Pole locations of LQ regulators for a dynamical system (D.S.) and its augmented state system (A.D.S.) are discussed. The A.D.S. considered is made by adding to the state variable the integral z of the output y of the original D.S. The performance index of the A.D.S. is made by adding to the integrand of the original performance index a quadratic form ε²z'Q2z. Poles of the LQ regulator for the A.D.S. can be divided into two classes. Any pole of the first class is shown to lie near a pole of the LQ regulator of D.S. with distance of O(ε2).
Poles of the second class can be approximated by ε/κ. Here, κ's are real and negative, and are given by square roots of principal values of a certain regular pencil. Therefore, it may be said that poles of the first class are not so much affected as those of the second class when ε varies. By using above mentioned properties of the pole location, a method is proposed to choose the quadratic weight to the z part of the state of A.D.S.

View full abstract

A design optimization problem in continuum structure is usually characterized by minimizing an objective function with shape parameters subject to a governing field equation and several constraints. The problem can be reformulated as a variational problem, introducing its adjoint system. If the problem is nonlinear with respect to the shape parameters, it is in general difficult to solve the problem analytically.
This paper proposes a practical method to solve the nonlinear variational problem numerically with the aid of the finite element discretization technic combined with the multiplier method. For numerical examples of an application, shape optimization problems of a fin, which is used to promote the heat exchanger efficiency, are considered herein to show the validity of the proposed scheme.

View full abstract

A method is obtained for the parameter identification of one-dimensional heat conduction systems with unknown boundary conditions. The boundary conditions are assumed to be known up to constant parameters. The identification problem is formulated as the simultaneous estimation of unknown parameters and input functions based on the overspecified boundary data. Applyinig the concept of elementary solutions, an integral representation is introduced so as to give the explicit expression for the measured data in terms of unknown parameters and estimated input functions. The least square method is applied for the identification of distributed and boundary parameters. The input functions are estimated by solving a second-kind Volterra's integral equation. The methodology is applied for the modeling of the dynamic behavior of ambient temperature based on the observations at the surface of the earth. The simulation results are compared with measured data to prove that the mathematical model proposed in this paper can describe the diurnal variation of ambient temperature caused by solar radiation.

View full abstract

Many design problems of process control systems have the the following two characteristic features: (1) To obtain a complete model of the process is extremely diffcult, (2) the specification for the closed-loop dynamic performance is puite moderate. The analog PID controller is regarded as the most suitable device for such a situation, because its simplicity in the configuration and the ease of implementation tolerate and/or balance the model uncertainty. In digital control systems, however, we have greater flexibility in the controller configuration beyond the PID actions, still keeping the simplicity and the ease of implementation.
In this paper, we propose an integrated design procedure of digital control systems including modelling, with the purpose of exploiting the flexibility of digital controller to incorporate the model uncertainty in a proper way. The procedure is essentially based on the partial model matching which offers a reasonable way to simplify the design and the controller configuration under model uncertainty. Given a “one-shot” stop response data of the plant contaminated by the noise, the procedure determines a controller in a simple form which is a natural extension of digital PID controller. The design methodology is applied to a digital control of a simple system.

View full abstract

A feedback control law for water distribution networks by piece-wise linearization is proposed and evaluated. In the proposed control law, several linearized control models in accordance with water demand are prepared in order to cope with nonlinearity of water distribution networks. The applicability range of each linearized control model is clarified for its practical use.
First, a problem of controlling water distribution networks is formulated, and based on this formulation a linearized feedback control model is derived. Then, stability condition of the derived model is obtained for a simple water distribution structure and the condition is extended for large scale distribution networks by computer simulation study. It is also shown that adjustment of the coefficients of the performance index can extend the stability range of the linearized model. Finally, a control law composed with multiple linearized models is proposed and it is demonstrated that the proposed law is practically useful for controlling actual water distribution networks by computer simulation.

View full abstract

This paper proposes a new type inverse, which is dynamical (that is, without pure differentiators) and stable, for linear time-invariant continuous-time systems. Most of the inverses considered in previous studies can be classified into two types:
Type 1; inverses which reproduce the input to the original systems by using differentiators.
Type 2; inverses which reproduce the integral value of the input without using differentiators.
But in type 1, the use of differentiators are not desirable from practical point of view. As for the inverses of type 2, from the knowledge of the integral value of input we cannot know the real value of input without differentiation. In addition to that, inverses of type 2 are generally unstable as shown later in this paper. So we propose a new type inverse called filtered inverse which is dynamical, reproduces the approximate value of the input, and can be realized as a stable system.
The definition and existence condition of the filtered inverse are given first. Then the stability and minimal realization of this inverse are considered. The stability of usual inverses are also considered to show the necessity of filtered inverses. Finally application of the filtered inverse to servo systems and a numerical example are presented.

