| Magnetic levitation ball system is nonlinear and unstable system, thecharacteristics of the system proposed higher requirements for the controlalgorithm. More over, the real-time of the system present high requirement,too.The contradiction between the real-time requirements and system requirementsof the algorithm makes the magnetic levitation ball system to be a goodexperimental platform for the test of control theory algorithm.In this paper, For the magnetic levitation ball system produced byShenZhen-Google-Technology, internal model control strategy is used in theunstable system.Firstly, for the unstable magnetic levitation ball system, using two steps todesign IMC-PID controller, the first step use a feedback PD controller tostabilize the unstable system, then for the generalize stable object, using theinternal model control theory to design IMC-PID controller. Research confirmsthat the method can achieve better control effect compared to the traditional PIDcontrol strategy.Secondly, for the parameters adjustment of IMC-PID controller, anadaptive tuning algorithm for the filter parameters is give based on the artificialtuning rules for parameter, theoretical studies shows that this method can obtaina more ideal control effect.Thirdly, Based on IMC-PID control algorithm, in order to get better controlperformance. The single neuron network is used to adjust the filter parameter ofIMC-PID controller, for the unstable magnetic levitation ball system, firstly, usea PD controller to stable the system, then use the single neuron network to adjustthe filter parameter of IMC-PID controller online, theoretical studies shows thatcompared with the the IMC-PID control strategy the given strategy can achievebetter performance, real-time control experiments also demonstrate thefeasibility and validity of this method.Finally, summarize was made for the designed control algorithm includingadvantages and disadvantages, which was perpared for the continue study that need to do. |
PDF Full Text Request