Papers by Keyword: Fuzzy Neural Network (FNN)

Paper TitlePage

Authors: Jie Dong, Hai Tao Chen, Sen Liu
Abstract: This paper presents an advanced method with hierarchical architecture for multirobot formation control. The control system consists of fundamental behavior module, supervisor module and velocity tuning module. Formation and obstacle avoidance behaviors are produced by fundamental behavior module, among which the obstacle avoidance behavior is produced by fuzzy logic technique. Then a FNN (Fuzzy Neural Network) is designed to fuse the two fundamental behaviors in supervisor module. The FNN is trained through reinforcement learning. At last, a velocity tuning module is designed to adjust the speed of the robot. Simulation results validate the feasibility of this method.
Authors: Jiao Luo, Miao Quan Li, Y.Q. Hu
Abstract: A constitutive equation has been established to describe the effect of grain size on the deformation behavior of Ti-6.62Al-5.14Sn-1.82Zr alloy during the high temperature. In this paper, firstly a steady flow stress model is proposed, and a function relating to the grain size is introduced to modify the steady flow stress model. Meanwhile, a microstructure model established by the fuzzy neural network method is applied to calculate the grain size of prior α phase during the high temperature deformation of Ti-6.62Al-5.14Sn-1.82Zr alloy. The calculated flow stress using the present constitutive equation shows a good agreement with the experimental flow stress of the Ti-6.62Al-5.14Sn-1.82Zr alloy. The relative maximum error was not more than 15%.
Authors: Hai Bin Su, Zhi Chong Cheng
Abstract: This paper proposes a algorithm of maximum power point tracking using fuzzy Neural Networks for grid-connected photovoltaic systems. The system is composed of a VSI converter, the maximum power point tracking algorithm based on fuzzy Neural Networks outputs a reference voltage as voltage loop import variable. The voltage controller outputs a reference current to control inverter output current in side grid. The fuzzy Neural Networks provide attractive features such as fast response, good performance. Therefore, the system is able to deliver energy to grid. This proposed algorithm is simulated and implemented to evaluate performance. From the simulation and experimental results, the fuzzy Neural Networks can deliver more power than the other algorithm.
Authors: Xiao Chen
Abstract: In this paper, a new DNA based genetic algorithm is proposed to optimize a fuzzy neural network model for a pH neutralization process. In the proposed algorithm, each individual presents a fuzzy neural network encoded by nucleotide base sequence, and modified DNA based crossover operation and three types of mutation operators are designed to improve the searching ability of the algorithm. The study on the performance for two functions shows that the proposed algorithm outperforms GA. Finally, to verify the effectiveness of the established model, it is compared with two models optimized with other methods. Comparison results show that the model optimized by DNA based GA is more accurate.
Authors: Jian De Wu
Abstract: In this paper, BP network is applied to structure multi-level evaluation model to implement evaluation for the kinematic concepts acquired by function analysis. Under this approach, the best concept can be selected once evaluation indicators of each candidate are fuzzily quantified, converted into evaluation attribute value, and fed into the trained network model. During the process, neural network is used to solve the bottle-neck problem of knowledge acquiring and expression, which can be viewed as knowledge base and reasoning engine for the evaluation. At the same time, it is effective in solving the problem of weight distribution in evaluation indicator system. Fuzzy logic is used to achieve the fuzzy quantization for the attribute value of evaluation indicator in evaluation system, which can be used as the I/O value for neural network.
Authors: X.M. Li, Ning Ding
Abstract: An adaptive fuzzy neural network control system in cylindrical grinding process was proposed. In this system, the initial cylindrical grinding parameters were decided by the expert system based on fuzzy neural network. Multi-feed and setting overshoot optimization methods were also adopted during the grinding process, and a human machine cooperation system (composed of human and two fuzzy – neural networks) could revise the process parameters in real-time. The experiment of the cylindrical grinding was implemented. The results showed that this control system was valid, and could greatly improve the cylindrical grinding quality and machining efficiency.
Authors: Zhong Chu Wang, Xin Zhao, Ran Bi
Abstract: Cast-roll hydraulic AGC is the main control means of the strip t hickness. The control effect of traditional PID is poor for adjusting this kind of model parameters. Therefore, a new type of fuzzy neural network self-learning and adaptive controller is proposed, and analyze the composition and basic performance. The simulation results show that the new controller can effectively improve its response, what’s more it has a better dynamic performance more than another control strategies.
Authors: Y.J. Huang, B.W. Hong, P. C. Wu, W. C. Chen
Abstract: In this paper, we propose an adaptive radio frequency identification (RFID) indoor positioning system technology for wheelchair home health care robot with wireless communication. The proposed RFID positioning system uses one reader and four tags which is low cost when applying in a large space of the indoor environment. It reduces the measured calculation by using multiple RFID tags instead of multiple RFID readers. While the measured experimental RFID data found with error leading to signal changes in different environmental parameters, we developed the adaptive fuzzy neural network technology to adjust the measurement data. Through the compensation of the measurement error, the actual wheelchair robot location-based application could be performed to overcome the uncertain environmental parameters. The positioning system provides very good accuracy and make home health care wheelchair robot positioning system available for navigation and guidance.
Authors: Gang Li, Da Peng Li, Yan Yan Zhou, Lin Wei Xu, Ji Feng He
Abstract: Firstly, the implication and purpose of the Combat Service Performance Evaluation for Air Defense Missile was elaborated in this paper. Then the Fuzzy Neural Network (FNN) Evaluating Method was put forward to apply to the combat service performance evaluation. Secondly, an evaluation index system was built on the basis of the characteristics of the combat service operator and the combat service process. At last, the correctness and validity were approved though the simulation results.
Authors: Fang Ju Ai
Abstract: The number of fuzzy rules directly determines the complexity and efficiency of Fuzzy Neural Network (FNN). The Iterative Pruning (IP) algorithm belongs to the Pruning Method, and it spends much time computing adjusting factors of the remaining weights. So the Improved Iterative Pruning (IIP) algorithm is put forward, which adopts dividing blocks strategy and uses the Generalized Inverse Matrix (GIM) algorithm to replace the Conjugate Gradient Precondition Normal Equation (CGPCNE) algorithm for updating the remaining weights. The IIP algorithm is applied in the rule-reasoning layer of FNN to simplify its rules and structure in a great extent and preserve a good level of accuracy and generalization ability without retraining after pruning. The simulation results demonstrate the effectiveness and the feasibility of the algorithm.
Showing 1 to 10 of 113 Paper Titles