Cluster analysis algorithm based on key data integration for cloud computingDong-rui, Li
doi: 10.1504/IJRIS.2017.090041pmid: N/A
In order to improve scheduling efficiency and a kind of cloud task scheduling algorithm of improved fuzzy cluster has been proposed. Firstly, cloud task scheduling algorithm of improved fuzzy cluster has been introduced, which mainly uses fuzzy FCM algorithm to complete resource cluster to three resource sets including computing type, storage type and bandwidth type in the context of using parallel processing to ensure the efficiency. The resource of cluster set with the longest time in completion will be liberated from the busy schedule to improve the utilisation ratio of resources, ensure load balance, reduce execution costs and enhance customer satisfaction; secondly, tasks have been allocated to each cluster through Min-Min heuristic algorithm and the results have been adjusted according to set threshold to obtain the better scheduling results. The experimental results show that the proposed algorithm is superior to the traditional algorithm without cluster in terms of execution time.
Parallel cluster analysis of multi city congestion based on spatial temporal potential correction in mobile phone APPJin, Shi
doi: 10.1504/IJRIS.2017.090040pmid: N/A
To improve urban road congestion detection and governance efficiency of city, this paper puts forward spatial-temporal analysis method for urban road congestion of multiprocessor parallel clustering based on potential field correction, establishes spatial-temporal model of urban road congestion based on temporal data of GIS four-dimensional space and constructs multiprocessor parallel clustering method by utilising potential field correction method and designs parallel multiprocessing detection method of urban road congestion of distance matrix, neighbourhood radius and density function. Experimental result verifies effectiveness of method mentioned and it shows that method mentioned can realise fast and effective detection analysis of urban road congestion and provides data support for urban road congestion management.
Cloud computing resource scheduling and leasing algorithm based on extreme price filterXiaoming, Liu; Zonghui, Li; Junjie, Wang; Xujiang, Xu
doi: 10.1504/IJRIS.2017.090036pmid: N/A
A cloud computing virtual resource leasing algorithm considering about extreme price filtering has been proposed in this paper, which realises optimal selection for handling leasing price for task virtual machine. First, three-function module cloud composed of virtual resource provider, cloud service provider and final user has been adopted to calculate environment and list calculation target for virtual resource leasing profit; second, distribution and task urgency of price has been fully considered. For weakly stationary price sequence, outlier detection method has been adopted to make extreme price filtering. At the same time, weak equilibrium operator has been designed, exponential function has been used to control the overall shape of curve, non-uniform mutation operator has been used to make local operating adjusting, realise effective prediction for price in the future and optimal selection and make optimal selection for handling leasing price of task virtual machine.
Training project arrangement for tennis athletes based on BP neural network modelHao, Wang; Hong, Yuan
doi: 10.1504/IJRIS.2017.090037pmid: N/A
In order to improve the prediction accuracy of athlete's tennis training effect, a kind of prediction method for athlete's tennis training effect of RBF (boundary value constraints radial basis function, BVC-RBF) neural network with boundary value constraints is proposed. Firstly, the internal and external factors that influence the athlete's tennis training effect is analysed, and the influence models of 12 indexes including quantitative load heart rate and body fat percentage are predicted and analysed emphatically; secondly, the RBF neural network algorithm with boundary value constraints is built to solve the boundary value constraint equation, so as to obtain the compensation function, and the least square method is used to train traditional RBF neural network, which achieves the improvement of prediction algorithm performance; finally, the simulation experiment shows that the proposed method provides higher prediction accuracy, which has a certain guiding value for tennis training.
Two echelon supply chain model of agricultural products based on stochastic fuzzy process of cost demandJie, Gao
doi: 10.1504/IJRIS.2017.090042pmid: N/A
Stochastic fuzziness existed in supply chain process has important influence on inventory maintenance and decision processing of normal system operation, especially under the situation of coupling existed in two echelon supply chain, this influence will be amplified to some extent. First, for the studied two echelon supply chain objects, joint cost has been taken as target to make optimal model design; second, considering about the stochastic fuzziness existed in two echelon supply chain, the demand rate of market for products as well as supply and distribution time have been selected as main research parameters, two echelon supply stochastic demand model based on triangular fuzzy number has been constructed and then model solving process has been deduced according to the characteristics of design model; last, verification for the performance of proposed model has been made based on parameter influence experiment and horizontal contrast.
