查看更多>>摘要：Semantic web technology plays an increasing role in performing smart home applied programs and it has led to improve semantic interoperability among different systems. However, classical ontologies fail to illustrate ambiguous, incomplete, and uncertain knowledge often available in the real world. On the other hand, the air quality assessment carried out to determine “the degree of pollution” lacks accurately specified boundaries; therefore, the conventional approach based on classic ontology cannot extract real-valued memberships and consequently fails to support ambiguous, incomplete, and uncertain knowledge. Integrating semantic web of things technology (SWOT) and type-2 fuzzy logic improves the capability of semantic reasoning to retrieve query information. Annotation of sensor-generated data and the ability to infer and represent knowledge based on type-2 fuzzy logic are extremely essential when the available data are ambiguous and uncertain. Hence, in this paper, we have provided a framework to build an IoT-based home air quality assessment system by using type-2 fuzzy ontology so that smart home systems can make a decision and control appropriately based on predefined rules by employing the provided semantic reasoning.
查看更多>>摘要：This work is developed to discuss the feasibility and efficiency of adopting Artificial Intelligence (AI) Deep Learning in smart city scenarios. A traffic flow prediction model is constructed based on the Deep Belief Network (DBN) algorithm. The target road section and its historical traffic flow data in Tianjin are collected and pre-processed. Then, several Restricted Boltzmann Machines (RBM) are stacked together to form a DBN, which is trained as a generative model. Finally, its performance is analyzed by the simulation experiment. The algorithm model proposed is compared with Neuro Fuzzy C-Means (FCM) model, Deep Learning Architecture (DLA), and Convolutional Neural Network (CNN) model. The results show that the Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) of the algorithm model proposed are 4.42%, 6.21%, and 8.03%, respectively. Its prediction accuracy is significantly higher than that of the other three algorithms. In addition, the algorithm can effectively suppress the spread of congestion in the smart city, achieving timely evacuation of traffic congestion. In short, the constructed Deep Learning-based traffic flow prediction model has a high-precision prediction effect and traffic congestion evacuation performance, which can provide experimental references for the later construction of smart cities.
查看更多>>摘要：Multipath transmission routing in dynamic wireless sensor network deployed with movable nodes is a multi-objective optimization issue with multi-parameter constraints. Traditional strategies have the deficiencies such as high computational complexity, long solving time, difficulty in obtaining optimal value and easy to fall into local solution especially in the scenario of large quantity of movable sensor nodes being deployed. Therefore, an improved immune particle swarm optimization based multipath transmission strategy (IPSMT) is presented. It contains three parts: improved immune particle swarm algorithm (IIPSO), IPSMT and fault tolerance strategy in multipath transmission routing (FTMT). IIPSO aims to improve the diversity of particle population and accelerate the convergence of multipath establishment respectively through concentration regulation mechanism and immune vaccine operation improvement. IPSMT and MTFT is to establish and optimize multiple transmission paths by considering the movable nodes’ moving distance and energy consumption quality by multi-objective optimization method. Network fault tolerance model has been established to analyze the algorithm's convergence, energy consumption balance, network fault tolerance and time complexity. Through the multi-objective optimization simulation and analysis of multipath establishment comparing with the related works, IPSMT shows good global searching ability and convergence performance as well as diversity of solution population to realize the optimization of multipath transmission routing. The whole network is proved to have good performance of transmission stability and fault tolerance.
查看更多>>摘要：Investigating Bitcoin price forecasting has attracted academic attention recently. However, despite some studies on potential economic determinants of Bitcoin price, a consensus on the best predictors is not reached yet. This paper investigates different predictors from various markets including Gold, Oil, S&P500, VIX, USDI, Ether and Ripple as well as Bitcoin historical price in predicting one-step-ahead Bitcoin returns. We propose a two-stage forecasting that comprises discrete wavelet transform as the decomposition method and a deep long short-term memory network as the forecasting algorithm. Beside analyzing forecasting for both univariate and multivariate regression, we design a simulated trading system to put the forecasts into practice and analyze the economic profitability of the predictors. In addition, we shed light on the black box method by implementing sensitivity analysis. To investigate the predictors’ efficacy through time and consider the effects of early 2018 price spike, the dataset is split into two periods: (1) prior to and including the spike and (2) after the spike. According to the experiments, it is hard to choose one predictor over the other in the first period. However, in the second period, Gold and Oil show the highest statistical accuracy, while S&P500 is the most profit-making predictor.
查看更多>>摘要：Classifiers based on evidential reasoning (ER) rule can well handle the uncertainty in the mapping relationship between input attributes and output classes. To avoid the number of model parameters increasing with the growing number of input attributes, this paper proposes a classification model based on attribute vectorization and evidential reasoning (AV-ER). Firstly, different input attributes are combined into attribute vectors by using principal component analysis (PCA). Then, all training samples are casted into reference attribute vectors, and the reference evidence matrix (REM) is generated by likelihood function normalization. After that, all pieces of activated evidence are fused through ER theory to generate the final classification decision. In the fusion process, parameters of the initial classification model are optimized by genetic algorithm (GA), and Akaike information criterion (AIC) is used to evaluate the model performance comprehensively considering the model complexity and classification accuracy. Finally, typical UCI benchmark datasets are applied to verify the proposed AV-ER classification model, and the results indicate that the classification performance of the AV-ER model is satisfying while the number of the model parameters decrease obviously as well.
