Nonlinear optimized adaptive trajectory control of helicopter

This paper attempts to develop an optimized adaptive trajectory control system for helicopters based on the dynamic inversion method. This control algorithm is implemented by three time-scale separation architectures. Pseudo control hedging (PCH) is used to protect the adaptive element from actuator saturation nonlinearities and also from the inner-outer-loop interaction. In addition, to augment the attitude control system, two online adaptive architectures that employ a neural network are used. By tuning the neural network based on the system model, a better and faster learning will be achieved, but this is a frustrating and time consuming process. Due to complexity in accurate tuning of neural network, this paper introduces a non-dominated sorting genetic algorithm II (NSGA-II) for off-line optimization of the neural network. Thus, in the proposed method, the neural network can compensate model inversion error caused by the deficiency of full knowledge of helicopter dynamics more accurately. The effectiveness of proposed method is demonstrated by numerical simulations.     

基于在线最小二乘支持向量机的广义预测控制

This paper proposes a practical generalized predictive control(GPC)algorithm based on online least squares support vector machines(LS-SVM)which can deal with nonlinear systems effectively.At each sampling period the algorithm recursively modifies the model by adding a new data pair and deleting the least important one out of the consideration on realtime property.The data pair deleted is determined by the absolute value of lagrange multiplier from last sampling period.The paper gives the recursive algorithm of model parameters when adding a new data pair and deleting an existent one,respectively,and thus the inversion of a large matrix is avoided and the memory can be controlled by the algorithm entirely.The nonlinear LS-SVM model is applied in GPC algorithm at each sampling period.The experiments of generalized predictive control on pH neutralizing process show the effectiveness and practicality of the proposed algorithm.     

Generalized Predictive Control with Online Least Squares Support Vector Machines

This paper proposes a practical generalized predictive control （GPC） algorithm based on online least squares support vector machines （LS-SVM） which can deal with nonlinear systems effectively. At each sampling period the algorithm recursively modifies the model by adding a new data pair and deleting the least important one out of the consideration on realtime property. The data pair deleted is determined by the absolute value of lagrange multiplier from last sampling period. The paper gives the recursive algorithm of model parameters when adding a new data pair and deleting an existent one, respectively, and thus the inversion of a large matrix is avoided and the memory can be controlled by the algorithm entirely. The nonlinear LS-SVM model is applied in GPC algorithm at each sampling period. The experiments of generalized predictive control on pH neutralizing process show the effectiveness and practicality of the proposed algorithm.     

A QoS-aware routing algorithm based on ant-cluster in wireless multimedia sensor networks

QoS-aware routing algorithm is important in wireless multimedia sensor networks. This paper formulates a generalized QoS-aware routing model on the basis of multiple routing metrics and priorities of packets. We first introduce a 2D plain-based routing algorithm IPACR which improves the standard ant colony algorithm by optimizing the initial distribution of artificial pheromone in order to accelerate the algorithm convergence rate. Then a clustering-based routing algorithm ICACR is presented which can be well applied in a large scale network. ICACR is a variation of IPACR because it can be suitable for clustering cases to satisfy the larger scale situations. Both the numerical algorithm performance analysis and simulation of IPACR and ICACR are given. The results show that ICACR outperforms IPACR in terms of both network lifetime and QoS-aware routing metrics in large scale wireless multimedia sensor networks. Moreover, the simulation based on the real video traces shows that by extending the multi-path to ICACR for different priorities of video frames better performance can be achieved.     

Research and Design of a Fuzzy Neural Expert System

We have developed a fuzzy neural expert system that has the precision and learning ability of a neural network.Knowledge is acquired from domain experts as fuzzy rules and membership functions.Then,they are converted into a neural network which implements fuzzy inference without rule matching.The neural network is applied to problem-solving and learns from the data obtained during operation to enhance the accuracy.The learning ability of the neural network makes it easy to modify the membership functions defined by domain experts.Also,by modifying the weights of neural networks adaptively,the problem of belief propagation in conventional expert systems can be solved easily.Converting the neural network back into fuzzy rules and membership functions helps explain the inner representation and operation of the neural network.     

