Font Size: a A A

Research On The Driving And Control System Of Ultrasonic Motor

Posted on:2015-04-16Degree:MasterType:Thesis
Country:ChinaCandidate:Z L HuangFull Text:PDF
GTID:2272330422491107Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
As a new kind of motor, there is a big different between traditionalelectromagnetic motor and ultrasonic motor. When it comes to control, we requirethis control algorithm has the function that can control the object which has strongnonlinearity. In addition, it should adapt to the changing environment. BecauseUSM is a kind of capacitive load, and works under Ultrasonic frequency, so when itcomes to drive, it requires the driving power to output some driving signals whichhave different phases, same frequency and same amplitude. Meanwhile, thefrequency is greater than20kHz and the output power must be greater than100Wand output signal is ultrasonic frequency.In this design process, we analysis the nonlinear factors of ultrasonic motorwhich using bending vibration transducer. Analyze the working characteristics bybuilding the contact model of ultrasonic motor. And come to the conclusion that ithas strong nonlinearity between stator and rotor. Then build the mold ofpiezoelectric ceramics when it works in free states. By this physical mold, analysisthe nonlinearity of motor when it works under different frequencies and differentamplitudes driving signals. What s more, analyze the heat producing process whenit works, and the influence of the temperature to the piezoelectric ceramics. Fromthe equations, we can see the rise of temperature is also the one cause of thenonlinearity of motor.In addition, based on the strong nonlinearity of ultrasonic motor, we use thefuzzy neural network method to control. Because the quantity of study samples istoo big, to make the value of data as small as possible, we analyze a new clusteringalgorithm which has the less complete graphs. This clustering algorithm can makethe clustering results keep less, and distance between every two data vertexes issmaller than the threshold value. After clustering, based on the fuzzy rule, designthe neural network and do reverse study by using BP study algorithm, and researchthe study process. Based on characteristic that the speed of convergence by usingBP study algorithm is low, we try many different learning rates and finally find outa learning rate which can make the convergence speed of error function relativelyfastest.Finally, we design the driving circuit of ultrasonic motor. The driving circuitcontains signal output module, first-order amplification module, isolation moduleand second-order power amplification module. To the signal output module, basedon the DDS technique, we use AD9852chips to output two sine signals whosefrequency and phase can be changed. We use operational amplifier to do the first- order amplification. And we use HCNR200optoelectronic isolation chips to doisolation between strong current and weak current. Finally, we use PA85chips to dopower amplification. After design and simulation, we make the PCB circuit, and dodriving ultrasonic motor test.
Keywords/Search Tags:ultrasonic motor, driving, fuzzy neural network, clustering
PDF Full Text Request
Related items