A neural model for eye–head–arm coordination

The coordinated movement of the eyes, the head and the arm is an important ability in both animals and humanoid robots. To achieve this, the brain and the robot control system need to be able to perform complex non-linear sensory-motor transformations in the forward and inverse directions between many degrees of freedom. In this article, we apply an omnidirectional basis function neural network to this task. The proposed network can perform 3-D coordinated gaze shifts and 3-D arm reaching movements to a visual target. Particularly, it can perform direct sensory-motor transformations to shift gaze and to execute arm reach movements and can also perform inverse sensory-motor transformations in order to shift gaze to view the hand.     

Adaptive neuro-fuzzy inference system–based model for elevation–surface area–storage interrelationships

In the developing of an optimal operation schedule for dams and reservoirs, reservoir simulation is one of the critical steps that must be taken into consideration. For reservoirs to have more reliable and flexible optimization models, their simulations must be very accurate. However, a major problem with this simulation is the phenomenon of nonlinearity relationships that exist between some parameters of the reservoir. Some of the conventional methods use a linear approach in solving such problems thereby obtaining not very accurate simulation most especially at extreme values, and this greatly influences the efficiency of the model. One method that has been identified as a possible replacement for ANN and other common regression models currently in use for most analysis involving nonlinear cases in hydrology and water resources–related problems is the adaptive neuro-fuzzy inference system (ANFIS). The use of this method and two other different approaches of the ANN method, namely feedforward back-propagation neural network and radial basis function neural network, were adopted in the current study for the simulation of the relationships that exist between elevation, surface area and storage capacity at Langat reservoir system, Malaysia. Also, another model, auto regression (AR), was developed to compare the analysis of the proposed ANFIS and ANN models. The major revelation from this study is that the use of the proposed ANFIS model would ensure a more accurate simulation than the ANN and the classical AR models. The results obtained showed that the simulations obtained through ANFIS were actually more accurate than those of ANN and AR; it is thus concluded that the use of ANFIS method for simulation of reservoir behavior will give better predictions than the use of any new or existing regression models.     

Nonlinear fuzzy model predictive iterative learning control for drum-type boiler–turbine system

Advanced control strategy is necessary to ensure high efficiency and high load-following capability in the operation of modern power plant. Model predictive control (MPC) has been widely used for controlling power plant. Nevertheless, MPC needs to further improve its learning ability especially as power plants are nonlinear under load-cycling operation. Iterative learning control (ILC) and MPC are both popular approaches in industrial process control and optimization. The integration of model-based ILC with a real-time feedback MPC constitutes the model predictive iterative learning control (MPILC). Considering power plant, this paper presents a nonlinear model predictive controller based on iterative learning control (NMPILC). The nonlinear power plant dynamic is described by a fuzzy model which contains local liner models. The resulting NMPILC is constituted based on this fuzzy model. Optimal performance is realized within both the time index and the iterative index. Convergence property has been proven under the fuzzy model. Deep analysis and simulations on a drum-type boiler–turbine system show the effectiveness of the fuzzy-model-based NMPILC     

Teaching–learning based optimization algorithm based fuzzy-PID controller for automatic generation control of multi-area power system

This paper deals with the design of a novel fuzzy proportional–integral–derivative (PID) controller for automatic generation control (AGC) of a two unequal area interconnected thermal system. For the first time teaching–learning based optimization (TLBO) algorithm is applied in this area to obtain the parameters of the proposed fuzzy-PID controller. The design problem is formulated as an optimization problem and TLBO is employed to optimize the parameters of the fuzzy-PID controller. The superiority of proposed approach is demonstrated by comparing the results with some of the recently published approaches such as Lozi map based chaotic optimization algorithm (LCOA), genetic algorithm (GA), pattern search (PS) and simulated algorithm (SA) based PID controller for the same system under study employing the same objective function. It is observed that TLBO optimized fuzzy-PID controller gives better dynamic performance in terms of settling time, overshoot and undershoot in frequency and tie-line power deviation as compared to LCOA, GA, PS and SA based PID controllers. Further, robustness of the system is studied by varying all the system parameters from −50% to +50% in step of 25%. Analysis also reveals that TLBO optimized fuzzy-PID controller gains are quite robust and need not be reset for wide variation in system parameters.     

A chaos and fractal theory based nonlinear dynamical model for China stock market

The China stock market is proved to be a chaotic system, a nonlinear dynamical model is established based on the study of the nonlinear dynamical properties of Shanghai stock, composite index sequence by using chaos and fractal theory. The phase space of the stock sequence is reconstructed and the correlation dimension is analyzed, which indicates that the dynamical system has finite degree of freedom. The nonlinear evolution mechanism is observed and the initial value sensitive characteristic of the system is demonstrated through Lyapunov exponent analysis. Finally, the stock sequence is reconstructed by using finite degree of freedom based fractal interpolation algorithm and gaining reasonably accurate replications. The experimental results indicate that the nonlinear dynamical model is more effective to describe the China stock market than the conventional ＂random walk＂ theory based stochastic models.     

