Search

In view of artificial neural network, there are some deficiencies in condenser fault diagnosis. The BP neural network used for condenser fault diagnosis is highly nonlinear pattern recognition and high precision of fault diagnosis. The PSO-BP neural network can effectively solve the problem of BP neural network that training time is long and training process is easy to fall into the local minimum. The training results of PSO-BP network in convergence speed and convergence effect are significantly improved. Under the condition of small samples, the calculation results of SVM method are better than the calculation result of the other two methods. Although the recognition ratio of improved PSO-BP(2) and SVM is the same, training time of improved PSO-BP(2) is longer than training time of SVM. The generalization ability of SVM is stronger, and the efficiency of SVM is higher than the neural network. With MATLAB programming, three different algorithms, which are BP neural network, PSO-BP neural network and SVM, are studied and compared for the performance of condenser fault diagnosis. In the models of this study, the research results show that condenser fault diagnosis based on SVM has the fastest convergence speed and the best accuracy.

Due to the complexity of the power plant production environment, it brings some difficulties to troubleshooting of turbine generator. Although the approach based on neural network has been widely used in fault diagnosis of equipment, the result of fault diagnosis, which is given by the single neural network, is often not ready to determine the fault type for turbine generator. In response to this situation, a fault diagnosis method based on multi-neural network is proposed on this paper. It means that the different neural network is to be used respectively for fault diagnosis of turbine vibration firstly. Then the results of these initial diagnoses are to be integrated with information fusion technology. Through this strategy, the reliable result of fault diagnosis is obtained and the disadvantage of inaccurate diagnosis based on a single neural network is overcome.

Power system short-term load forecasting have a significant role in the power system planning and reliable operation. This paper compared the result of two prediction models BP neural network and wavelet neural network (WNN). According to comparison results, WNN model in the case of the same number of nodes have better predictive accuracy. In order to achieve the balance of the least input neurons and prediction precision, the regionality human body amenity indicator is used as the input of meteorological factors. Meanwhile, in order to take full advantage of the historical load data, improve the prediction precision, the concept of innovation is introduced and the load data of forecast base day is included in the input range. The model is verified by using the actual data, the results show that the prediction results is accurate and the model in practical and effective.

Power system short-term load forecasting have a significant role in the power system planning and reliable operation. According to the deficiency of time variation and easy to fall into the local least value in feed forward neural networks, a forecasting Model is provided by using Elman feedback neural network. Hourly weather factors can improve the prediction precision. In order to achieve the balance of the least input neurons and prediction precision, the regionality human body amenity indicator is used as the input of meteorological factors. Meanwhile, in order to take full advantage of the historical load data, improve the prediction precision, the concept of innovation is introduced and the load data of forecast base day is included in the input range. The model is verified by using the actual data, the results show that the prediction results is accurate and the model in practical and effective.

The monitoring of operation states and fault diagnosis system of turbines in power plant are significant to guarantee the units long-term safety and economic operation. The fault diagnosis of turbines is influenced by various factors, the information provided by every sensor need to be used in comprehensive to enhance the accuracy and reliability of fault diagnosis. The paper has presented a fault diagnosis system based on D-S evidence theory and neural network. Firstly, the inputs of neural network were fuzzed through fuzzy membership function, using BP neural network to train and simulate them, but in some cases, the diagnostic results of neural network were unable to determine the fault type accurately, therefore the information fusion was need. The D-S evidence was applied. This method has been proved by the fault diagnosis of turbine equipment in power plant, which can determine the fault type accurately. Particularly, in the case when the fault type can't be determined by using neural network. Therefore, the method was reliable and effective, for fault diagnosis in power plant equipment, it was significant.

For the reason that the application of the neural network in the condenser fault diagnosis has some restrictions, the result of the fault diagnosis based on neural network is unsatisfied. On the basis of the fault diagnosis of condenser based on neural network, the thought of data fusion is introduced in this paper and a method of condenser integrated fault diagnosis based on multiple neural networks and D-S evidence theory is proposed. According to BP neural network and CPN network, the respective diagnosis results regarded as the primary evidences of D-S theory evidence in decision layer are obtained first, and then these results are fused by using of the evidence theory to obtain the final diagnosis result. The result of simulation shows that: Comparing with the result from the single network, this method has a smaller error and higher diagnosis reliability.

A new method of fault diagnosis for turbines using preserving nonlinear and quasi-Newton algorithm for the learning process of BP neural network was introduced in the paper, Which take advantages of preserving nonlinear and quasi-Newton algorithm with fast speed, significant superiority for high-dimensional problems and more accurate mathematical models comparing with traditional BP neural network. Though simulation studies on typical fault diagnosis examples of turbines in power plant, the results have shown that the performances of BP neural network based on preserving nonlinear and quasi-Newton algorithm is superior to the traditional BP algorithm, improving the ability of online diagnosis for turbines and having broad application prospects and value.

In order to enhance the accuracy rate of the fault diagnosis of condenser, a fault diagnosis method based on consensus information fusion is presented. The decision-making distance is calculated and the closest supporting group is found. By integrating the results of the closest supporting group, the best decision-making value of fault diagnosis is obtained. This method is applied in the simulation experiment of condenser fault diagnosis in this paper. The experimental results show that the consensus information fusion method is able to improve the reliability of the fault diagnosis and reduce erratic elements evidently.