Dissertation
Dissertation > Industrial Technology > Electrotechnical > Motor > DC motor

Brushless Dc Motor Control System Design Based on the Nonlinear Dynamic PID Neural Network

Author LinDongMei
Tutor ZengZhao
School Changsha University of Science and Technology
Course Circuits and Systems
Keywords Brushless DC Motor Neural Network PID Single Neural Network Dynamic PID Neural Network
CLC TM33
Type Master's thesis
Year 2009
Downloads 193
Quotes 2
Download Dissertation

The brushless DC motor is simple in structure and operates reliably, so it is widely used in national product and life. Following social progress, the requirements of brushless DC motor performance are becoming higher. Improving the performance of brushless DC motor can bring great economic and social benefits. The methods include optimal design of motor body, control of power electronic equipment and advanced control strategies.Brushless DC motor has characteristics of nonlinear, multivariable and strong coupling, for this reason, traditional PID control algorithm is difficult to meet the performance requirements of high precise servo system, thus it is difficult to guarantee motor running in high precision. Traditional neural network algorithm has strong adaptive ability and can learn on-line, which has satisfactory results to the controlled object with characteristics of nonlinear, time-varying and strong coupling. But the control structure is too complex and linear adjustment is not easy to achieve for too many parameters.In view of the above problems, this paper puts forward a kind of nonlinear dynamic PID neural network control algorithm, in order to realize the brushless dc motor control precision.For general neural network algorithm is limited to solve the above problems, this dissertation studies the types of neural network excitation function, has given the domain of neural network learing rate keeping neural network algorithm converge. The coefficient of traditional PID (proportion, integral and differential) is varied following the error, which is also described to curve. Furthermore, we often denote the curve as exponential error function or nonlinear function. Next, a nonlinear dynamic PID control law is constructed. And then three weight coefficients are trained online by neural network, making brushless DC motor achieve nonlinear intelligent control.At last, this dissertation constructs the model of brushless DC motor cascade control system. Traditional PID control algorithm and nonlinear dynamic PID neural network control algorithm is used in external speed control loop respectively. Simulation results indicate that nonlinear dynamic PID neural network control algorithm has reached good effect, superior to traditional PID control algorithm.

Related Dissertations
More Dissertations