Study on Intelligent Control Method for Double Input-Double Output Piezoelectric Hydroelectric Servo Valve
when the double input-double output piezoelectric hydroelectric servo valve (two ends of the sliding valve core are driven by two piezoelectric stacks)is working, the piezo-actuators need to be added bias voltage. When the valve core moves to one end, the voltage of the piezo-actuators of this end need to reduce and the voltage of the piezo-actuators of the other end need to increase; in principle, when the reducing length of one end piezo-actuators is equal to the increasing length of the other end piezo-actuators, (he valve core can move with celerity, precision and tranquilization; however , the hysteresis and nonlinearity of the piezoelectric stack make the outputs of two piezo-actuators of the valve core two end have very coupling role, which makes the speed, precision and tranquilization of the valve core reduce. To adopt the common PID controller could realize decoupling control at a certain extent, however, the design and realization of the decoupling controller is complex and the decoupling precision depends on the precision of the system recognition model. So a parameter self-tuning double input-double output NN PID controller basing on BP NN is introduced to decouple control for the double input-double output piezoelectric hydroelectric servo valve. Basing on the certain enactment control law , the NN PID controller adjust the proportion, integral and differential parameters of PID controller online by network, then we can utilize the classical PID controller to gain two control values which participate in control, the NN PID controller can realize decoupling control well in this process and it doesnt need stylebooks to leam off-line. The result of experiment shows that the performance of the NN PID controller is better man that of the PID controller in control of the piezoelectric hydroelectric servo valve.
piezoelectric hydroelectric servo valve piezo-actuators decoupling control NN PID controller
Cao Feng Jiao Zongxia Liang Lei Wang Shaoping
School of Automation Science and Electrical Engineering, Beijing University of Aeronautics and Astronautics,Beijing, 100083,China
国际会议
北戴河
英文
507-512
2007-06-06（万方平台首次上网日期，不代表论文的发表时间）