A Research On The Optimization Of The Electronic Triaxial Handheld Stabilizer Control System Based On BP-PID

With the popularity of photography,followed by a series of problems encountered in the process of filming.One of the main issues of photographic images in the process of shaking.Up to now,people are used to enhance the stability of the equipment is not much,there are mainly the tripod and Stan Nikon and so on.Developed in recent years,electronic triaxial handheld stabilizer is the research object of this article,it is a kind of photography in the process of enhance the stability of the equipment,but the current stabilizer response speed slow,lack of stability is still the main problem.Based on the electronic handheld stabilizer,the deficiency of the three axis is proposed using Artificial Colony Algorithm(Artificial Bee Colony Algorithm,ABC)and nanophosphor must Search(Beetle Antennae Search Algorithm,BAS)improved BP(Back Propagation)neural network,and the improvement of BP neural network used in the choice of the parameters of the PID control system,namely BP-PID control system,so as to realize the online parameter setting and adaptive,improve electronic triaxial handheld stabilizer parameters self-tuning PID controller and adaptive ability,in order to improve the rapidity and stability of the system.This paper main research work is as follows:(1)In this paper with reaction speed is slow,lack of stability of electronic handheld stabilizer PID control system,three axes are put forward using the BP neural network used in the selection of PID control parameters,in order to realize the parameter self-tuning of PID controller.First of all,according to the requirements of project design reasonable electronic triaxial video stabilizer PID control system model,and expounds the BP neural network PID control,in view of the BP neural network dynamic optimization performance,slow convergence speed and learning time long shortcomings,error function of BP neural network and activation function,make the neural network can get rid of the flat area,avoid falling into local optimum.The simulation results show that the improved BP-PID control system has better dynamic performance.(2)BP-in order to further improve the dynamic performance of PID control system,using artificial colony algorithm and longicorn must search these two kinds of heuristic algorithm,weights of neural network PID controller for optimization.The longicorn must search algorithm as a new algorithm,there is no person in the optimization of neural network PID control system are studied,the particular form of the algorithm does not need to function,no gradient information,its strong ability to jump out of local optimum,sawyer,must be a single individual convergence speed is fast,made in training neural network only do binary classification problems,can reduce the complexity of the network,reduce network coupling.Tested BP-PID control system based on ABC optimization and the optimization of BP-BAS PID control system,and compares the two BP-without tuning and step response of PID control system.Simulation results show that BAS training BPPID control system with ABC training algorithm of BP-PID control system have their own advantages and disadvantages,prove sawyer,must search for BP-the effectiveness of the proposed tuning PID control system.In this paper,this article uses MATLAB language to realize the two kinds of BAS and ABC algorithm,then the BAS was measured and the optimization process of ABC,the BAS-BP-step response of PID and the step response of ABC-BP-PID,again through the MATLAB simulation software to analyze it.The results show that the BAS-BP-the PID operation efficiency is higher than the ABC-BP-PID.