View full abstract

Urban planning should be evaluated not only from the side of administration, but also from the side of population in order to reflect population's opinions and their interests. It is surely necessary to include population's urban evaluation model in the urban decision-making process, which quantitatively represents population's value of urban life environment, and by introducing this model it would be easily accomplished to help deciding a plan which is sufficiently incorporated population's dissatisfaction with urban environment and their conflict between various interested groups and several subregions. In this paper, we propose the population's urban evaluation model and describe a method to construct the model and the practical application to Tobata ward in Kitakyushu city. The model is constructed by introducing an additive multiattribute utility function.
The method mainly consists of the following procedures.
1) Constructing a goal hierarchy of urban environment.
2) Gathering population's questionairs which contain relative weights among goals and an unidimensional utility function in each goal.
3) Classifying population into sevaral common interested groups based on their weights and utility functions. In this stage, principal component analysis is usefuly applied.
4) Specifying a group utility function in each group.
5) Specifying the population's utility function by aggregating group utility functions. In this stage, we propose three criterions on which indivdual utility functions are aggregated. One is equality criterion, another is goal utility satisfying criterion, and the other is goal consequence satisfying criterion.
As the result of application to Tobata ward, we obtained the population's urban evaluation model very well expressed problems which are currently confronted in Tobata and that, very effective informations about their welfare are obtained by their urban evaluations.

View full abstract

The rotor and stator of an ordinary electric motor are composed of materials with electrical and magnetic properties that do not change due to temperature change, so that the temperature dependence can almost be neglected in the motor performance.
But, motors using temperature-sensitive magnetic cores with a relatively low Curie temperature, as a part of the magnetic path, have a different torque and speed characteristics as compared to conventional motors when considering temperature change.
The temperature dependence of the temperature-sensitive magnetic core's reluctance controlling flux in the magnetic circuit affects the circuit constants related to the torque and speed of the motor.
With the rise of temperature, the dc motor torque and speed have rise and descent characteristics under load conditions, and ac motor torque and speed show descent chracteristics. A speed up range along with a rise of temperature can be used for a cooling motor in devices which exhaust high temperature gas or liquid. A speed reduction range along with a rise of temperature may find application to the fuel pump motor for keeping a heating system at a constant temperature.
However, conventional speed control methods used with temperature changes require many attachments, therefore the extremely simplified method has many merits for stability in performance both because of the possibility of a temperature sensor and flux control with aid of the temperature-sensitive magnetic core.
The devices using the temperature-sensitive magnetic core have already been proposed in the field of static electrical apparatus, but few papers have reported regarding the rotating electric machines.
The temperature dependence of torque and speed characteristics of the small dc and ac motor with temperature-sensitive magnetic core is described in this paper.

View full abstract

This paper is concerned with vortex shedding from rings and an application of this phenomenon to a vortex flowmeter for pipe flow.
Vortex rings shed alternately from the inside and outside of a ring rotate opposite to each other and form stable vortex rows on the common axis of symmetry. The configulation of the rows is found to be different from that of Karman vortex street, that is, the spacing between the alternating vortices is not equal. Characteristics of vortex shedding from rings with various shapes are also experimentally investigated.
Vortex shedding from a ring in pipe flow is not disturbed by the existence of the wall of the pipe and axisymmetrical velocity profile. The frequencies of vortex shedding from a ring can be detected more stably at the behind of the center of the ring comparing to vortex shedding from a plate. Thus, a ring is thought to have advantages as a bluff body of a vortex flowmeter for pipe flow.

View full abstract

In this report theoretical equations for the optical potential device (OPD) are introduced to clarify the output voltage characteristics of the OPD, based on an assumed equivalent circuit in which photoconductive portion of the OPD is represented by an assemblage of small partitions of unit length equal to a slit width and brush resistances for each partition of the photoconductor other than the photoreceiving partition are taken into consideration.
The results obtained from the experiment and calculation can be summarized as follows:
i) Output voltage characteristics of the OPD that can not be fully explained by the conventional equivalent circuit are made clear by the proposed equivalent circuit.
ii) Output voltage sensitivity of the OPD to the displacement of light spot depends upon slit width, light intensity, etc.
iii) Linear conditions of the OPD are made clear.
The above points may serve as a guide to the make or use of the OPD.
A rotating potentiometer which was made experimentally as an example of the application of the OPD has the following advantages:
i) free from problems arising from mechanical contact, ii) simple in structure and easy to make, iii) effective angle of rotation is wide, iv) superior in temperature characteristics, v) various non-linear output voltages are obtainable directly by simply changing the slit shape.

View full abstract

Some properties of random numbers generated by Tausworthe-Lewis-Payne (TLP) method are described.
It is shown that the generated random number w_i has the same period as that of n-th order M-sequence, unless the starting matrix W₀ which consists of w₀, w₁, …, w_n-1, is zero matrix. In order for wi to be uniform in the sense that W₀ appears only once in a period, it is necessary that W₀ has independent columns, which is guaranteed if the delay between rows r is such that (r, N)=1, where N=2ⁿ-1, the period of the M-sequence.
The autocorrelation function φ_ww(k) of w_i is shown to be obtained from sampling, every d digits, the autocorrelation function φ_xx(k) of the binary L-tuples from the original Msequence. Namely, φ_ww(k)=φ_xx(kd) where r·d=1 (mod N).

