Papers by Keyword: BP Network

Abstract: In order to improve the prediction accuracy and reduce the influence caused by earthquake disasters, the method of predicting the earthquake is proposed. The method aim at nonlinear problems, limitation in training samples and unbalanced distribution that exist in the earthquake prediction, according to the sample the method build training and forecast BP network. The results of simulation experiments prove that the method is feasible and also helpful to improve the accuracy of earthquake prediction, because of less error.

Abstract: Aimed at Neural Network can approach any nonlinear system with arbitrary accuracy, the frame of distributed NN decoupling system are proposed to decouple the MIMO nonlinear system. In this paper, we designed and finished the Distributed Control System based on ABB’s Freelance 800F, and collected experimental data to model the thermostatic heater, then we have carried out the mathematical model by means of MATLAB dynamic simulation. In sequence, we trained the neural network controller in MATLAB. When the decoupling is completed, we used controller to control the MIMO nonlinear system in DCS. Experiment result shows that it is conscientiously feasible and deserves to be widely applied in the process of controlling industry.

Abstract: Currently the BP networks which used in friction welding defect recognition existents the problem of slow convergence and easy to fall into local optimum. This text will introduce the germ arithmetic into the BP network to turn a beginning to start the power value, putting forward the improvement BP network of the blemish identify. Pass to verify the blemish signal of the some GH4169 heat metal alloys detected by a super voice C, which expresses that the method makes blemish qualitative analysis classification acquired higher accuracy, and it can lower artificial and subjective factor consumedly to the influence that the blemish identify.

Abstract: How to ensure the health of pile, which is a problem that engineers study much. The principle of back stress wave method is used to detect the health of pile. It studies the synergetic technology of BP algorithm and genetic algorithm. The paper analyses the performance of the synergetic model and applies the model in the area of pile's health detection. Experimental results show that the method is better than the traditional methods.

Abstract: In view of the virtue and shortage of genetic algorithm and BP network, this paper proposes a new BP network training method based on improved genetic algorithm (IGA-BP). This algorithm uses hierarchical code, adaptive crossover and mutation, pruning similar chromosomes, dynamic supply new chromosomes and other operations, so the network structure and weight are optimized at the same time and the "premature" phenomenon is avoided. The simulation results show that the IGA-BP network architecture is simple, the convergence rate is quick, and has good approximation and generalization ability.

Abstract: Artificial Neural Network(ANN) is an effective method of data analysis and way of processing, which is widely used. In the article BP artificial neural network was used to set up groundwater quality assessment model in Industrial Park Catchment. The matlab toolbox was used to calculate it, and the result was satisfactory. The results of the study show that BP artificial neural network in the appraisal of water quality has the advantages of simplicity and practical.

Abstract: The barrel life is decided by explosive payload, tempreture, firing interval, ammunition type and other factors. Thus the zone charge’s auto-recognition is the first question to realize the automatic prediction technology of the barrel life. The zone charge’s recognition technology was proposed based on BP Network. The theory support could be provided by the technology to realize the zone charge’s auto-recognition.

Abstract: Sadness, one of the negative emotions, may cause undesirable impact to the daily life. Therefore, it is desirable to automatically detect sadness emotion in human-machine interactions in order to adopt measures to impair the negative effects caused by it. Speech is one of the means used by human to express emotions, therefore, it is reasonable to detect sadness emotion using speech samples. In this paper, we analyzed relevant speech features, and proposed an improved Back Propagation (BP) network for sadness recognition. The experimental results show that the improved BP network proposed has better performance than traditional BP networks in detecting sadness emotion.

Abstract: A new kind of swarm intelligence algorithm called stochastic focusing search(SFS) is proposed and applied to optimize sheet metal forming process in this paper. The steps of the optimization procedure include combining numerical simulation technology with orthogonal experiments to provide training samples for BP net, and producing the fitted function as optimization function for SFS algorithm. The validation of the final optimization results by a rectangular box part stamping case shows that this kind of optimization methodology is correct and reliable for the design of deep drawing process. Advantages of the SFS algorithm are demonstrated that SFS has good global searching ability and fast convergence speed in finding optimal solutions, which means the optimization method using SFS algorithm can provide a competitive way of solving the optimization design problems in sheet metal forming.

Abstract: This paper is to present a novel design to predict the State of charge (SOC) of the batteries for the Electric Vehicles (EV) using a voltage descent model which has been built based on the analysis of adaptive fuzzy neural intelligent algorithm (AFNIA) and the charge/discharge experimental data of Electric Vehicle. In this design, an improved BP neural network has also been proposed to indicate the correlation between open circuit voltage and SOC. An experiment employed a Lateral Moving and In Situ Steering EV built by Shenzhen Polytechnic. The test and simulation results showed that the intelligent methods can accurately predict the SOC of lithium batteries. The combination of fuzzy control and neural network can achieve an effective way of predicting the SOC of batteries.