with the uncertainties in power network planning.Consideration of the low probability scenarios which have severe impacts on power system is important for enhancement of the planning schemes' reliability.There still exists defectiveness in the way to generate these extreme scenarios,such as inaccurate recognition and insufficient quantity.Therefore,a method for recognition of extreme scenarios based on structural vulnerability index and state vulnerability index is proposed,and the extreme scenario set is then generated from sample selection with Pareto optimal theory.With this method,recognition of extreme scenarios is more accurate in systems integrated with high penetration intermittent power sources.4.Because the distribution of system power supply ability is only determined by the peak load level,the conventional composite power system effective load duration curve(CMELDC)method has problem in accuracy for probabilistic production simulation.In order to improve its applicability in power network planning,an improved CMELDC method is introduced in this paper.The effect of variation of load points,load level on distribution of system power supply ability is considered in this method.The variation of load level is taken into account through clustering based on load proportions.The concept of correlation factor is adopted to reflect the correlation between load level and load proportions,and the maximum arrival power values of load points are calculated by the clustering center thought sensitivity method.With this method,accuracy of reliability evaluation is improved and the analytical methods' advantage on calculation speed is preserved.It is suitable for reliability evaluation in power network planning.5.The frequency and duration(FD)method for probabilistic production simulation evaluates the dynamic cost in system operation through frequency information of fluctuation of net load.Its result can be inaccurate because of unfeasible calculation of the start-up/shut-down operations of generators in short time.To express the effect of such fluctuations on the system operation cost more accurately,the concept of interval frequency distribution is proposed to consider the frequency information and time distribution information of load fluctuation simultaneously.An analytical probabilistic production simulation method based on equivalent interval frequency distribution is proposed.This method is more accurate than traditional FD method for evaluation,and it is more suitable for cost evaluation in power network planning.6.The schemes of power network planning should have adaptability to the growth of load and power sources beyond the preset conditions.Quantitative evaluation and pertinence optimization for this kind of adaptability are rarely considered in the existing power network planning.In this dissertation,the adaptability is evaluated from the angle of redundancy,and the correlating concepts of redundancy optimization model for power network planning are proposed.A method to classify the development models of regional load and power sources are proposed based on different load grow speed.With this method,four different development models of the planning region are formulated.Four specific forms of redundancy optimization model for power network planning are presented according to the characteristics and demands of different development models,which are illustrated as follows:For improving the adaptability and wide applicability of optimization in different development models,a power network planning model considering transmission capacity redundancy is built.With a sub-objective concerning comprehensive optimization of redundancy,this model considers adequacy and proportionality of transmission capacity redundancy synthetically.For the development model with rapid load growth,a power network planning model considering redundancy of load growth is proposed.The adaptability of schemes to load growth is reflected through redundancy of load growth,and the load growth redundancy index which quantitatively evaluates this adaptability is proposed.Proportionality of redundancy of load points and adequacy of the overall system's redundancy are considered for evaluation of the schemes' redundancy condition.For the development model with rapid growth of intermittent power source,a power network planning model considering potential income from wind generation is proposed.Redundancy of ability for acceptance of wind power was reflected from the view of economic income considering the evaluation result of potential income from wind generation.For the development model with rapid growth of load and intermittent power source,a power network planning model based on extreme scenario set and redundancy of system expansion was proposed.Adaptability of schemes to simultaneous increase of load and power source is considered based on redundancy of system expansion,and the system expansion redundancy index which quantitatively evaluates this adaptability is proposed.Redundancy of system expansion,as well as construction cost and the proportionality of system are optimized in the model.The models are in conjunction respectively with the probabilistic load flow method,the improved CMELDC method,the probabilistic production simulation method based on equivalent interval frequency distribution,and the extreme scenario set in this dissertation.In the development models corresponding to these power network planning models,the adaptability of schemes to future growth can be enhanced,the bottlenecks can be eliminated,and the benefits can be improved.