View full abstract

In fluidic digital circuits using wall attachment amplifiers, branch line systems are often used. The pressure wave form at the input port of a downstream element differs according to constructing conditions of line system such as position of branch point, the diameter of tubes, etc. So it is necessary to design the line system so that the desirable wave form might be transmitted.
In this paper, first we discussed the change of pressure wave due to reflection and transmission at branch point and end points by using the lossless line model and the method of characteristics. It was found that the change of pressure wave was given by the product of reflection and transmission coefficients and this relation was expressed by a simple diagram. The outline of transient response can be obtained easily from this diagram.
Next we discussed the actual conditions about the diameter and the length of tubes to get desirable wave form by using a practical example.
Last, transient response was calculated by applying the quasi-method of characteristics to the line model which considered viscosity and heat transfer. Experimental pressure wave forms were measured and compared with calculated results. The validity of discussions about conditions of line system was verified and usefulness of calculation was confirmed.

View full abstract

Two sets of bellows are located on the same axis in an enclosure which is opened by an exhaust vent to the atmosphere. A nozzle is put on a movable end of one of the sets of bellows (feedback bellows) and faces the movable closed end of the other set (input bellows). The inside of the feedback bellows is connected to a supply chamber, of which the pressure is regulated to a constant value, through an orifice provided in the fixed end of the feedback bellows. The pneumatic circuit which is composed of the nozzle, the orifice and the feedback bellows is so arranged to have bistable transfer characteristics. Another path with a restrictor connects the supply chamber with the inside of the enclosure. The chamber in the feedback bellows is also connected to another set of bellows through an adjustable restrictor built in the fixed end of the bellows, The third bellow is also set in the enclosure and its closed movable end is used to control the flow out area of the exhaust vent.
When a trigger pulse is applied to the input bellows, the nozzle contacts its counterface and the output pressure (in the feedback bellows) jumps from a low level to a high level equal to the supply pressure. Then, the pressure of the third bellows increases gradually with a speed given by the resistance of the adjustable restrictor and the capacity of the bellows. This pressure increase makes the pressure in the enclosure high, because of the resistance increase of the vent. When the pressure in the enclosure reaches a constant level, the nozzle separates abruptly from its counterface by the compressive action of the two sets of bellows and the output pressure falls down to the low level.

View full abstract

This paper deals with the multi-variable servomechanisms autonomously to generate a trajectory corresponding to a given path for the purpose of the trajectory control of a manipulator.
The principle of trajectory generation is as follows. At first, the path is described by the function on X-Y plane. Then, the set of nonlinear differential equations corresponding to the motion of a point on the path, that is, the trajectory, is implicitly presented by two vectors. That is, one is the tangential vector of advancing direction at the point on the path and the other is the recovery vector which is orthogonal to the path and proportional in magnitude to the deviation from the path. Accordingly, the trajectory can be realized autonomously by constructing the system that solves the set of nonlinear differential equations mentioned above.
The features of this method are as follows:
(1) The system can be easily constructed by using a microprocessor.
(2) The trajectory has strong stability against the disturbance and the changes of parameters in the servomechanism.
(3) The representation of the paths by general equations will make it easier to express the complex paths required and also to reduce the data corresponding to them.
In this paper, the principle of the trajectory generation was described at first, Then the stability and the steady state performance were analyzed selecting the line and the circle as the practical trajectories. Furthermore the computer simulation of the system was made to clarify the effects of sampling time to the stability and the accuracy since the trajectory control may become the sampled-data control by using the digital computer in practice.

View full abstract

This paper describes a simulation study of a demand bus system for Tsukuba Science City. The demand bus system is especially for the transportation poor in the daytime. and is fully computerized. Input of demands by customers and information exchange between the control center and buses are performed at bus stops linked to the center through telephone lines. The residential districts and commercial districts are scattered in the system area of 3km×10km. For the simulation many bus stops in these districts are represented by 11 points, and based on traffic studies number of demands is set to the range of 10/h to 1000/h, and number of buses 1 to 50. After giving the numbers of demands and buses the simulation proceeds by tracing generation of demands, allocation and dispatch of a bus to a new demand and, if any bus stops at a bus stop, passengers' getting on and/or off at each discretized time. Then, characteristics of the system under steady state are considered, and relations between system parameters (number of buses, number of demands per unit time) and system characteristics such as level-of service, which is defined as a ratio of wait time plus ride time to direct ride time, customers' wait time at bus stops and maximum number of passengers in a bus are obtained. The heuristic algorithm is applied for allocation and dispatch of buses to demands, and a new type of the performance index for allocation and dispatch is proposed. The results show that the demand bus system is superior to the conventional bus system in respect to both level-of service and wait time, and that the demand bus system is effective in the present city.

View full abstract