In this paper,an interacting multiple-model(IMM)method based on datadriven identification model is proposed for the prediction of nonlinear dynamic systems.Firstly,two basic models are selected as combination componen...In this paper,an interacting multiple-model(IMM)method based on datadriven identification model is proposed for the prediction of nonlinear dynamic systems.Firstly,two basic models are selected as combination components due to their proved effectiveness.One is Gaussian process(GP)model,which can provide the predictive variance of the predicted output and only has several optimizing parameters.The other is regularized extreme learning machine(RELM)model,which can improve the overfitting problem resulted by empirical risk minimization principle and enhances the overall generalization performance.Then both of the models are updated continually using meaningful new data selected by data selection methods.Furthermore,recursive methods are employed in the two models to reduce the computational burden caused by continuous renewal.Finally,the two models are combined in IMM algorithm to realize the hybrid prediction,which can avoid the error accumulation in the single-model prediction.In order to verify the performance,the proposed method is applied to the prediction of moisture content of alkali-surfactant-polymer(ASP)flooding.The simulation results show that the proposed model can match the process very well.And IMM algorithm can outperform its components and provide a nice improvement in accuracy and robustness.展开更多
This paper presents an adaptive rationalized Haar function approximation method to obtain the optimal injection strategy for alkali-surfactant-polymer(ASP) flooding. In this process, the non-uniform control vector par...This paper presents an adaptive rationalized Haar function approximation method to obtain the optimal injection strategy for alkali-surfactant-polymer(ASP) flooding. In this process, the non-uniform control vector parameterization is introduced to convert original problem into a multistage optimization problem, in which a new normalized time variable is adopted on the combination of the subinterval length. Then the rationalized Haar function approximation method, in which an auxiliary function is introduced to dispose path constraints, is used to transform the multistage problem into a nonlinear programming. Furthermore, an adaptive strategy proposed on the basis of errors is adopted to regulate the order of Haar function vectors. Finally, the nonlinear programming for ASP flooding is solved by sequential quadratic programming. To illustrate the performance of proposed method,the experimental comparison method and control vector parameterization(CVP) method are introduced to optimize the original problem directly. By contrastive analysis of results, the accuracy and efficiency of proposed method are confirmed.展开更多
基金supported by National Natural Science Foundation under Grant No.60974039National Natural Science Foundation under Grant No.61573378+1 种基金Natural Science Foundation of Shandong province under Grant No.ZR2011FM002the Fundamental Research Funds for the Central Universities under Grant No.15CX06064A.
文摘In this paper,an interacting multiple-model(IMM)method based on datadriven identification model is proposed for the prediction of nonlinear dynamic systems.Firstly,two basic models are selected as combination components due to their proved effectiveness.One is Gaussian process(GP)model,which can provide the predictive variance of the predicted output and only has several optimizing parameters.The other is regularized extreme learning machine(RELM)model,which can improve the overfitting problem resulted by empirical risk minimization principle and enhances the overall generalization performance.Then both of the models are updated continually using meaningful new data selected by data selection methods.Furthermore,recursive methods are employed in the two models to reduce the computational burden caused by continuous renewal.Finally,the two models are combined in IMM algorithm to realize the hybrid prediction,which can avoid the error accumulation in the single-model prediction.In order to verify the performance,the proposed method is applied to the prediction of moisture content of alkali-surfactant-polymer(ASP)flooding.The simulation results show that the proposed model can match the process very well.And IMM algorithm can outperform its components and provide a nice improvement in accuracy and robustness.
基金Supported by the National Natural Science Foundation of China(61573378)the Fundamental Research Funds for the Central Universities(15CX06064A)
文摘This paper presents an adaptive rationalized Haar function approximation method to obtain the optimal injection strategy for alkali-surfactant-polymer(ASP) flooding. In this process, the non-uniform control vector parameterization is introduced to convert original problem into a multistage optimization problem, in which a new normalized time variable is adopted on the combination of the subinterval length. Then the rationalized Haar function approximation method, in which an auxiliary function is introduced to dispose path constraints, is used to transform the multistage problem into a nonlinear programming. Furthermore, an adaptive strategy proposed on the basis of errors is adopted to regulate the order of Haar function vectors. Finally, the nonlinear programming for ASP flooding is solved by sequential quadratic programming. To illustrate the performance of proposed method,the experimental comparison method and control vector parameterization(CVP) method are introduced to optimize the original problem directly. By contrastive analysis of results, the accuracy and efficiency of proposed method are confirmed.
