For linear periodic systems, the assignment of a set of characteristic multipliers is an important problem. Brunovsky pointed out that the necessary and sufficient condition for assigning the characteristic multipliers was the system to be completely controllable, and Kohno considered the same problem for discrete systems. But there were left some unresolved problems to implement the results.
This paper discusses the necessary and sufficient condition of controllability for such systems that have analytic coefficients. Under this condition, we propose an easily programmable type algorithm for assigning a set of characteristic multipliers, where the feedback matrix is composed of a sequence of Dirac delta functions. This type of feedback is easy to implement using proper approximation to the delta function and is proved to be useful by applying it to a system governed by Mathieu equation.

View full abstract

This paper is concerned with a least-squares state estimation problem for noisy observations with feedback. The statistics of the observation noise is assumed to be dependent on the state variables so that we can accept a wider class of applications than the case of usual white Gaussian noise.
Firstly, a form of equivalent data transformation is shown which transforms noisy data, via feedback, into one with a conevnient format for signal processing. It is also shown that any equivalent transformation can be represented in this form.
Secondly, nonlinear filtering formulas are derived for the observation obtained through an equivalent feedback transformation. The result previously obtained by the author for feedbackless observations is applied to get a generalized Bayes' formula for the optimal estimate, and approximate formulas in the Bayes' fashion and in the stochastic differential equation. The approximation is made by applying additional noise, and it is shown that the approximate estimate converges to the optimal one as the variance of the additional noise approaces to zero.

View full abstract

This paper describes a new attempt at the non-linear state estimation for the stochastic environmental system with varying S/N ratio. A recursive algorithm of estimating higher order statistical quantities of arbitrary function type, not to mention mean or variance, is obtained by introducing a new expansion form of Bayes' theorem. Furthermore, this method is widely applicable for the actual case when the random fluctuation is of non-Gaussian type. The algorithm proposed in this paper agrees completely with the well-known Kalman filtering theory as a simplified special case when the stochastic system is a linear type with Gaussian random excitation. Finally, the validity and effectiveness of the theory are experimentally confirmed by applying it to the actually observed room acoustic data and the road traffic noise data.

View full abstract

The decoupling problem for linear time-invariant multivarialbe systems has been extensively investigated. Many results which have been obtained so far are important from the theoretical point of view, but it is also important to develop efficient and accurate computational procedures for sythesizing decoupling control systems. But, thorough study along these lines has not yet been made. In this paper, we give a new decoupling algorithm. On the basis of Hessenberg forms, we give a necessary and sufficient condition for state feedback decoupling and a procedure for constructing a decoupling control law. Since this algorithm involves only orthogonal transformations, the numerical stability is guaranteed. A system representation similar to the CD system is introduced in order to check the internal stability. A numerical example is given.

View full abstract

In this paper, some considerations about improvement of transient characteristics of control systems with proportional and integral terms of output errors are described. Those are done as follows. That is, the control systems mentioned above are reduced to gain output feedback control systems by introduction of augmented systems. And reductions of overshoots or maximum/minimum peaks for changes of set points are considered from the viewpoint of eigenvalues assignment.
For a controllable and observable multi-input system with n states, m inputs and l outputs, if n≤m+l-1, then almost arbitrary sets of closed-loop eigenvalues are assignable by using output feedback. Then, in general, some degrees of freedom remain in gain matrices after the eigenvalues have been assigned. Therefore, transient characteristics may be improved by making use of these degrees of freedom.
In this paper, this improvement method is described and its effectiveness is shown by the numerical example.

View full abstract

Most of work on the identification theory for viscoelastic systems and its application have been restricted to the isotropic systems. These models are not enough to express all the actual viscoelastic systems because improvements in material properties and more effective engineering capabilities in utilizing anisotropic behavior have made an increase of anisotropy in viscoelastic systems.
In this papar we propose the identification method of three-dimensional hyperbolic distributed parameter systems with anisotropic viscoelasticity. First of all, we derive the Green function of systems whose boundary conditions are given as one free surface and another fixed surfaces. In this Green function, the eigenvalues coresponding to the free surface are expressed in so complex a manner that we approximate them explicitly for the identification of the system.
The validity of the proposed method has been confirmed by digital simulation and we applied this method to the actual anisotropic viscoelastic system to estimate the system prameters. The results of the estimates from the actual physical system have shown that the proposed method is useful.

View full abstract

The velocity distribution at the inlet of a capillary viscometer is determined by the flow before the viscometer. Therefore, this flow is considered to affect the measurements of viscosity by the capillary viscometer.
By the theoretical considerations applying Bernoulli's theorem and Newton's law of motion to the flow, the effect is indicated as πa²Δp and let M_A be the momentum flowrate at the inlet of the capillary tube, it is known as M_A>πa²Δp>1/2M_A.

View full abstract

This paper is concerned with the problem of designing model reference adaptive control systems (MRACS) for single-input single-output linear discrete time systems having deterministic disturbances such as sine waves, biases, ramp waves, etc. An exponential type polynomial disturbance model is introduced as an approximation of disturbance. The MRACS is designed for the composite system which consists of the plant and the disturbance model. The composite system parameters are estimated by using an adaptive observer with exponential rate of convergence and the adaptive input is determined based on these parameters so that the output tracking error tend to zero exponentially. It is shown that the algorithms will ensure that the system inputs and outputs remain bounded for all time.
The effect of the introduction of disturbance model is confirmed experimentally by applying he proposed method to the MRAC problem of a second-order thermal system.