Heuristic dynamic programming-based learning control for discrete-time disturbed multi-agent systems

Owing to extensive applications in many fields, the synchronization problem has been widely investigated in multi-agent systems. The synchronization for multi-agent systems is a pivotal issue, which means that under the designed control policy, the output of systems or the state of each agent can be consistent with the leader. The purpose of this paper is to investigate a heuristic dynamic programming (HDP)-based learning tracking control for discrete-time multi-agent systems to achieve synchronization while considering disturbances in systems. Besides, due to the difficulty of solving the coupled Hamilton– Jacobi–Bellman equation analytically, an improved HDP learning control algorithm is proposed to realize the synchronization between the leader and all following agents, which is executed by an action-critic neural network. The action and critic neural network are utilized to learn the optimal control policy and cost function, respectively, by means of introducing an auxiliary action network. Finally, two numerical examples and a practical application of mobile robots are presented to demonstrate the control performance of the HDP-based learning control algorithm.     

Support Vector Machine and Artificial Neural Networks for Hydrological Cycles Classifications of a Water Reservoir in the Amazon

This paper presents a methodology based on computational intelligence techniques for classification of hydrological cycles that can infer the change in the physico-chemieal parameters and metals from the water of a reservoir in the Amazon. The methodology initially consists in perform a pre-processing the data to select the most relevant variables of the samples. After that, we compared two different machine learning classifiers, namely SVM （support vector machine） and ANN （artificial neural network）. The automatic model selection is made to choose the parameters of the classifiers. The results indicate that the support vector machine classifier using radial basis function or polynomial kernel exhibited superior results to ANN in terms of overall accuracy and robustness. The SVM classifier accuracy （89.1%） can be considered satisfactory, since there is a great variability of physico-chemical parameters and metals in the hydrological cycles and in the different ecosystems where are the sampling station.     

Adaptive output-feedback regulation for nonlinear delayed systems using neural network

A novel adaptive neural network （NN） output-feedback regulation algorithm for a class of nonlinear time-varying timedelay systems is proposed. Both the designed observer and controller are independent of time delay. Different from the existing results, where the upper bounding functions of time-delay terms are assumed to be known, we only use an NN to compensate for all unknown upper bounding functions without that assumption. The proposed design method is proved to be able to guarantee semi-global uniform ultimate boundedness of all the signals in the closed system, and the system output is proved to converge to a small neighborhood of the origin. The simulation results verify the effectiveness of the control scheme.     

Adaptive neural network dynamic surface control for a class of nonlinear systems with uncertain time delays

This paper presents a solution to tracking control problem for a class of nonlinear systems with unknown parameters and uncertain time-varying delays. A new adaptive neural network(NN) dynamic surface controller(DSC) is developed. Some assumptions on uncertain time delays, which were required to be satisfied in previous works, are removed by introducing a novel indirect neural network algorithm into dynamic surface control framework. Also, the designed controller is independent of the time delays. Moreover,the dynamic compensation terms are introduced to facilitate the controller design. It is shown that the closed-loop tracking error converges to a small neighborhood of zero. Finally, a chaotic circuit system is initially bench tested to show the effectiveness of the proposed method.     

应用信度分配的模糊CMAC实现非线性系统的容错控制

The adaptive fault-tolerant control scheme of dynamic nonlinear system based on the credit assigned fuzzy CMAC neural network is presented. The proposed learning approach uses the learned times of addressed hypercubes as the credibility, the amounts of correcting errors are proportional to the inversion of the learned times of addressed hypercubes. With this idea, the learning speed can indeed be improved. Based on the improved CMAC learning approach and using the sliding control technique, the effective control law reconfiguration strategy is presented. Thesystem stability and performance are analyzed under failure scenarios. The numerical simulation demonstrates the effectiveness of the improved CMAC algorithm and the proposed fault-tolerant controller.     

