This paper presents a novel method for estimation of high-dimensional state variables by means of particle filtering (PF). One of the major drawbacks of PF is that a large number of particles is generally required for accurate estimation of state variables lying in a high-dimensional space, whose maintenance is time-consuming. In many applications, the high-dimensional state variables can be divided into two groups according to the hierarchy that an object model possesses; one group (higher layer) is easily integrated out analytically from the posterior densities, and we have only to carry out PF for the other group (lower layer) whose state space is reduced from the whole state space. In this paper, we propose a novel proposal distribution in the lower layer, which is a mixture of approximate prediction densities computed in the higher and lower layers. An adaptive determination of the mixture ratio, which implements a mutual interaction between the layers as a mixture state transition model in PF, is realized by means of on-line EM algorithm for adapting to complicated real environments. The effectiveness of the proposed method is demonstrated by computer simulations of head pose estimation of a car driver.
The object-oriented transaction processing characterized by recursive invocation of method for object is realized by concurrency control which satisfies object-oriented serializability. As the conventional semantic concurrency control is executed by switching subtransactions, the transaction which has arrived subsequently can be partly executed by suspending the precedent transaction which has not been completed, as far as the object-oriented serializability is guaranteed. Despite of allowance of partial execution for the subsequent transaction, the semantic concurrency control faces the left off problem about precedent transactions, which means that the precedent transactions cannot be executed for a long time. This paper presents an object-oriented transaction processing by introducing priority ceiling to resolve the left off problem. Furthermore, to evaluate performance of our proposed method detailed experiments were carried out from various viewpoints. Especially, from experimental results about execution time of transaction, left-off transactions, execution time of transaction, suspended time and waiting time of subtransactions, we obtain detailed consideration of the performance evaluation and show effectiveness of our proposed object-oriented transaction processing by the priority-ceiling.
This paper is concerned with intrinsic control performance limitations achievable by feedback for single-input and multi-output (SIMO) possibley unstable and non-minimum phase linear discrete-time systems. We investigate two typical H2 optimal control problems, namely optimal tracking control problem and minmal energy control problem. We derive closed-form analytical expressions of the best achievable tracking error and minmal energy H2 norms. The former consists of plant gain as well as unstable poles and non-minimum phase zeros of the plant, while the latter is only given by unstable poles and non-minimum phase zeros of the plant. Those results are confirmed by several numerical examples, and they characterize easiness and difficulty of plants to be controlled.
This paper considers robust control analysis and synthesis problems via the expression of our previously proposed summational type state equation. The summational type state equation is a mathematical expression to solve two essential problems, i.e., one of them is a physical problem of discontinuity in mathematical expressions e.g. the controllable canonical form for different orders of the existing state equation, and the other one is an engineering problem of disunification in which continuous-time and discrete-time systems are not described with consistency. First, this paper introduces the summational type state equation. Next, a robust stability analysis condition and a feasible condition for the scaled H∞ control synthesis problem are derived from the summational type state equation. Furthermore, the relation between analysis/synthesis conditions based on summational and difference type state equations is clarified. Finally, the effectiveness of the summational type state equation is shown by numerical examples on a sensitivity minimization problem. From these results, this paper points out that the summational type state equation is one of possible important mathematical expressions in control theory.