Gangquan Si

Enhancing effectiveness of density-based outlier mining scheme with density-similarity-neighbor-based outlier factor

This paper proposes a density-similarity-neighbor-based outlier mining algorithm for the data preprocess of data mining technique. First, the concept of k-density of an object is presented and the similar density series (SDS) of the object is established based on the changes of the k-density and the neighbors k-densities of the object. Second, the average series cost (ASC) of the object is obtained based on the weighted sum of the distance between the two adjacent objects in SDS of the object. Finally, the density-similarity-neighbor-based outlier factor (DSNOF) of the object is calculated by using both the ASC of the object and the ASC of k-distance neighbors of the object, and the degree of the object being an outlier is indicated by the DSNOF. The experiments are performed on synthetic and real datasets to evaluate the effectiveness and the performance of the proposed algorithm. The experiments results verify that the proposed algorithm has higher quality of outlier mining and do not increase the algorithm complexity.

Neighborhood effective information ratio for hybrid feature subset evaluation and selection

Feature selection has been widely discussed as an important preprocessing step in machine learning and data mining. Evaluation criterion designing arises as a main aspect for constructing feature selection algorithms. In this paper, a new feature evaluation criterion, called the neighborhood effective information ratio (NEIR), is proposed to compute discernibility capability of categorical and numerical features. Based on the evaluation criterion, a general definition of significance of hybrid features is presented. Then a greedy selection algorithm for hybrid feature subsets based on the proposed evaluation criterion is constructed for data classification. We compare the proposed algorithm with other feature selection algorithms. Both theoretical and experimental analysis verifies the effectiveness and the efficiency of the proposed algorithm.

A Hybrid Controller of Self-Optimizing Algorithm and ANFIS for Ball Mill Pulverizing System

For ball mill pulverizing system of the thermal power plant, a hybrid controller of self-optimizing algorithm and adaptive neuro-fuzzy inference system(ANFIS) is proposed. In order to keep the ball mill pulverizing system working at the optimum point all along, the self-optimizing algorithm is presented. The self-optimization algorithm can automatically find out the extreme point and adjust the control set values in time. The adaptive neuro-fuzzy inference system, which integrates the advantages of the neural network and the fuzzy control, uses the learning ability of the neural network to optimize the membership functions and fuzzy logic rules of fuzzy control. Such combined framework makes fuzzy control more systematic and less relying on expert knowledge. Simulations results verify that the controller can control the ball mill pulverizing system effectively and has higher control quality.

Adaptive generalized function matrix projective lag synchronization between fractional-order and integer-order complex networks with delayed coupling and different dimensions

This paper investigates generalized function matrix projective lag synchronization between fractional-order and integer-order complex networks with delayed coupling, non-identical topological structures and different dimensions. Based on Lyapunov stability theory, generalized function matrix projective lag synchronization criteria are derived by using the adaptive control method. In addition, the three-dimensional fractional-order chaotic system and the four-dimensional integer-order hyperchaotic system as the nodes of the drive and the response networks, respectively, are analyzed in detail, and numerical simulation results are presented to illustrate the effectiveness of the theoretical results.

A density-based quantitative attribute partition algorithm for association rule mining on industrial database

Quantitative attribute partition is an important work of association rule mining, which is widely applied in industrial control at present, and the current partition methods are not suitable for the industrial database, which is generally large, high-dimensional and coupling. The paper proposes a density-based quantitative attribute partition algorithm for industrial database. The proposed algorithm uses an improved density-based clustering algorithm to detect the clusters. The clusters are agglomerated to form the new clusters according to the proximity between clusters and the new clusters are projected into the domains of the quantitative attributes. So the fuzzy sets and the membership functions used for partition are determined. We performed the experiments on a test database and a real industrial database. The experiments results verify the proposed algorithm not only can partition the quantitative attributes of industrial database successfully but also has the higher partition effectiveness.