Toward reverse engineering on secret S-boxes in block ciphers

In this paper, we extend the deterministic learning theory to sampled-data nonlinear systems. Based on the Euler approximate model, the adaptive neural network identifier with a normalized learning algorithm is proposed. It is proven that by properly setting the sampling period, the overall system can be guaranteed to be stable and partial neural network weights can exponentially converge to their optimal values under the satisfaction of the partial persistent excitation (PE) condition. Consequently, locally accurate learning of the nonlinear dynamics can be achieved, and the knowledge can be represented by using constant-weight neural networks. Furthermore, we present a performance analysis for the learning algorithm by developing explicit bounds on the learning rate and accuracy. Several factors that influence learning, including the PE level, the learning gain, and the sampling period, are investigated. Simulation studies are included to demonstrate the effectiveness of the approach.     

Efficient multi-sequence memory with controllable steady-state period and high sequence storage capacity

Sequential information processing, for instance the sequence memory, plays an important role on many functions of brain. In this paper, multi-sequence memory with controllable steady-state period and high sequence storage capacity is proposed. By introducing a novel exponential kernel sampling function and the sampling interval parameter, the steady-state period can be controlled, and the steady-state time steps are equal to the sampling interval parameter. Furthermore, we explained this phenomenon theoretically. Ascribing to the nonlinear function constitution for local field, the conventional Hebbian learning rule with linear outer product method can be improved. Simulation results show that neural network with nonlinear function constitution can effectively increase sequence storage capacity.     

An optimized RBF neural network algorithm based on partial least squares and genetic algorithm for classification of small sample

Radial basis function (RBF) neural network can use linear learning algorithm to complete the work formerly handled by nonlinear learning algorithm, and maintain the high precision of the nonlinear algorithm. However, the results of RBF would be slightly unsatisfactory when dealing with small sample which has higher feature dimension and fewer numbers. Higher feature dimension will influence the design of neural network, and fewer numbers of samples will cause network training incomplete or over-fitted, both of which restrict the recognition precision of the neural network. RBF neural network has some drawbacks, for example, it is hard to determine the numbers, center and width of the hidden layer’s neurons, which constrain the success of training. To solve the above problems, partial least squares (PLS) and genetic algorithm(GA)are introduced into RBF neural network, and better recognition precision will be obtained, because PLS is good at dealing with the small sample data, it can reduce feature dimension and make low-dimensional data more interpretative. In addition, GA can optimize the network architecture, the weights between hidden layer and output layer of the RBF neural network can ease non-complete network training, the way of hybrid coding and simultaneous evolving is adopted, and then an accurate algorithm is established. By these two consecutive optimizations, the RBF neural network classification algorithm based on PLS and GA (PLS-GA-RBF) is proposed, in order to solve some recognition problems caused by small sample. Four experiments and comparisons with other four algorithms are carried out to verify the superiority of the proposed algorithm, and the results indicate a good picture of the PLS-GA-RBF algorithm, the operating efficiency and recognition accuracy are improved substantially. The new small sample classification algorithm is worthy of further promotion.     

Deterministic learning and neural control of a class of nonlinear systems toward improved performance

A deterministic learning theory was recently presented which states that an appropriately designed adaptive neural controller can learn the system internal dynamics while attempting to control a class of nonlinear systems in normal form. In this paper, we further investigate deterministic learning of the class of nonlinear systems with relaxed conditions, and neural control of the class of system toward improved performance. Firstly, without the assumption on the upper bound of the derivative of the unknown affine term, an adaptive neural controller is proposed to achieve stability and tracking of the plant states to that of the reference model. When output tracking is achieved, a partial PE condition is satisfied, and deterministic learning from adaptive neural control of the class of nonlinear systems is implemented without the priori knowledge on the upper bound of the derivative of the affine term. Secondly, by utilizing the obtained knowledge of system dynamics, a neural controller with constant RBF networks embedded is presented, in which the learned knowledge can be effectively exploited to achieve stability and improved control performance. Simulation studies are included to demonstrate the effectiveness of the results.     

