Research On Indoor Positioning Method Based On Multi-sensor Information Fusion

DC/DC converter is generally used in the traditional industry,electronic information and other industries,has great application potential.At present,in the field of DC/DC converters,researchers can usually be classified into two categories: one is committed to designing appropriate control strategies to provide reliable power for the system and guarantee the stability of the system;the other is focused on the development of various topologies to improve the efficiency of the converter.This paper takes Buck converter as the research object,mainly by comparing various control algorithms,selects RBF neural network algorithm to transform the control strategy of traditional sliding mode controller,and designs a sliding mode controller based on neural network,aiming at improving the performance of the converter.The main research contents of this paper include the following aspects:First of all,the switching power supply were studied,including the historical background,current situation of the development of switch power supply as well as the basic theoretical knowledge,discusses the type Buck step-down DC/DC converter of several kinds of modeling method,and using the state space method,established the mathematical model of target converter,discusses several classical control strategy,and,on a good mathematical model has been established in accordance with the design idea of the sliding mode variable structure control,sliding mode variable structure controller is completed to meet the requirements,and to reflect the performance of the sliding mode variable structure controller is also set up the experimental comparison of traditional PID controller.After the design of the sliding mode variable structure controller is completed,the reliability of the system is confirmed by using the Lyapunov energy function,and the rationality of the design is verified.Finally,a series of simulation experiments are carried out on the traditional sliding mode variable structure controller designed by using the MATLAB simulation software.Then,according to the completion of the sliding mode variable structure controller design,aiming at the existing chattering problem of sliding mode controller,the load disturbance and voltage fluctuation influence,through in-depth study of neural network algorithm,select RBF neural network algorithm,the retrofit design of the traditional sliding mode variable structure controller,neural network approximation item design and the adaptive law is designed,built,based on the neural network sliding mode control of DC/DC converter simulation experiment platform,use energy lyapunov function of the system verified the feasibility and stability analysis.Finally,the simulation design is carried out in the Matlab environment,and the simulation comparison experiment is completed to analyze the control effect and robustness of the neural network sliding mode controller to the switching converter.Finally,in order to verify the performance of the neural network sliding mode controller is designed,using STM32 control chip build machine system experimental platform finish transducer measurement experiment,at the same time and the traditional PID controller and traditional sliding mode variable structure controller complete contrast experiment,the experimental results show that the design of the neural network sliding mode controller can effectively weaken chattering controller,at the same time also has the advantages of resistance to load disturbance and reduce interference system,the sliding mode control based on neural network of the transducer shows good output performance.