One of the basic issues in the navigation of autonomous mobile robots is the obstacle avoidance task that is commonly achieved using a reactive control paradigm where a local mapping from perceived states to actions i...
详细信息
One of the basic issues in the navigation of autonomous mobile robots is the obstacle avoidance task that is commonly achieved using a reactive control paradigm where a local mapping from perceived states to actions is acquired. A control strategy with learning capabilities in an unknown environment can be obtained using reinforcement learning where the learning agent is given only sparse reward information. This credit assignment problem includes both temporal and structural aspects. While the temporal credit assignment problem is solved using core elements of the reinforcement learning agent, solution of the structural credit assignment problem requires an appropriate internal state space representation of the environment. In this paper, a discrete coding of the input space using a neural network structure is presented as opposed to the commonly used continuous internal representation. This enables a faster and more efficient convergence of the reinforcement learning process.
Internal thermally coupled distillation column (ITCDIC) is the frontier of energy saving distillation research. In the paper, an evaluation method on operating cost and its saving in the ITCDIC processes of ideal mixt...
详细信息
ISBN:
(纸本)0780372980
Internal thermally coupled distillation column (ITCDIC) is the frontier of energy saving distillation research. In the paper, an evaluation method on operating cost and its saving in the ITCDIC processes of ideal mixtures is presented. A mathematical model for optimization is first derived. The benzene-toluene system is studied as an illustrative example. The optimization results show that the energy cost saving potential is close to 40% and annual energy cost saving is about 0.3 million dollars compared with the operating cost of a conventional distillation column operated at the minimum reflux ratio. The process analysis is also carried out. The optimal operating conditions and some useful results are obtained. These pave the way for the smooth operation and the further optimal design of ideal ITCDIC processes.
As a matter of fact that the most plants are nonlinear in practice, the proper T-S fuzzy model is obtained by recursive fuzzy identification method to describe the real complex nonlinear systems. First, a fuzzy-cluste...
详细信息
ISBN:
(纸本)0780375084
As a matter of fact that the most plants are nonlinear in practice, the proper T-S fuzzy model is obtained by recursive fuzzy identification method to describe the real complex nonlinear systems. First, a fuzzy-clustering algorithm is used to competitively learn the centers of the input-areas online, Second, the radius of input- areas are determined based on the superpose degree of the predetermined rules, while the parameters in the inference can be achieved by the RLS algorithm, then the fuzzy local linear model can be used as the CARIMA model at every instant to predictive the output of the plant. Finally, a novel-tuning algorithm for PID controller parameters is presented based on the idea of GPC and the relation between the GPC and PID controller. The optimal index includes the weighting term on the square of error between the predicted process steady-state value and the desired-state value. What's more, the stability of the system has also been studied. The simulation results both in theory and paper pulp process show the effectiveness of the proposed method.
In this paper, the design of available bit rate (ABR) congestion control in ATM (asynchronous transfer mode) networks is considered. The goals of congestion control are high link utilization, low cell loss, and low ne...
详细信息
In this paper, the design of available bit rate (ABR) congestion control in ATM (asynchronous transfer mode) networks is considered. The goals of congestion control are high link utilization, low cell loss, and low network delay. Various algorithms for congestion control can be found in the literature. Most of them are based on a network model, which is assumed to be known. In this paper a recursive least square (RLS) algorithm for on-line identification of the network model is implemented and combined with a generalized predictive controller (GPC) and with an I-controller based on the Smith predictor. Simulations were carried out to prove the effectiveness of the proposed algorithms.
A laboratory model for experimental investigations of the rubber-asphalt sliding pair has been designed with the purpose of better understanding of dynamic behavior of the friction force in the contact patch between t...
详细信息
This paper presents a laboratory testing of genetic algorithm (CA) based self-tuned PI controller for the speed control of interior permanent magnet synchronous motor (IPMSM). A radial basis artificial neural network ...
详细信息
This work presents dominant pole design (DPD) methodology, one MATLAB implementation of DPD and its usage. DPD for PI and PID controllers is derived, and on this basis MATLAB toolkit for three-pole method (3PM) has be...
详细信息
This work presents dominant pole design (DPD) methodology, one MATLAB implementation of DPD and its usage. DPD for PI and PID controllers is derived, and on this basis MATLAB toolkit for three-pole method (3PM) has been developed. The least squares procedure for finding functional dependence of PID controller parameters on process parameters is described. This least squares procedure has been applied to the first order process with dead time (FODT) obtained functional dependence, expressed through tuning formulas, is presented. Performance of PID controller tuned with 3PM tuning formulas is compared to the performance of PID controller tuned with integral criteria derived tuning formulas. Observed characteristics of DPD method are given at the end of the paper.
Support vector machine (SVM) based classification is applied to fault diagnostics of an electrical machine. Numerical magnetic field analysis is used to provide virtual measurement data from healthy and faulty operati...
详细信息
Support vector machine (SVM) based classification is applied to fault diagnostics of an electrical machine. Numerical magnetic field analysis is used to provide virtual measurement data from healthy and faulty operations of an electric machine. Power spectra estimates of a stator line current of the motor are calculated with Welch's method, and SVMs are applied to distinguish the healthy spectrum from faulty spectra. Multiple SVMs are combined with a majority voting approach to reconstruct the final classification decision.
controller performance assessment is a challenging task in industrial processcontrol. In this work, we review the state of research in controller performance assessment and monitoring, and present some commercial imp...
详细信息
controller performance assessment is a challenging task in industrial processcontrol. In this work, we review the state of research in controller performance assessment and monitoring, and present some commercial implementations of the technology in process industries. Challenges related to the controller performance evaluation are outlined for research and applications.
Distillation columns constitute a significant fraction of the capital invested in the refineries around the world; their control requires a major part of the total operating cost of chemical processes, if the used str...
详细信息
Distillation columns constitute a significant fraction of the capital invested in the refineries around the world; their control requires a major part of the total operating cost of chemical processes, if the used strategy is not adequate. This article presents the application of optimal fuzzy control to reduce the energy consumption of a Benzene-Toluene distillation column. This method is based on the determination of the specific values of the fuzzy controller parameters such that certain performance criterion is minimised. Results of a simulation study are presented showing the potential improvement offered by this method.
暂无评论