Research On Load Frequency Control Based On Adaptive Echo State Network

The research of load frequency control is related to the national economy and people’s life.With the continuous upgrading of power system structure,the control object is becoming more and more complex and nonlinear,the model is complex,and it is difficult to achieve accurate modeling.The classical control theory is difficult to support the modern load frequency control problem.Due to the rapid development of artificial neural network,it has been successfully applied to load frequency control system.However,there are still some problems in the application of artificial neural network,such as large amount of calculation and large amount of training.Due to the unique advantages of"fast calculation","high-precision training results" and "randomly generated connection of neurons",echo state network is introduced into the load frequency control system..In view of the fact that the disturbance often occurs in the power system,the traditional control parameters are difficult to obtain,the amount of calculation is large,and it is difficult to adapt to the experience,this thesis proposes a load frequency control based on adaptive echo state network,designs the controller,and deeply analyzes and discusses the stability of the system.And,considering the bandwidth constraints under network control,this thesis adds event triggering mechanism in the system to save communication resources.The main contents of this thesis include:The second chapter of the thesis introduces some basic knowledge needed in this thesis.Firstly,the basic principles of power generation,speed regulation,load,tie line and auxiliary control are introduced in detail,and the corresponding transfer function is given.Then,the state space expression of the load frequency control system is introduced,which lays a theoretical foundation for the controller design and related proof.Secondly,the basic structure and model knowledge of echo state network are given.Finally,some lemmas used in this thesis are listed to provide mathematical technical support for the subsequent proof.In the third chapter of this thesis,aiming at the problem that the traditional PID controller parameters are too large in calculation and too dependent on human experience,which makes it difficult to adjust adaptively,this thesis proposes the combination of echo state network and load frequency control for the first time,and studies the load frequency control problem based on adaptive echo state network.Firstly,according to the practical significance and control requirements of load frequency control,the echo state network model is established.Then,the echo characteristics of the echo state network are studied,and the stability conditions of the network are derived.Secondly,the adaptive load frequency controller is designed by using echo state network.According to the input and output of the load frequency control system,the training process of the network is given.Finally,the simulation model is built by Simulink,and the effectiveness of the method is verified by comparing the results and relevant data.In the fourth chapter,the load frequency control system with limited network communication bandwidth is studied.A large amount of information transmission with limited bandwidth will bring burden to the power system,waste resources to a certain extent,and even bring economic problems to the load frequency control system.Therefore,this thesis studies the load frequency control problem based on event triggered adaptive echo state network.Firstly,the echo state network is introduced into the load frequency control,and the echo state network model is established.At the same time,the output feedback controller of the load frequency control system based on event triggered adaptive echo state network is designed.The online training steps of the network under the controller are given.The stability of the load frequency control system is analyzed,and the conditions for the bounded stability of load frequency control system are calculated,the Zeno phenomenon is proved.Finally,the effectiveness of this method is verified by the simulation of two area interconnected system in MATLAB.In the fifth chapter,the whole thesis is summarized.Firstly,this thesis proposes a load frequency control based on adaptive echo state network,analyzes it,gives the specific steps and the network training process under this method.Then the load frequency control of event triggered adaptive echo state network considering communication bandwidth is proposed.The threshold design,controller design and stability analysis are carried out to eliminate the Zeno phenomenon.Finally,the further prospect of the selection of ESN parameters and the setting of trigger threshold is given.