Recently, a design of model matching and adaptive control systems under the presence of measurement noise is reported by employing a non-minimal order state observer. But this method can be only used for the case where the plant has a stable inverse. In this paper, we will show that the idea is also applicable for a nonminimum phase plant. We will consider the plant having not only measurement noise but also input noise. The freedom which arises by employing non-minimal order observer will be used for noise reduction. To clarify the degrees of freedom, a parametrization using an extended division algorithm will be shown.
It is important in developing intelligent scheduling systems to build an effective method for acquisition of decision rules. We have proposed a method for acquisition of scheduling rules by using inductive learning. Cases needed for rule acquisition are generated by interchanging two jobs in a sample schedule, and job attributes and location properties of jobs are used as the basic information for the rule acquisition. The proposed method for rule acquisition has been applied to single machine problems and flowshop problems. This paper extends the authors' previous work to jobshop scheduling problems and describes a method to acquire scheduling rules by generating a number of sample schedules and analyzing their properties. The effectiveness of the proposed method for rule acquisition and the applicability of such obtained rules are demonstrated by applying them to the exploration process of local search.
In this paper, we propose a view invariant method for face discrimination using two kinds of three-layered neural networks. The first neural network learns informations of face obtained from different view point. Considering the human visual mechanism, two images from different views are fed into the first neural network. The second neural network integrates the outputs of the first neural network and recognizes a person. Our method has achieved a recognition rate of 100% on a database of 20 persons containing 150 images per person.
Neural network has been widely applied to business problems such as marketing, management decision making, stock market and so on. In this kind of fields, the data are usually very noisy and simple application of back propagation is not possible. For nonlinear regression problems, a method has been proposed to realize the monotonicity of input-output relation by using constraint of coefficients. In this paper, we propose a method to the BP training with prior knowledge by using gradient reference points. The effectiveness of this algorithm is shown by numerical simulation.
Recently, a design of model matching and adaptive control systems under the presence of measurement noise is reported by employing a non-minimal order state observer. But this method can be only used for the single-input, single-output plant which is stably invertible. In this paper, it will be shown that the idea is also applicable to the multivariable plant which may not have a stable inverse. Since the proposed method can be implemented by simple calculation, it is adequate for an adaptive control in which the controllers must be calculated on line.
This paper deals with how to evaluate the effectiveness of carbon tax and energy tax for regulating the carbon dioxide emissions. For this purpose we mainly deal with a primal problem and its dual problem of dynamic linear programming model. The primal problem is formulated by extending Leontief type input-output model and the basic idea of commodity stocks. It represents the balance of materials. The dual problem is obtained and interpreted as cash balance. It is clarified in this paper whether the carbon tax and energy tax are effective to decrease the total amount of carbon dioxide emissions.
An algorithm which simultaneously applies the Fuzzy c-Means clustering and the correspondence analysis is developed. Maximization of an objective function yields memberships of fuzzy clusters and assigns values to categories and patterns. A regularization term is introduced in the objective function and we apply the Lagrangean multiplier method and an adaptive method using eigenvalues in the correspondence analysis.