Forum pracy nad projektem wskrzeszenia The Forge

[lin91957]

Improving neural network algorithm efficiency in the hydro-generating unit Calculation


46-142443ll4.1286-23.652715.2456ll】 = I-1.1752330733-2l7036l ( l3.0642-27.9o6017.088lll76473. 33.1578-3O7851l 【2. B2】 = [-o1359027430.48760.3690-1.6181.0361】 table test data and calculated results with l the number of training plans 2iIII Yan difference curve (continued on page 128) Power System Engineering Research 2001, recorded l7 observed the characteristic curve and compared with the original curves to compare the results can be used to take preventive measures. This will improve plant performance and low cost, to maintain the best running sense. (4) generator monitoring and diagnostic software package. Support system can be detected earlier and capture the deviation of generator operating parameters. (5) process optimization package. State of the control parameters to assess the current state of the plant. Gives the percentage of the cost differential. Running on the current state of computing power, but also to calculate the cost of the year. 5 system defects (】) PROCONTROLP systems and networks during a remote bus. Is centrally managed by the bus communication. On the bus have a bus controller. Responsible for communications control. Once the bus controller fails, the impact on the remote bus communication. Although the use of redundant remote bus, but the damage caused by failure is still difficult to estimate. (2) bridge implementation. Between the two networks. To have a bridge to implement the forwarding of information. PROC0NTROLP system, the bridge is inserted in a bus station on the Ethernet card to achieve. Ethernet card is equipped with an Ethernet cable A plug connected to the server with the Communication, the Ethernet card through the bus station to receive information on the remote bus, and then forwarded to the server from the management network to send information through the Ethernet bus station,[link widoczny dla zalogowanych], and then Station bus controller then be sent to the remote bus. Because the bridge is not directly connected with the remote bus,[link widoczny dla zalogowanych], the bus must stop to transfer, so that the state of the bus station to the bridge is bound to affect the work, thus affecting the two O information transmission. 6 Conclusion PROCONTROLPDCS system is a new generation of distributed systems, whether in performance or reliability has its own unique, it results from its application can be seen. However, remote bus communication and bridge design,[link widoczny dla zalogowanych], there is still need for improvement. 【Reference port 11 set to teach the king often power system applications pulling machine to pick the trial】 Industry Press bar l999. [2】 Ren Chengyu computer controlled constant | Technology and Systems, published by Du 19'85 hydropower Editor: Xiao Min (continued from page 7】 l) of the amount of holding,% holding smoke 氲 ■ lap 7, NO 4 row pull out the relationship between the amount of conclusions After changes in different coal boiler, under different loads. About emissions 200pPm N (410mg, m; O.121mg, kJ). To the existing areas of Beijing holding a class put standards. Can effectively reduce NO emissions, and high efficiency, low cost. The corresponding transformation for the boiler. Population [1】 Mao Jian heap shame + D field clean-burning coal. Beijing: Science Press dictionary of Figure Han Bin tangential combustion control No. J】 【emission wood boiler technology. t995 (9) 【3】 GBI3 ~-l99 power plant air pollutant emission standards [s】 fine New eye of an ax bad burning coal with low N0. - A PM combustion clamor. Mitsubishi Heavy Industries Technical Report, 1995,32 (1) Editor: Wen Chang (Continued from page 120) 3 Conclusion Based on the existing basis of BP network learning algorithm. With quasi-Newton nonlinear optimization method. BP network is given a training on the new learning algorithm effectively improves the speed of BP network learning convergence is obtained than the conventional BP learning algorithm for better convergence performance and learning speed. Will be applied to the efficiency of hydroelectric generating set curve calculation can be quickly and accurately the efficiency of typical test data under the head to learn, and then right after the end of the training and threshold value and the selected network structure model, can be easily to calculate the efficiency of any operating conditions, the results of high precision,[link widoczny dla zalogowanych], strong adaptability. L】 【References mouth Yong Shu Chuan Wan Station of economic principles. Beijing: China Water Conservancy and Hydropower Publishing well-being. 1998. f2】 Zhou Hui, Du Cong Ho based on artificial neural network of water shoes generator efficiency curve calculated ah. The beginning of eddy water science. 19g120l32 ~ l35 [3] Yuan Yaxiang. San contingent nonlinear algorithms regulatory Ba Pai: Shanghai Science and Technology Publishing Du wood. 1992 David King stabbed the wall with neural network application groan】 Xiamen, Xi'an,[link widoczny dla zalogowanych], Xi'an Jiaotong University Press, 19
More articles related to topics：


[link widoczny dla zalogowanych]

oakley brillen _5657 New combined microwave oven t

[link widoczny dla zalogowanych]