View full abstract

In this paper, we discuss a design of exact model matching systems (EMM) and its application to model reference adaptive control systems (MRACS) and to adaptive finite time settling control systems (AFSC) for linear discrete time systems.
In the EMM design, the mtarix inversion to solve a polynomial equation is not required and the dimension of feedback compensators are kept to be minimum, which makes the design method applicable to MRACS. In the design of MRACS, a projection type parameter identification scheme, called residual projection method (RPM) in this paper, is used, which ensures the parameter convergence in finite steps.
The design method is also applied to AFSC with the RPM, in which the monotonic convergence of the output error is shown.
A numerical example is given to illustrate the effectiveness of the adaptive finite settling control systems.

View full abstract

Acceleration control using detected external force, is a scheme to control a manipulator to carry an object with the acceleration proportional to the external force. A special type manipulator named suspend-type manipulator was designed for the controlled manipulator.
Suspend-type manipulator is a paralled type manipulator that hangs an object with wires and controls its motion by driving those wires. It has an ability same as the transfer machines such as cranes and can control the position of the object more exactly.
Adaptive control to the external force is useful for aiding the manual transfer operation. Most control systms or mechanisms that have developed to generate the adaptive motion to the external force, are too much dedicated to some task or hardware mechanism, even if some of them are widely used in industries. They do not respond to various situations and requirements.
We have developed a control system responding to the external force as follows. First, we decompose the whole control system into three parts; a part to detect the external force, a part to generate desired motion, and a part to make the manipulator follow the desired motion. With this configuration, we can make the control indepedent from the mechanism. Second, for the desired motion, the acceleration proportional to the external force is generated. It makes the desired motion continuous both in position and velocity, and an operator may change its characteristics by altering the ratio. Third, a method of external force detection is adopted that can identify the external force even when the manipulator is in motion.
The control system was implemented in an experimental system with a suspended-type manipulator. It worked successfully under some conditions.

View full abstract

The precision in measuring SAR image shift between looks using three look correlation is analyzed and verified. In this method, in order to improve the measurement precision and to produce optimally focused SAR imagery, n pairs of two relative image shifts among three looks are measured and selected by the rule that the difference between the pair should be less than a threshold of T.
So, the precision of three look correlation is described as a function of, n, T and σ, where σ is a newly defined measure of difficulty in measuring image shift. This function leads to the optimal threshold that maximizes the precision in the method.
The results of the image shift measurement test shows that the proposed method can reduce the measurement error to up to 2/3 of a conventional one caused by two look correlation.

View full abstract

A novel magnetic flowmenter has been proposed to eliminate insulating liner and to improve reliability and performance of magnetic flowmeters. The proposed magnetic flowmeter keeps the boundary condition between fluid and metallic pipe wall identical to that of non-conducting pipe wall by applying electric current automatically to form potential distribution on the wall. Subsequently, the output signal is the same as that of conventional magnetic flowmeters.
The feasibility and the effectiveness of the proposed flowmeter have been clarified by the theory and the experiments in previous papers, where the optimum position of the electrodes to feed electric currents and the optimum servo gain of the control cicuit were given by the theoretical analysis of the static performance of wall and fiuid potential.
The present paper describes the dynamic characteristics of the flowmeter. When a usual rectangular wave excitation is adopted, the effect of an electric double layer at the boundary of fiuid and metallic pipe wall cannot be neglected. Since it is difficult to analyze the distributed physics of the layer, a lumped equivalent circuit model which consists of parallel capacitance and resistance with series resistance is constructed and overall dynamics of the flowmeter is studied through computer simulations. The simulated output signal coincides with the experimental one for the various fiuid conductivity ranging from about 100 to 5000S/cm, when the parameters of the model are adequately chosen. Also, the simulation shows that both the capacitance of the layer and the conductivity of the fluid affect the waveform of the output signal.
Therefore, feedback circuit of potential formation is revised to eliminate these effects and to make the flowmeter output equal to conventional magnetic flowmeter. The signal for the feedback is sampled and held at the each end of the half periods of the rectangular wave excitation, where above effects become negligibly small as is demonstrated by the simulation. Both results of the simulation and the experiment of the new circuit configuration show good performance and the suppression of the effects of the capacitance and the conductivity.

View full abstract

A non-contact method of determining displacement of a rigid plate facing the open end of an acoustic resonance tube by measuring the amplitude of acoustic pressure inside the tube with constant excitation (or alternately, variation of the resonance frequency) is described. The method features itself by being independent of electric/magnetic properties of the object, not requiring compressed air supply, and not excerting any force on the object. By suitable choice for the excitation frequency and position of sound pressure pick up, rms pressure amplitude can be made change almost linearly with the displacement, by a factor up to 10 For a 2-dimensional case, both a finite difference calculation and experiment indicated good prospect for such method. The validity of the calculation was confirmed by agreement of its results with the experimental. By numerical simulation based on the calculation, linearity of output-displacement relation, magnitude of fractional change in the output, susceptibility of the output to pick-up position change, and relative level of sound pressure for constant frequencies and pick-up positions. Results are summarized in a map of the above-named characteristics with frequency and pick-up position as the coordinates. It is seen that a frequency a little lower than the fundamental (or the 3rd harmonic) resonance frequency of a tube with one end closed and the other end open gives satisfactory condition.

View full abstract