mac cosmetics discount Wavelet neural network adap

Wavelet neural network adaptive inverse control of hydropower units

△ a run-length unit 2S, the signal amplitude of 3 【,[link widoczny dla zalogowanych], (Chi) applied to the controlled object, the corresponding received y (K), the {(pot), y (k a 1), U (pot), U ( pot-1)} as the input stimulus applied to the identification WNI WNI learning, WNI will be trained access control system, inverse controller WNC 3-5-1 three-layer network structure used,[link widoczny dla zalogowanych], the network input is: {Y (ten pots 1), (pot), U (pot-1)}, control the amount of 77 to take the weighted coefficient of 4.75, the required network parameters training Jacobian data provided by the WNI. Figure 2, Figure 3,[link widoczny dla zalogowanych], Figure 4, solid line for the load rejection speed of 20 (X), head (H) and the guide vane opening (y) the control results, dotted line under the same conditions 1, activation function with s (z) a ÷ __ 1T God by having only one network control effect. It can be seen from the figure the wavelet neural network adaptive inverse control method can effectively control the hydroelectric units, and better than the neural network control, manifested as small overshoot, settling time is short. Figure 5 shows the effect of parameters to control 10, when system instability. 0l40l20.100080060040.020-002 (21) Figure 22O% load disturbance response of 154 College of the control of the first 18 0l0203040506o7080f / s Figure 320% load disturbance change process 0.0-0.1-0.2 10 head. Figure 3 a 0.4-0.5-0.6-070】 O203040506o7080t vane 420% load disturbance change process 06O. 40-2 Yuk 0.0 a 02 a 04-06-0.80l020304050607080t / s Figure 5 parameter 7 = 0 3 Conclusion system response to achieve by wavelet neural network model of the controlled object is identified, and using the inverse model as a controller,[link widoczny dla zalogowanych], the control by constructing a generalized weighted objective function, for non-minimum phase systems with nonlinear hydropower unit control were studied. Theoretical analysis and simulation show that by taking appropriate control weighting factor, based on wavelet neural network adaptive inverse control can be used for the control of hydropower generating units, and the wavelet neural network control better than the neural network control. 【References [1] AUTOMATION AND Zhou Shixiang. BP network intelligent PID control unit of water [J]. China Institute of Metrology, 2006,17 (1) 72-74. [2] Sun Wei, Yao-nan. Wavelet network-based fuzzy neural network for the robot tracking control [J]. Control Theory and Applications, 2003,20 (1) :49-53. [3] Hu Weili. Inverse control of nonlinear internal model control [J]. Automatica Sinica, 2002,28 (5) :715-721. [4] AUTOMATION AND SHEN Zu-yi. Based on Wavelet Network on Control Application on the hydro-generating unit [J]. Large motor technology, 2005,[link widoczny dla zalogowanych],15 (2) :56-59. [5] Liu Jinkun, Liu Tao. Wavelet neural network based adaptive inverse control and its application [J]. Systems Engineering and Electronics, 2003,25 (5) :591-594. [6] WRAYJ, GREENGR. Neuralnetworks, approximationtheoryandfiniteprecisioncomputation [J]. NeuralNetworks, 1995,8 (1) :31-37. [7] ZHANGQH, BENVENISEA. Waveletnetworks [J]. IEEETransonNeuralNetworks. 1992,3 ( Cool

:889-898. [8] HYDROELECTRIC POWER. Sensor Based on Wavelet Network for Nonlinear characteristics of linear [J]. China Institute of Metrology, 2004,15 (3): 191194. [9] I, IX, CHENzQ, YUANzz. Simplerecurrentneuralnetworkcontrolfornon-minimumphasenonlinearsystem [J]. ControlTheoryandApplications, 2001,18 (3) :456-460. [1O] AUTOMATION AND SHEN Zu-yi. Wavelet network based adaptive control of hydraulic turbine [J]. Hydroelectric Engineering, 2005,24 (1) :118-121. m ∞ 0OOOOOOOO one hundred and eleven {H

相关的主题文章：

[link widoczny dla zalogowanych]

[link widoczny dla zalogowanych]

[link widoczny dla zalogowanych]

The post has been approved 0 times