Research On Closed-loop Continuous Active Control For Beam String Structure

The current research and innovation of spatial structure focuses on the optimization of the structure and the pursuit of lighter and more efficient structure.The rise of the field of artificial intelligence has also promoted the development of intelligent structure.The unique characteristics of beam string structure ensure its good prospect in the field of structural static control.In this paper,the beam string structure is taken as the research object,and the realization of closed-loop continuous active control is considered as the research target.From the research contents of control algorithm innovation,single-objective control sample research,multi-objective control sample research,closed-loop continuous control and so on,the active beam string structure is further studied.Firstly,the stochastic search algorithm which is widely used in the field of structural active control is introduced.Combined with the characteristics of discrete optimization problems and the advantages of genetic algorithm,on the basis of gradient genetic algorithm,discrete crossover operator and discrete mutation operator are introduced to make the genetic operator search in discrete space.At the same time,the discrete gradient operator is proposed to improve the local optimization ability of the genetic algorithm and realize the fast discrete optimization of the algorithmSecondly,taking the active beam string structure as an example,the structural control problem is transformed into a mathematical optimization problem.Taking the minimum displacement of the structure and the minimum stress of the main beam as the optimization objective,the displacement and internal force control model of the active beam string structure is established.And taking the discrete actuator control quantity as the optimization variable,the discrete sample generation mechanism based on the improved genetic algorithm is proposed.Then considering the control requirements of various objectives,the multi-objective optimization problem of active beam string structure in discrete domain is studied.The multi-objective fuzzy optimization model of the beam string structure based on discrete variable is established,and it is solved by a two-step fuzzy optimization algorithm.A variety of control options are provided and multi-objective control strategy that meet the expectations of decision makers is developedFinally,the method based on BP neural network to realize continuous prediction of the active beam string structure is proposed.According to the control requirements of different objectives,the stress state information is selected,and it is combined with the discrete actuator control quantity as the sample input quantity of the neural network to establish a one-to-one corresponding relationship with the optimal control quantity.An improved neural network is formed which can be applied to the prediction of continuous structural control under the specified control objective.