This paper presents a modification of reference-model adaptive control with a layered network of higher-order neural units (HONUs) as adaptive state-feedback controllers. The degree of freedom of such neural controller is deemed here as the number of applied HONUs of a customizable polynomial order and as of their individually customizable input vectors. Furthermore, the control scheme is enhanced because potentially occurring disturbances of controlled variable can result in that the sample-by-sample adapted controllers may tend to adaptively force the plant output to merely follow the reference model output because input data are affected by the error of disturbance, while the overall control loop dynamics would be more accurately adapted if no perturbation occurs or if the reference model is actuated with actually measured variables. Furthermore, the controller weight updates usually involve some step-delayed computations that might be in fact recalculated with the latest updated weights, so the controller learning can be improved.

A recurrent neural network is proposed to deal with the convex optimization problem. By employing a specific nonlinear unit, the proposed neural network is proved to be convergent to the optimal solution in finite time, which increases the computation efficiency dramatically. Compared with most of existing stability conditions, i.e., asymptotical stability and exponential stability, the obtained finite-time stability result is more attractive, and therefore could be considered as a useful supplement to the current literature. In addition, a switching structure is suggested to further speed up the neural network convergence. Moreover, by using the penalty function method, the proposed neural network can be extended straightforwardly to solving the constrained optimization problem. Finally, the satisfactory performance of the proposed approach is illustrated by two simulation examples.

Design of an adaptive controller for complex dynamic systems is a big challenge faced by the researchers. In this paper, we propose a novel method for the design of an error- based robust adaptive controller to make the system response reasonably fast with no overshoot. Here the control action is designed by introducing the notion of `error-based adaptive controller' (EB-AC). In the design of this feedback adaptive controller, parameters of the controller are designed as a function of the system error. For example, the position feedback parameter Kp(e, t), which controls the bandwidth of the system as well as the dynamic response, is a function of the system error e(t). In the design of the position feedback parameter Kp(e, t), for large error Kp(e, t) is kept large, thus increasing the bandwidth of the system which yields a fast response, whereas for decreasing errors, Kp(e, t) is continuously decreased to a small value. Thus, during the dynamic response of the system, the bandwidth of the system is continuously controlled by the system error e(t). Similarly, the velocity feedback parameter Kv(e, t) which controls the damping of the system is kept very small for large errors, and continuously increased to a large value for decreasing value of error. Hence, in the design of the proposed adaptive controller, the position feedback Kp(e, t) and the velocity feedback Kv(e, t) are formulated as a function of the system error, and this approach for formulating the adaptive controller yields a very fast response with no overshoot. In this paper, we present an error-based robust adaptive control design methodology for a linear system.

In this paper, we propose a new computer aided diagnosis method of pulmonary nodules in X-ray CT images to reduce false positive (FP) rate under high true positive (TP) rate conditions. An essential core of the method is to extract and combine two novel and effective features from the raw CT images: One is orientation features of nodules in a region of interest (ROI) extracted by a Gabor filter, while the other is variation of CT values of the ROI in the direction along body axis. By using the extracted features, a principal component analysis technic and any pattern recognition technics such as neural network approaches can then used to discriminate between nodule and non-nodule images. Simulation results show that discrimination performance using the proposed features is extremely improved compared to that of the conventional method.

A recurrent neural network for the optimal control of a group of interconnected dynamic systems is presented in this paper. On the basis of decomposition and coordination strategy for interconnected dynamic systems, the proposed neural network has a two-level hierarchical structure: several local optimization subnetworks at the lower level and one coordination subnetwork at the upper level. A goal-coordination method is used to coordinate the interactions between the subsystems. By nesting the dynamic equations of the subsystems into their corresponding local optimization subnetworks, the number of dimensions of the neural network can be reduced significantly. Furthermore, the subnetworks at both the lower and upper levels can work concurrently. Therefore, the computation efficiency, in comparison with the consecutive executions of numerical algorithms on digital computers, is increased dramatically. The proposed method is extended to the case where the control inputs of the subsystems are bounded. The stability analysis shows that the proposed neural network is asymptotically stable. Finally, an example is presented which demonstrates the satisfactory performance of the neural network.