基本样条循环神经网络及其非线性建模

提出一种新的基于基本样条逼近的循环神经网络,该网络易于训练且收敛速度快。此外为克服定长学习步长训练速度慢的问题,提出一种用于该网络训练的自适应权值更新算法,给出了学习步长的最优估计。该最优学习步长的选择可用于基本样条循环神经网络的训练以及对非线性系统的建模。     

Learning from neural control of nonlinear systems in normal form

A deterministic learning theory was recently proposed which states that an appropriately designed adaptive neural controller can learn the system internal dynamics while attempting to control a class of simple nonlinear systems. In this paper, we investigate deterministic learning from adaptive neural control (ANC) of a class of nonlinear systems in normal form with unknown affine terms. The existence of the unknown affine terms makes it difficult to achieve learning by using previous methods. To overcome the difficulties, firstly, an extension of a recent result is presented on stability analysis of linear time-varying (LTV) systems. Then, with a state transformation, the closed-loop control system is transformed into a LTV form for which exponential stability can be guaranteed when a partial persistent excitation (PE) condition is satisfied. Accurate approximation of the closed-loop control system dynamics is achieved in a local region along a recurrent orbit of closed-loop signals. Consequently, learning of control system dynamics (i.e. closed-loop identification) from adaptive neural control of nonlinear systems with unknown affine terms is implemented.     

Learning from neural control

One of the amazing successes of biological systems is their ability to "learn by doing" and so adapt to their environment. In this paper, first, a deterministic learning mechanism is presented, by which an appropriately designed adaptive neural controller is capable of learning closed-loop system dynamics during tracking control to a periodic reference orbit. Among various neural network (NN) architectures, the localized radial basis function (RBF) network is employed. A property of persistence of excitation (PE) for RBF networks is established, and a partial PE condition of closed-loop signals, i.e., the PE condition of a regression subvector constructed out of the RBFs along a periodic state trajectory, is proven to be satisfied. Accurate NN approximation for closed-loop system dynamics is achieved in a local region along the periodic state trajectory, and a learning ability is implemented during a closed-loop feedback control process. Second, based on the deterministic learning mechanism, a neural learning control scheme is proposed which can effectively recall and reuse the learned knowledge to achieve closed-loop stability and improved control performance. The significance of this paper is that the presented deterministic learning mechanism and the neural learning control scheme provide elementary components toward the development of a biologically-plausible learning and control methodology. Simulation studies are included to demonstrate the effectiveness of the approach.     

非线性系统模糊神经网络变结构自适应控制

在非线性系统的模糊动力学模型基础上,提出一种模糊神经网络变结构自适应控制器;网络的结构根据非线性系统特性动态构成,基于该网络提出非线性预测器,基于梯度法提出了一种网络参数学习算法,并分析了收敛性及其性质。将网络预测器与参数学习算法相结合,构成自适应控制算法,证明了算法的收敛性。仿真结果证实了算法的有效性。     

混沌系统的混合遗传神经网络控制

在小扰动控制技术基础上,将暂态误差预测方法和遗传算法结合起来,提出了一种混合遗传神经网络控制非线性混沌系统的新方法(简称HyGANN).通过增强学习训练,HyGANN可产生控制混沌状态的小扰动时间序列信号,Henon映射的计算机仿真结果表明,它不仅有效镇定混沌周期1,2等低周期轨道,还可成功将高周期混轨道变成期望周期行为.     

非零和微分博弈系统的事件触发最优跟踪控制（英文）

近年来,对于具有未知动态的非零和微分博弈系统的跟踪问题,已经得到了讨论,然而这些方法是时间触发的,在传输带宽和计算资源有限的环境下并不适用.针对具有未知动态的连续时间非线性非零和微分博弈系统,本文提出了一种基于积分强化学习的事件触发自适应动态规划方法.该策略受梯度下降法和经验重放技术的启发,利用历史和当前数据更新神经网络权值.该方法提高了神经网络权值的收敛速度,消除了一般文献设计中常用的初始容许控制假设.同时,该算法提出了一种易于在线检查的持续激励条件(通常称为PE),避免了传统的不容易检查的持续激励条件.基于李亚普诺夫理论,证明了跟踪误差和评价神经网络估计误差的一致最终有界性.最后,通过一个数值仿真实例验证了该方法的可行性.