Parameter tuning of boiler thermal process based on SVM neural net optimisationPeng, He
doi: 10.1504/IJRIS.2017.090043pmid: N/A
Because of complex characteristics, such as multivariable coupling in boiler thermal process of circulating fluid bed, parameter turning, there is relatively large difficulty in automatic accurate control so that a kind of self-adaptive controller algorithm is put forward. Fuse fuzzy control and equivalent method of BP neural net usage structure to fuzzy BP neural net and bring in weight of genetic algorithm optimisation BP neural net by aiming at defects, such long convergence time of neutral net and realise self-adaptive accuracy control to boiler thermal process of circulating fluid bed by feed-forward compensation decoupling device. It is showed from experiment results that the algorithm can adapt to working condition of variable parameter boiler thermal process of circulating fluid bed and it has realised uncoupling of bed temperature and main steam pressure.
Design of human-computer interaction interface considering user friendlinessHong, Chen
doi: 10.1504/IJRIS.2017.090044pmid: N/A
Analysis and modelling for interactions on self-service terminal interface have been made in this paper based on distributed cognition theory, which is to confirm the relationship between interactions and information presentation in human-computer interaction and propose interface interaction design method of self-service terminal based on user cognitive ability. With adoption of this method, designers need to make research on users and identify the user group that needs to be taken care of firstly, and then make analysis of cognitive ability, set up user cognitive load model, describe the interactive behaviour of users, confirm basic interaction frame, and then establish interaction design matrix with universal usability design model and propose interactive design program. We take hotel self-service terminal as example, adopt this method to make design program and then verify the effectiveness of the proposed design method through making comparisons with design program formed by traditional method. This interactive design method can help designers develop self-service terminal interface that is suit-able for people to understand and use, decrease the cognitive load of users and meet the diversified demands of self-service terminal users for cognition.
Complex electromechanical system condition monitoring based on improved particle swarm optimisation RBF for audio visual fusionJiangwei, Xu; Tiejun, Li; Liang, Zhang
doi: 10.1504/IJRIS.2017.090045pmid: N/A
To improve transient stability of multi-generator power system, continuous high-order sliding mode excitation control strategy is put forward. Power angle deviation of each generator is the variable of sliding mode. Nonlinear and uncertain high-order sliding mode control of multi-generator power system is transferred into limited time stability problem of uncertain integral chain system. Limited time convergence of system condition is realised and overcoming uncertainty, such as unmodeled dynamics of system, measuring error and external disturbance, etc., through combination of controller and geometric homogeneous continuous control law and second-order sliding mode super-twisting algorithm. Power angle differential with precise robust differentiator is observed. Limited time stability of closed-loop system theoretically is analysed and verified. High-order sliding mode excitation controller designed can keep voltage stability at generator terminal and improves transient stability of power system effectively. Simulation result aimed at three-generator system verifies effectiveness of control method mentioned.
Algorithm of key data ensemble clustering and approximate analysis in cloud computingWendong, Xia; Yuanfeng, Liu; Deli, Chen
doi: 10.1504/IJRIS.2017.090038pmid: N/A
One collaborative data fusion recommendation algorithm is SFS-TOPSIS based on customer satisfaction degree. First, it starts from calculation efficiency and recommended precision angle that improves service recommendation algorithm, makes real-time updating and algorithm improvement for it by combining time-varying weight method TOPSIS fusion algorithm and designs a collaborative data fusion recommendation algorithm based on customer satisfaction degree. Second, for the problem of inadequate definition of traditional similarity for resolution, improvements have been made based on user evaluation confidence. Last, time-varying weight method has been adopted to improve standard TOPSIS fusion, improve time-varying attribute of TOPSIS decision fusion and realise effective attribute fusion of user similarity data; through making simulation comparison on two standard testing sets of MovieLens and BookCrossing, it indicates that the service recommendation performance of SFS-TOPSIS is superior. The proposed SFS-TOPSIS algorithm can improve service recommendation accuracy effectively and it is with certain application value.
Prediction and optimal allocation of agricultural non-point source pollution based on chaos theoryChenyang, Li; Na, Cheng; Nan, Sun
doi: 10.1504/IJRIS.2017.090046pmid: N/A
In order to promote the accuracy of agricultural pollution source prediction algorithm, an agricultural pollution source prediction algorithm based on chaotic differential evolution algorithm neural network is proposed in the article. Firstly, it initialises the population of differential evolution algorithm with chaos theory, to promote the diversity for initial population solution; then it improves the differential evolution algorithm by using mean entropy and perturbation variation method to promote its optimise performance. Secondly, it optimises the neural network parameter learning process by using the improved differential evolution algorithm to promote the accuracy of parameters optimisation. Lastly, it applies the proposed algorithm into the example of local agricultural pollution source prediction and the results showed that the proposed method could effectively increase the accuracy of agricultural pollution source prediction.