In this paper, we analyze neural spike dynamics of a double feedback neural unit (DFNU). An essential emphasis of the analysis is on use of the DFNU's simple formulations that can provide quantitative analytic results. Comparing dynamics of Hodgkin-Huxley model to that of the DFNU, it is shown that dynamics of the DFNU is also physiologically plausible under a condition. The results suggest that high-frequency firings are relatively appropriate for a neural informational carrier due to the reliability and robustness to noisy inputs. To realize such reliable spike communication, we improved the DFNU's performance by using extra noisy inputs with appropriate amplitudes. Simulation studies show that there is optimal region of the amplitude that makes the DFNU possess the noise-enhanced reliable communication ability as similar to stochastic resonance phenomena.

In this paper, a new type of six-degrees of freedom (DOF) flexible parallel mechanism (FPM) is presented. This type of parallel mechanism possesses several favorable properties: (1) its number of DOFs is independent of the number of serial chains which make up the mechanism; (2) it has no kinematical singularities; (3) it is designed to move on rails, and therefore its workspace is much larger than that of a conventional parallel manipulator; and (4) without changing the number of DOFs and the kinematics of the mechanisms, the number of the serial chains can be reconfigured according to the needs of the tasks. These properties make the mechanism very preferable in practice, especially for such tasks as joining huge ship blocks, in which the manipulated objects vary dramatically both in weights and dimensions. Furthermore, the mechanism can be used as either a fully actuated system or an underactuated system. In the fully actuated case, the mechanism has six DOF motion capabilities and manipulation capabilities. However, in the underactuated case, the mechanism still has six DOF motion capabilities, but it has only five DOF manipulation capabilities. In this paper, both the inverse and forward kinematics are studied and expressed in a closed form. The workspace and singularity analysis of the mechanism are also presented. An example is presented to illustrate how to calculate the kinematics of the mechanism in both fully-actuated and underactuated cases. Finally, an application of such a mechanism to manufacturing industry is introduced.

We propose a self-organizing neural structure with dynamic and spatial changing weights for a feature space representation of concept formation. An essential core of this self-organization is based on an unsupervised learning with incomplete information for the dynamic changing and an extended Hebbian rule for the spatial changing. A concept formation problem requires the neural network to acquire the complete feature space structure of a concept information using an incomplete observation of the concept. The connection structure or self-organizing network can store with the information structure by using the two rules. The Hebbian rule can create a necessary connection corresponding to a feature space substructure of the complete information. On the other hand, unsupervised learning can delete unnecessary connections. Finally concept formation ability of the proposed neural network is proven under some conditions.

Visual servo control is needed for realizing automatic bio-micromanipulation and increasing the accuracy of micromanipulator. A dual-hand micromanipulation system with micro-visual-servo was developed for automating cell manipulation in biotechnology. This paper presents a complete solution for micro-visual-servo of the system, involving micro-visual-servo architecture, control law, modeling of image Jacobian and recognition algorithms of micro objects. The experimental results of micro-circle trajectory tracking and automatic transgenic operation verify the effectiveness of the micro-visual-servo solution.

The conventional dynamic backpropagation (DBP) algorithm proposed by Pineda does not necessarily imply the stability of the dynamic neural model in the sense of Lyapunov during a dynamic weight learning process. A difficulty with the DBP learning process is thus associated with the stability of the equilibrium points which have to be checked by simulating the set of dynamic equations, or else by verifying the stability conditions, after the learning has been completed. To avoid unstable phenomenon during the learning process, two new learning schemes, called the multiplier and constrained learning rate algorithms, are proposed in this paper to provide stable adaptive updating processes for both the synaptic and somatic parameters of the network. Based on the explicit stability conditions, in the multiplier method these conditions are introduced into the iterative error index, and the new updating formulations contain a set of inequality constraints. In the constrained learning rate algorithm, the learning rate is updated at each iterative instant by an equation derived using the stability conditions. With these stable DBP algorithms, any analog target pattern may be implemented by a steady output vector which is a nonlinear vector function of the stable equilibrium point. The applicability of the approaches presented is illustrated through both analog and binary pattern storage examples.

