Reliability Evaluation And Optimization For Power Generating Systems With Limited Size Of Performance Sharing Group

With the fast development of smart grid,electricity market,and renewable resources,power systems have already become one of the largest and the most complex engineered systems where are required to work in a real-time manner.The fundamental task of power systems is to provide safe and reliable service to consumers.As the producer of electrical powers,power generating systems play a vital role in ensuring reliability and operation of power systems.However,the load models in the existing reliability assessment methods for generation systems are usually based on the subjective predictions and experiences of experts or designers.The deviations between the forecasted loads and the real values are,therefore,inevitable.Additionally,it was also assumed that the load models among subsystems are statistically independent of each other in reliability evaluation of interconnected power generating systems.In fact,due to the influence of geographical and meteorological conditions,the load models are oftentimes statistically dependent with one another.This thesis devotes to develop a mathematical tool for reliability evaluation of interconnected power generating systems suffering uncertain and correlated loads.Furthermore,the traditional reliability models of interconnected power generating systems usually suppose that all the power elements could be connected to the common bus for performance sharing.However,it is extremely difficult or even unworthy to achieve such a huge interconnection structure due to the physical and economic restrictions.This thesis further explore the reliability evaluation and optimization method of interconnected power generating systems with limited size of performance sharing group,so as to facilitate further decision makings of power generating systems in terms of design and expansion.The major contributions of this thesis are summarized as follows:(1)Development of a reliability evaluation method for interconnected generating systems with considering the uncertainty of loads.In this thesis,firstly,the reliability model of interconnected generation system is established.To solve the problem of uncertainty of the predictive load model,the normal distribution is introduced.Meanwhile,a loss of load probability method based on the universal generating function is presented,which can evaluate the reliability of generation systems under uncertain loads in a computationally efficient manner.(2)Development of a reliability evaluation method for interconnected generating systems with considering the correlations of loads.The loss of load probability method for uncertain loads is further extended to the case where the loads are statistically correlated.Furthermore,by taking account of the restrictions on geography,economy,environment and technology,the reliability assessment for generating systems with limited interconnected resources is studied,and a new reliability evaluation method for such generating systems is proposed.As observed in our comparative studies for static connection and dynamic connection,the significance of the interconnection structure of the generating systems has been revealed.(3)Development of the reliability optimization model for interconnected power generating systems in the context of the limited size of performance sharing group.Under the limited connected elements,a single-objective and a multi-objective optimization problems are formulated for generating system reliability with either static connection or dynamic connection.The genetic algorithm is used to solve the single-objective problem whereas the non-dominated sorting genetic algorithm is implemented to address the multi-objective problem.As demonstrated in the case studies,the dynamic connection can greatly improve the availability of interconnected power generating systems.