【正文】 m monitoring data according to the circuit characteristics to consider redundancy control circuit the system of the munist party of China 17 frequency analog and 3 road, which is divided into 6 groups, signals are grouped as shown in table 1, including 17 road analog was divided into 5 groups, 3 road frequency quantity is 1 set alone, each group use the same signal disposal circuit undertake recuperating, therefore, the redundancy of the system circuit design, make full use of the conditions in each group share a redundant signal regulate circuit, namely each group use the same modulation circuit, respectively, using the original control signal in normal operation of circuit, when the set is in any way the signal conditioning circuit failure occurs, can switch to the standby redundancy control circuit, continuing state monitoring, so as to improve the accuracy and reliability of system, due to the signal in each group share a conditioning circuit, and will initially need to increase redundant 20 road son signal disposal circuit module circuit, reduced to simply add redundancy sub circuit for signal disposal circuit, greatly reducing the regulate the number of sub circuit, and improve the accuracy and reliability of the system.According to the above analysis of the monitoring system for monitoring signal and modulation circuit, design of the nursing child son/redundant circuit switching circuit, the event of a failure regulate circuit and redundant circuit switching process is as follows： （14）W (t) is the output of the hysteresis that w (t) is equal to zero, then the system is not stable, because for any x. the simulation, using the generalized RBF neural network for modeling of inverse hysteresis control signal for r = t and the output of the neural network inverse model for v (t), as shown in figure 3, the model by the neural network inverse hysteresis system under the action of single output for 1 w (t), as shown in figure can be seen from the figure 3 and figure 4, greatly weakened, hysteresis andhysteresis phenomenon basically of PID controller is: p = 70 K, K (I) = K d = Bang control rule is: A = , K B = 50, Sp and Sp 1 2 are and figure 5 is not plus Bang Bang control simulation results, namely for the NPID control, figure 6, 7, 8, and Bang Bang control after NBPID control simulation results.4 conclusionHas good characteristics through the use of generalized RBF neural network, direct inverse model of hysteresis system modeling, and then use the inverse model of the implementation of feedforward control, can greatly reduce the hysteresis joining Bang Bang control, can effectively control can be seen from the simulation results, based on feedforward control plus Bang Bang control of PID control, the hysteresis system can be effectively method can also be extended to other types of hysteresis control system.References:[ 1] Hamdan M, Gao Z Q. A novel PID controller fo r pneumat icproportio nal valv es with hyster esis [ J ] . IEEE, 2000, 2:1198 1201.[ 2] Zhao Hongwei, etc. Piezoelectric ceramic actuator in the application of flexible manipulator robot research [J]. Journal of piezoelectric and acoustics, 2000, 22 (3) : 173176.[ 3] Choi G S, Kim H S, Cho i G H. A study on position controlof piezoelectric acuators [ A] . ISIE’ 97 [ C] . Portug al,1997.[ 4] Tao G, Kokotovic P V. Adaptive contr ol of plants w ith unknow n hysteresis [ J] . IEEE Tr ans. Autom. Contr. 1995,40( 2) : 200 212.[ 5] SU C Y, Stepanenko Y, Svoboda J, et al. Robust adaptive contro l of a class of nonlinear systems [ J] . IEEE T ran. . Con. 2000, 45( 12) : 2427 2432.[ 6] Cruz Her nndez J M, Hayward V. Phase control approach to hyster esis reduction [ J] . IEEE Tran. on Contr . . 2001, 9( 1) : 17 26.[ 7] Hwang C L, Jan C, Chen Y H. Piezomechanics using intelli
gent variable structure control [ J] . IEEE Tran. o n Industrial Electo nics. 2001, 48( 1) : 47 59.[ 8] Han J M. T. A. Adriaens. Willem L. de Koning , ReinderBanning. Modeling piezo electric actuators [ J ] . IEEE/ASME Tran. Mech. 2000, 5( 4) : 331 341.[ 9] wang yong ji stuff. Neural network control [M]. Machinery industry press,[ 10] Hay kin S. Neur al Networks[ M] . Prentice Hall I nc. 1999.航空發(fā)動(dòng)機(jī)狀態(tài)監(jiān)測(cè)系統(tǒng)設(shè)計(jì)研究康文雄、李華聰、楊勇柯( 1. 華南理工大學(xué), 廣東 廣州510640。Under 16 way switch quantity after adopting the light into a the data processing part of the need for continuou