Papers by Keyword: BP Network

Abstract: Artificial Neural Networks have been used for a wide variety of chemical applications because of their ability to learn system features.This paper presents the use of BP neural networks for soft senor of the cell concentration.For microbial concentration in fermentation process of the problem of real-time online detection,from the perspective of biological cell metabolism in the process of fermentation were analyzed by the correlation of metabolic parameters,then these parameters as the input variables, established the cell concentration BP neural network model of soft senor for biomass parameters detection in complex system.The simulation results show that the soft senor method has highly accuracy and can well applied in the biological fermentation process.

Abstract: Because of the defect in traditional BP network of cementing quality prediction at present which is sensitive with the initial weights, easy to fall into the local least value,low forecast precision and slow convergence speed occurred. In order to overcome the shortcomings of traditional BP network, the paper introduced the particle swarm optimization (PSO) algorithm based on the random global optimization into the neural network training. The PSO is used to optimize weights of BP network. The simulation results show that this method has shorter training time and higher prediction accuracy than the BP network, and it can improve cementing quality and realize prediction and tracking analysis of cementing quality. It has good serviceability for predicting all kinds of information not known in cementing. It has provided a new method for cementing quality prediction.

Abstract: The purpose of the research is to put up a fault diagnosis algorithm of the aero-engine’s fuel flow sensors and verify the platform through simulation based on the QAR (Quick Access Recorder) data. By analyzing the correlations of the parameters that affect the conditions of the engine, a three-layer BP network model is established. Then, in order to solve the influences to the BP model in the full envelope due to the fluctuations of transition between different flight phases, a region partition method is used to divide the whole flight envelop into several regions and corresponding BP model is established. Finally, the QAR data are used as the training samples to build the BP network for different regions, then, the real-time data are used as the inputs to verify the platform. The simulation results show that the region partition method can effectively detect the fault of the fuel flow sensors.

Abstract: There are many mathematical models in GPS height conversion fitting algorithms, such as polynomial surface fitting, Multi-surface fitting function, BP neural network and so on. The paper is focused on the fitting algorithms and summary of these kinds of GPS height conversion approaches and something useful is concluded.

Abstract: The classification and prediction of load is very important, in the power market .In order to improve the accuracy and speed of forecast, it is proposed that the mixed algorithm of particle swarm and back propagation network and model. And model is established on the basis of one city electric power bureaus electric power load data. Using the PSO - BP algorithm to the load for forecasting .According to the results of prediction, this method converges fast, prediction accuracy improved significantly. Application in the power market analysis and forecasting have very good effect and prospect.

Abstract: Traditional transformer fault diagnosis based on single source of information has significant limitation in identification of transformer fault type because of power transformers complex structure and changeable operating environment. So fusion technology is introduced into the fault diagnosis of power transformer. This method divides the progress of transformer fault diagnosis into two fusion levels. The first level is to ascertain whether it is overheated or discharged by content of gases dissolved in transformer oil. The second level is to ascertain the location or cause of the fault by electric data. The intelligence algorithms which are used in these two levels are both the improved BP neural network algorithm. Finally, the effectiveness is validated by the result of practical fault diagnosis examples.

Abstract: Classification of lettuce growth peroid is the premise of records of lettuce growth information. In this study, lettuce images in every growth period are collected. And visible images are preprocessed to extract features to establish initial feature library of lettuce images. Through R cluster analysis on many features, good image eigenvector are obtained. Classification of the lettuce samples are obtained by modeling and analysis of the neural networks. The experimental classification results compared with practical classification results, the recognition accuracy is up to 88.4%.

Abstract: Because of the existing problems of BP network: easily falling into local minimum point, slowly converging and generalization ability can not be guaranteed, the evaluation method of water quality based on BP network is not satisfactory. Therefore, an algorithm of improved particle swarm optimizing is used to optimize the BP network. On the basis of this, an evaluation method of water quality based on improved PSO-BP network is designed. Proved by experiments, the BP network optimized by improved PSO is stable. So, the performance and efficiency of the water quality evaluation based on improved PSO-BP is pretty good.

Abstract: The text introduced a system based on BP network for the prediction of grape disease. It included the design of a network structure, the selection of parameter for network study, the processing of sample data etc. With the use of BP network model, this system can forecast the extent of grape disease, so it is applicable to the conditions which have many influencing factors, complicated relationship, difficulty of analyze quantitatively and requirement of long-term prediction. Using this system to the prediction of grape disease in Zhuo Lu area Zhang Jia Kou city, the authors obtained a good effect, which is of value to the prediction of grape disease occurrence.

Abstract: Aiming at the problems of BP network algorithm easily falling into local minimum point, slow converging and the problem that generalization ability can not be guaranteed, a method to improve the PSO is proposed. This method of improved PSO can strengthen the parameters of BP network. Based on this, a license plate recognition algorithm is designed. Some conclusions can be drawn from the experiments: (1) the improved PSO-BP network is stable and robust which can avoid falling into flat areas and local minimum point. (2) the performance and efficiency of license plate recognition based on the improved PSO-BP network is pretty good.