The emergence of the theory of dynamic neural computing has made it possible to develop neural learning and adaptive schemes that can be used to obtain feasible solutions to complex control problems, such as inverse kinematic control for robotic systems. In this paper, such a neural learning scheme using a multilayered dynamic neural network (MDNN) is proposed. The basic dynamic computing element of MDNN is a dynamic neural unit (DNU) developed in this paper. The learning and adaptive capabilities of DNU can be used for developing complex dynamic structures. In this paper, we have used DNU for developing a MDNN for the inverse kinematic control of a two-link robot. The validity of the proposed scheme is demonstrated through computer simulation studies.

A local navigation technique with obstacle avoidance, called adaptive navigation, is proposed for mobile robots in which the dynamics of the robot are taken into consideration. The only information needed about the local environment is the distance between the robot and the obstacles in three specified directions. The navigation law is a first-order differential equation and navigation to the goal and obstacle avoidance are achieved by switching the direction angle of the robot. The effectiveness of the technique is demonstrated by means of simulation examples.

A method for estimating the equilibrium capacity of a general class of analog feedback neural networks is presented in this brief paper. Some explicit relationships between upper bound of the number of possible stable equilibria and the network parameters such as self-feedback coefficients, weights, and gains of a feedback neural network are obtained. Increasing the equilibrium capacity using multimodal sigmoidal functions is also discussed. Some examples are provided to demonstrate the effectiveness of the analytical results presented.

Some globally asymptotical stability criteria for the equilibrium states of a general class of discrete-time dynamic neural networks with continuous states are presented using a diagonal Lyapunov function approach. The neural networks are assumed to have the asymmetrical weight matrices throughout the paper. The resulting criteria are described by the diagonal stability of some matrices associated with the network parameters. Some novel stability conditions represented by either the existence of the positive diagonal solutions of the Lyapunov equations or some inequalities are given. Using the equivalence between the diagonal stability and the Schur stability for a nonnegative matrix, some simplified global stability conditions are also presented. Finally, some examples are provided for demonstrating the effectiveness of the global stability conditions presented.

The problem of learning control for a general class of discrete-time nonlinear systems is addressed in this paper using multilayered neural networks (MNNs) with feedforward connections. A suitable extension of the concept of input-output linearization of discrete-time nonlinear systems is used to develop the control schemes for both output tracking and model reference control purposes. The ability of MNNs to model arbitrary nonlinear functions is incorporated to approximate the unknown nonlinear input-output relationship and its inverse using a new weight learning algorithm. In order to overcome the difficulties associated with simultaneous online identification and control in neural networks based adaptive control systems, the new learning control architectures are developed for both adaptive tracking and adaptive model reference control systems with online identification and control ability. The potentials of the proposed methods are demonstrated by simulation examples.>

An analysis of the absolute stability for a general class of discrete-time recurrent neural networks (RNN's) is presented. A discrete-time model of RNN's is represented by a set of nonlinear difference equations. Some sufficient conditions for the absolute stability are derived using Ostrowski's theorem and the similarity transformation approach. For a given RNN model, these conditions are determined by the synaptic weight matrix of the network. The results reported in this paper need fewer constraints on the weight matrix and the model than in previously published studies.

Two coupled nonlinear first-order systems whose dynamic behavior reflects the neural states exhibited by a spatially localized population of excitatory and inhibitory nerve cells are described. The dynamics of each constituent neural subpopulation represents a fundamental neural information processing element (PE) of a complex neural system. Phase plane analysis is used in this paper to show how such antagonistic positive acting (excitatory) and negative acting (inhibitory) PEs can generate diverse steady-state and temporal phenomena when the nonlinear system parameters of the PEs are altered. By modifying a selected set of parameters, it is possible to program the positive and negative PEs to exhibit various dynamic attributes such as multiple stable states, transient response behavior and limit-cycle oscillations. These dynamic attributes may be used to perform a variety of useful computational tasks in signal processing and vision systems such as short-term memory (STM), temporal filtering (TF) and pulse frequency modulation (PFM). Computer simulations are presented throughout this paper in order to illustrate these dynamic attributes.

A multitask neural network is proposed as a plausible visual information processor for performing a variety of real-time operations associated with the early stages of vision. The computational role performed by the processor, named the positive-negative (PN) neural processor, emulates the spatiotemporal information processing capabilities of certain neural activity fields found along the human visual pathway. The state-space model of this visual information processor corresponds to a bilayered two-dimensional array of densely interconnected nonlinear processing elements (PE's). An individual PE represents the neural activity exhibited by a spatially localized subpopulation of excitatory or inhibitory nerve cells. Each PE may receive inputs from an external signal space as well as from itself and the neighboring PE's within the network. The information embedded in the external input data which originates from a video camera or another processor is extracted by the feedforward subnet. The feedback subnet of the PN neural processor generates a variety of transient and steady-state activities. Their various computational roles are applicable to gray level, edge, texture, or color information processing. Computer simulations involving gray level image processing are used to illustrate the versatility of the PN neural processor architecture for machine vision system design.