Sound learning–based event detection for acoustic surveillance sensors

Multimedia Tools and Applications - This study proposes an event detection technique for acoustic surveillance that detects emergency situations by using acoustic sensors. Most surveillance systems...     

A self-organization mining based hybrid evolution learning for?TSK-type fuzzy model design

In this paper, a self-organization mining based hybrid evolution (SOME) learning algorithm for designing a TSK-type fuzzy model (TFM) is proposed. In the proposed SOME, group-based symbiotic evolution (GSE) is adopted in which each group in the GSE represents a collection of only one fuzzy rule. The proposed SOME consists of structure learning and parameter learning. In structure learning, the proposed SOME uses a two-step self-organization algorithm to decide the suitable number of rules in a TFM. In parameter learning, the proposed SOME uses the data mining based selection strategy and data mining based crossover strategy to decide groups and parental groups by the data mining algorithm that called frequent pattern growth. Illustrative examples were conducted to verify the performance and applicability of the proposed SOME method.     

SmartGantt – An intelligent system for real time rescheduling based on relational reinforcement learning

With the current trend towards cognitive manufacturing systems to deal with unforeseen events and disturbances that constantly demand real-time repair decisions, learning/reasoning skills and interactive capabilities are important functionalities for rescheduling a shop-floor on the fly taking into account several objectives and goal states. In this work, the automatic generation and update through learning of rescheduling knowledge using simulated transitions of abstract schedule states is proposed. Deictic representations of schedules based on focal points are used to define a repair policy which generates a goal-directed sequence of repair operators to face unplanned events and operational disturbances. An industrial example where rescheduling is needed due to the arrival of a new/rush order, or whenever raw material delay/shortage or machine breakdown events occur are discussed using the SmartGantt prototype for interactive rescheduling in real-time. SmartGantt demonstrates that due date compliance of orders-in-progress, negotiating delivery conditions of new orders and ensuring distributed production control can be dramatically improved by means of relational reinforcement learning and a deictic representation of rescheduling tasks.     

Dempster–Shafer structure based fuzzy logic system for stochastic modeling

The Dempster–Shafer (D–S) theory of evidence is introduced to improve fuzzy inference under the complex stochastic environment. The Dempster–Shafer based fuzzy set (DFS) is first proposed, together with its union and intersection operations, to capture the principal stochastic uncertainties. Then, the fuzzy inference will be modified based on the extensional Dempster rule of combination. This new approach is able to capture the stochastic disturbance acting on fuzzy membership function, and provide a more effective inference under strong stochastic uncertainty. Finally, the numerical simulation and the experimental prediction of the wind speed are conducted to show the potential of the proposed method in stochastic modeling.     

Adaptive control based on IF–THEN rules for grasping force regulation with unknown contact mechanism

An industrial gripping application with unknown contact mechanism is considered as a class of unknown nonlinear discrete-time systems. The control scheme is developed by an adaptive network called multi-input fuzzy rules emulated network (MiFREN) within discrete-time domain. The network structure is directly constructed regarding to IF–THEN rules related to gripper and contact mechanism properties. All adjustable parameters require only the on-line learning phase to improve the closed loop performance. The time varying learning rate is devised for gradient reach with the proof of stability analysis. Furthermore, the estimated sensitivity of system dynamic is directly considered within the parameter adaptation. The experimental system with an industrial parallel grip model WSG-50 validates the performance of the proposed controller.     

A generation alternation model for user–system cooperative evolutionary computation

User?Csystem cooperative evolution (CEUS) is an evolutionary computation (EC) method to optimize quantitative and qualitative criteria. In previous work of CEUS, the whole population update is performed at every generation, and the user observes very few individuals. This paper proposes a generation alternation model designed for CEUS. The proposed model allows a user to find widely varied individuals in addition to the best individuals by replacing just one individual in a population for each generation, and consequently, contributes user??s idea generation by enhancing divergent thinking.     

Pulse neural network–based adaptive iterative learning control for uncertain robots

An adaptive iterative learning control algorithm based on pulse neural network (PNN) is proposed for trajectory tracking of uncertain robot system. Sliding mode variable structure control is used to improve the robustness to disturbance and perturbation, and boundary layer is used to eliminate the chattering of sliding mode control. In the iterative domain, the unknown parameters are tuned and used for part of the controller. Running in parallel, the PNN can perform real-time state estimation for improving the system convergence. We analyze the stability and convergence of this algorithm by using the Lyapunove-like methodology. The simulation results show that the expected control purpose can be achieved using the proposed algorithm.     

Agent–cellular automata model for the dynamic fluctuation of EV traffic and charging demands based on machine learning algorithm

Neural Computing and Applications - Electric vehicles (EV) comprise one of the foremost components of the smart grid and tightly link the power system with the road network. Spatial and temporal...     

