| Distributed optimization is an effective task to achieve optimization through cooperation and coordination between individuals.It can be used to solve large-scale and complex optimization problems that many centralized algorithms can’t handle.It has become one of the hot issues in large-scale complex networks.On the other hand,to some extent,event-based control strategy,that is,the control is updated only when a certain control amount exceeds a given threshold value(depending on the system status or time),can reduce the waste of unnecessary computing resources and communication resources.And it attracts great attention from scholars in the theoretical and engineering fields.Taking the advantages of event-based control and distributed optimization,the event-based distributed optimization is considered in a class of continuous systems.The details are as follows:1.Basic knowledge of algorithm design and demonstration methods are discussed for distributed optimization of discrete systems and continuous systems,respectively.Through comparison,it is found that although the systems are not the same in form,there are essential similarities in algorithm design and demonstration methods.The detailed analysis and comparison of perturbation analysis theories lays a foundation for the further study of event-based distributed optimization theory.2.For a class of continuous systems,we first design the trigger function that depends on the step-rate attenuation and the corresponding trigger condition determines the sampling time points.Then we design the distributed optimization algorithm using the state information at the sampling time point,and use the disturbance analysis theory to demonstrate the feasibility of the algorithm,which can achieve the same optimization goal as designing a distributed optimization algorithm using continuous state information with minimal computational resources and communication resources.Finally,the validity of the result was verified by simulation. |
PDF Full Text Request