Finite-time generalized function matrix projective lag synchronization of coupled dynamical networks with different dimensions via the double power function nonlinear feedback control method

This paper investigates the problem of finite-time generalized function matrix projective lag synchronization between two different coupled dynamical networks with different dimensions of network nodes. The double power function nonlinear feedback control method is proposed in this paper to guarantee that the state trajectories of the response network converge to the state trajectories of the drive network according to a function matrix in a given finite time. Furthermore, in comparison with the traditional nonlinear feedback control method, the new method improves the synchronization efficiency, and shortens the finite synchronization time. Numerical simulation results are presented to illustrate the effectiveness of this method.

Generalized Tagaki-Sugeno fuzzy rules based prediction model with application to power plant pulverizing system

This paper proposes a generalized Tagaki-Sugeno (TS) fuzzy rules based prediction model and apply it to estimate the pulverizing capability of ball mill pulverizing system of thermal power plant. The proposed method improves the core idea of the adaptive neuro-fuzzy inference system and does not use the neural network to interpret the model structure and the training process. Hence, the proposed model has generalization in a certain extent and could be applied efficiently on a variety of multi-variable and nonlinear dataset. For the proposed method, the Gaussian kernel fuzzy clustering algorithm is firstly used to determine the initial rules, and then the membership functions and the consequent parameters of TS fuzzy rules are tuned by the iterative optimization algorithm that minimizes the measure of the potential of data. The proposed model is performed on the field data obtained from a real thermal power plant and the experiments results verify the effectiveness of the proposed model.

Adaptive Inverse Optimal Control of a Novel Fractional-Order Four-Wing Hyperchaotic System with Uncertain Parameter and Circuitry Implementation

An efficient approach of inverse optimal control and adaptive control is developed for global asymptotic stabilization of a novel fractional-order four-wing hyperchaotic system with uncertain parameter. Based on the inverse optimal control methodology and fractional-order stability theory, a control Lyapunov function (CLF) is constructed and an adaptive state feedback controller is designed to achieve inverse optimal control of a novel fractional-order hyperchaotic system with four-wing attractor. Then, an electronic oscillation circuit is designed to implement the dynamical behaviors of the fractional-order four-wing hyperchaotic system and verify the satisfactory performance of the controller. Comparing with other fractional-order chaos control methods which may have more than one nonlinear state feedback controller, the inverse optimal controller has the advantages of simple structure, high reliability, and less control effort that is required and can be implemented by electronic oscillation circuit.

An Improved Simulated Annealing for Ball Mill Pulverizing System Optimization of Thermal Power Plant

This paper proposes an improved simulated annealing for ball mill pulverizing system optimization of thermal power plan. The proposed algorithm combines the simulated annealing and Tabu search and for the annealing operations, the current calculated solution is evaluated according to the neighborhood of the values in Tabu list. Moreover, some rules for the generation of the neighborhood solution are presented based on the characteristics of the ball mill pulverizing system. The proposed algorithm is performed on the real field data. The results of the experiments verify that the proposed algorithm could determine the optimal values of process variables correctly and has faster convergence speed. In addition, the proposed algorithm has been put into practice and the statistic data show that the working time of ball mill pulverizing system is decreased and the energy consumption would be reduced.

Application of information fusion based on RBF neural networks and fuzzy control to ball mill pulverizing system

Ball mill pulverizing system is a typical multi-input and multi-output (MIMO) system with the characteristics as strong coupling, nonlinearity, large delay and time-varying. The running conditions and the precise mathematic model of the system can not be obtained easily for its complex characteristics, so the conventional control strategy is not effective. An information fusion based method including multi-sensors, data preprocessing, RBF neural networks, fuzzy controller is put forward in this article. By combining multiple sensorspsila information, the RBF neural networks extract features and estimate the running conditions, and then the adjustments of outputs are decided by fuzzy control. By using the fusion method, the running conditionspsila measurements are more accurate and reliable than that using single sensor method and the control system can adapt to the time-varying characteristic of the ball mill pulverizing system. The practical application indicates that the information fusion based control system runs stably and efficiently.