Multi-feature sea–land segmentation based on pixel-wise learning for optical remote-sensing imagery

Robust sea–land segmentation in optical remote-sensing images is challenging because of the complex sea–land environment and scene diversity. Here, we propose a novel multi-feature sea–land segmentation method via pixel-wise learning for optical remote-sensing images. Multiple features such as greyscale, local statistical information, edge, texture, and structure are first extracted from each pixel in training images and then used to learn a multi-feature sea–land classifier, which transforms the segmentation issue into pixel-wise binary classification problem. In our approach, a new multi-feature sea–land segmentation algorithm is put forward based on the approximation of Newton method. Experiments on Google-Earth, Venezuelan Remote Sensing Satellite-1 (VRSS-1) and Gaofen-1 images demonstrate that the proposed approach yields more robust and accurate sea–land segmentation results.     

Wavelet based non-parametric NARX models for nonlinear input–output system identification

Wavelet based non-parametric additive NARX models are proposed for nonlinear input–output system identification. By expanding each functional component of the non-parametric NARX model into wavelet multiresolution expansions, the non-parametric estimation problem becomes a linear-in-the-parameters problem, and least-squares-based methods such as the orthogonal forward regression (OFR) approach, coupled with model size determination criteria, can be used to select the model terms and estimate the parameters. Wavelet based additive models, combined with model order determination and variable selection approaches, are capable of handling problems of high dimensionality.     

Numerical strategies for the system of first-order IVPS using the RK–Butcher algorithm

In this article, a new method of analysis for first-order initial-value type ordinary differential equations using the Runge–Kutta (RK)–Butcher algorithm is presented. To illustrate the effectiveness of the RK–Butcher algorithm, 10 problems have been considered and compared with the RK method based on arithmetic mean, and with exact solutions of the problems, and are found to be very accurate. Stability analysis for the first-order initial-value problem (IVP) has been discussed. Error graphs for the first-order IVPs are presented in a graphical form to show the efficiency of this RK–Butcher method. This RK–Butcher algorithm can be easily implemented in a digital computer and the solution can be obtained for any length of time.     

Hybrid online–offline learning to rank using simulated annealing strategy based on dependent click model

Knowledge and Information Systems - Learning to rank (LTR) is the process of constructing a model for ranking documents or objects. It is useful for many applications such as Information retrieval...     

Applying fuzzy arithmetic to the system dynamics for the customer–producer–employment model

This paper presents a system dynamics analysis based on the application of fuzzy arithmetic. Traditional crisp system dynamics observe that some variables/parameters may belong to the uncertain factors. It is necessary to extend the system dynamics to treat also the vague variables/parameters. The evaluation of fuzzy system dynamics may provide the decision-maker information regarding the system's behavior uncertainties. In this paper, the customer–producer–employment model is examined with the fuzzy system dynamics in two types of fuzzy arithmetic, α-cut fuzzy arithmetic and T_ω weakest t-norm operator. Symmetrical and nonsymmetrical triangular fuzzy number (TFN), varied amount of fuzzy inputs’ fuzziness, and length of the system time delay are examined with useful results provided. Particularly, it is revealed that (1) both types of fuzzy arithmetic can provide the steady-state analysis of the system's variables as their counterpart, the crisp arithmetic analysis. (2) The α-cut arithmetic realizes the fuzziness of the model interactive variables fuzzier than that of the T_ω fuzzy arithmetic due to the accumulating phenomenon of fuzziness of the α-cut arithmetic. The fuzzier the inputs, the higher the level and/or oscillation of the cyclically steady or stable pattern of the stability of these variables exhibit with the α-cut arithmetic. (3) The T_ω arithmetic gives a smaller fuzziness and defuzzified values due to the concept that it takes only the maximal fuzziness encountered and calculated in the operation. In this case, the T_ω arithmetic provides more stable (or conversely less sensitive) results to the amount of fuzziness and nonsymmetricity (fuzziness) of input.     

Stable indirect adaptive switching control for fuzzy dynamical systems based on T–S multiple models

A new indirect adaptive switching fuzzy control method for fuzzy dynamical systems, based on Takagi–Sugeno (T–S) multiple models is proposed in this article. Motivated by the fact that indirect adaptive control techniques suffer from poor transient response, especially when the initialisation of the estimation model is highly inaccurate and the region of uncertainty for the plant parameters is very large, we present a fuzzy control method that utilises the advantages of multiple models strategy. The dynamical system is expressed using the T–S method in order to cope with the nonlinearities. T–S adaptive multiple models of the system to be controlled are constructed using different initial estimations for the parameters while one feedback linearisation controller corresponds to each model according to a specified reference model. The controller to be applied is determined at every time instant by the model which best approximates the plant using a switching rule with a suitable performance index. Lyapunov stability theory is used in order to obtain the adaptive law for the multiple models parameters, ensuring the asymptotic stability of the system while a modification in this law keeps the control input away from singularities. Also, by introducing the next best controller logic, we avoid possible infeasibilities in the control signal. Simulation results are presented, indicating the effectiveness and the advantages of the proposed method.