The theory of discriminant operators, a robust mathematical framework for the extraction of transitions, gradients, and edges from signals and images in the presence of noise, is introduced. The discriminant operator, which is briefly introduced, is an attempt to emulate the robust information extraction properties of the receptive fields found in the retina and the visual channel leading to the visual cortex. The discriminant operator (or the Θ operator) has differential properties with inherent aggregation. The differentiation properties help to extract the transition and edge information while the aggregation operation provides noise immunity. Also introduced is the differential operator, and then the Θ operator is used in both the continuous and discrete domains. The robustness of the theory is illustrated by means of a set of simulation studies on one-dimensional signals. This theory will be used in the development of robust algorithms for robotics and feedback control problems

A machine that can see, perceive and recognize images and objects will be an integral part of the future, autonomous intelligent robotic systems for applications to industrial automation, remote sensing, medical imaging, and space exploration [1-10]. Basic research on and development of such a machine pose a scientific challenge to both the academic and industrial communities.

A dynamic neural network with neural computing units that exhibit hysteresis phenomena is proposed as a mechanism for visual memory. The neural network, named the PN-processor, is loosely based on a mathematical theory proposed by Wilson and Cowan to describe the functional dynamics of conical nervous tissue. The individual neural computing units of the network are programmed to exhibit localized hysteresis phenomena. This neural network structure is capable of storing visual information without physical changes to its synaptic connections. External stimuli move the network's neural activity around a high-dimensional phase space of state attractors until the overall response is stabilized. Once stabilized, the response remains unperturbed by weak or familiar stimuli and is changed only by a sufficiently strong new input. In this paper we briefly describe several aspects of this type of visual memory.

In the analysis and design of expert control systems, when linguistic types of control algorithms are used, it is important to model a physical system using fuzzy theoretic approach. The modeling, analysis, and synthesis of multivariable structures of fuzzy control systems is discussed. First, a multivariable open-loop fuzzy system is described by a set of fuzzy equations. Then a block diagram of the system consisting of functional blocks and intersectional blocks is presented. Series, parallel, and series-parallel connections of multivariable open-loop systems are also analyzed. Multivariable feedback structures are depicted by a set of fuzzy equations and block diagrams. Then the theory of multivariable-multilevel structures is presented. An example of two interconnected tank liquid-level system is given.

Controllability properties of fuzzy dynamic systems are presented using graph representation. The indices of stability and controllability are introduced. A necessary condition for controllability is formulated. An algorithm for finding acceptable controls, based on a Liapunov-like approach, is presented. A numerical example is given.

A dynamic system is stable if its total energy decreases monotonically until a state of equilibrium is reached. The stability of a fuzzy dynamic system is based on a generalization of this notion. If a free fuzzy dynamic system has an asymptotically stable equilibrium state, the stored energy of the system displaced within the domain of attraction decays with time until it assumes its minimum value at the equilibrium state. An energy `measure' of a fuzzy dynamic system is proposed, and `an energy function' is formulated. A heuristic algorithm for determining the stability of the fuzzy system is also proposed. To illustrate the applications of the algorithm, some numerical examples are given.

In this paper an on-line multilevel suboptimal control procedure is developed for large-scale, time-invariant, and discrete linear systems with quadratic performance criterion, and measurement disturbances having arbitrary statistical distributions. The on-line coordination procedure as reported here employs an upper-level intervention through interaction estimates and goal modifications in order to modify lower-level subsystem optimization problems. Their individual controls are applied to the system and the resulting output is estimated by Kalman Type filter and performance information is fedback to the upper level unit. A reference control approach is used to up-date intervention variables over a sequence of time instants to improve on the control being applied to the process. In the computer simulation of the process, it was found that the actual overall cost for systems using this on-line multilevel control procedure is within ±10% of the cost resulting from a conventional overall optimal control policy.