查看更多>>摘要：In Markov games, accurately detecting opponent policies and reusing optimal response policies is still a challenging problem. Most previous works assume that opponents switch their policies infrequently only at the end of an episode. However, the opponents may change their policies at high-frequency or even within an episode. Besides, the agent may achieve inconsistent optimal returns because of different opponent behaviors, which brings greater challenges to policy detection. This paper studies how to deal with the non-stationary opponent with abrupt policy changes through accurate policy detection and direct policy reuse. Specifically, we propose a context-aware Bayesian policy reuse (CABPR) algorithm to accurately identify and track the multi-strategic opponent. To continuously infer the opponent policy, an intra-episode belief is introduced taking advantage of opponent models. Within an episode, an inter-episode belief using Bayesian inference and the intra-episode belief are jointly used to detect the opponent type based on its behaviors and episodic rewards. Then the agent reuses the best response policies accordingly. We demonstrate the advantages of the proposed algorithm over several state-of-the-art algorithms in terms of episodic rewards, accumulated rewards, and detection accuracy in four competitive scenarios.
查看更多>>摘要：This paper proposes two ensemble strategies for the backtracking search algorithm (BSA). The first one is an ensemble of two sets of evolutionary operators that balances exploration and exploitation abilities. The second one is an ensemble of values for each parameter associated with the evolutionary operators. The second strategy provides diverse search moves with various search step lengths that are essential for searching different search landscapes. In addition to the ensemble strategies, another strategy is used to reinitialize specific individuals of the population to escape from local optima. Sixteen variants of the BSA are built based on different combinations of these strategies or their modified versions. The best variant for solving 29 problems of CEC2017 test suite is statistically compared with nineteen state-of-the-art algorithms. The results confirm its superiority to all the considered algorithms. Remarkably, according to the Wilcoxon rank-sum test with a significance level of 0.05, it is better than others for solving at least 20 and 18 functions with 30 and 50 dimensions, respectively. Furthermore, it is applied to five engineering design optimization problems. Its solutions are at least as well as or better than those obtained by the best existing algorithms in the literature for three problems.
查看更多>>摘要：Many engineering optimization problems are characterized by large scale and complex constraints. High optimization efficiency and reliable constraint handling are two major challenges. The traditional optimization methods hard to obtain practical and feasible solutions in reasonable time. To get better solutions and enhance the global search capability, a constrained cooperative adaptive multi-population differential evolutionary (CCAM-PDE) algorithm is proposed in this paper. The main contributions of this paper are in three aspects. First, a hyperspace dynamic constraint handling region between feasible region and infeasible region is proposed. Second, according to the feasible rate of population, a “one to one” or “one to many” subpopulation generation scheme is adopted for improving the global searching ability. Third, the selection operation of differential evolution algorithm is replaced by the elimination mechanism through the constraint handling technology. Eight economic load dispatch problems and CEC2017 Benchmark test functions are used to testify the performance of the CCAM-PDE algorithm. The experimental results shown that the CCAM-PDE algorithm has a strong constraint-handling efficiency and better global searching ability, its search accuracy and the speed of convergence against the other state-of-the-art algorithms.
查看更多>>摘要：This paper presents a design application of Interval Type-2 (IT2) Takagi–Sugeno (T–S)? Fuzzy Model Based (FMB) control system for generic aircraft. The IT2 T–S FMB flight control system consists of an IT2 T–S fuzzy model and a fuzzy controller connected in a closed-loop. The IT2 T–S? fuzzy model is obtained by linearizing the nonlinear aircraft dynamics about various representative points (equilibrium points) of the flight envelope with some fuzzy rules. The aircraft flight envelope parameters, i.e., operating altitude and aircraft speed are characterized as premise parameters and elements of stability and control derivative matrix are identified as consequent parameters of fuzzy model. Longitudinal dynamics of X-29A research aircraft is selected for design of IT2 T–S FMB control system. To achieve the optimum design flexibility, an imperfectly matched premise and membership function dependent (MFD) stability analysis is considered. In MFD stability conditions, the information of flight envelope parameters is included to capture the nonlinearity pertaining to variation in flight conditions. The closed-loop response of IT2 T–S FMB controller is presented at trim equilibrium points by taking four initial angle of attack flight conditions. The performance analysis of designed controller is also presented and discussed in comparison with Fuzzy LQR and LQR based optimal controllers. The simulation results reveal that proposed IT2 T–S FMB controller not only stabilizes the aircraft dynamics but also provides improved transient performance as compared with Fuzzy LQR and LQR based optimal controllers. This demonstrates the utility of IT2 T–S FMB control system for aircraft/ UAV's related application.
查看更多>>摘要：Radial basis function network-based autoregressive with exogenous input (RBF-ARX) models are useful in nonlinear system modelling and prediction. The identification of RBF-ARX models includes optimization of the (model lags, number of hidden nodes and state vector) and the parameters of the model. Previous works have usually ignored optimizations of the model's architecture. In this paper, the RBF-ARX architecture, which includes the selection of lags, number of nodes of the RBF network, lag orders and state vector, is encoded into a chromosome and is evolved simultaneously by a genetic algorithm (GA). This combines the advantages of the GA and the variable projection (VP) method to automatically generate a parsimonious RBF-ARX model with a high generalization performance. The highly efficient VP algorithm is used as a local search strategy to accelerate the convergence of the optimization. The experimental results demonstrate the effectiveness of the proposed method.
NSTL
Elsevier