In this paper, we propose a time domain design of the deadbeat digital controller with series compensation of the integral-action and local feedback compensation. It is proved that the control system proposed here gives the deadbeat response not only for the step change of the desired value but also the disturbance, and the simulation result makes the fact clear.
The dynamic characteristic of the process is expressed by the step or impulse response, then the sequence of the control signal which makes deadbeat response can be easily obtained by the matrix computation, using sampled data of these responses. The designing method of the integral-action gain constant and feedback compensator is derived by using the sequence of the control signal.
This method is very simple comparing with other methods which use the pulse transfer function or state equation, and can be systematically applied to the system including the lag time or the restriction of the manipulated variable.

抄録全体を表示

This paper deals with an LQ problem in which the plant and/or the cost functional contains a finite Markov chain as a random parameter. Although the existence of the optimal solution has not been completely cleared, this problem has been almost solved under the assumption that the value of the parameter can be measured as well as the state variables of the plant. The optimal control signal is generated by a linear state feedback law with a gain matrix dependent on the parameter. In this paper, we are interested in the case where the parameter has a nominal value and stays at the other values for very short time in average. We prove the existence of the optimal solution, and consider the property of an approximate solution constructed from the solution of the matrix Riccati equation with the nominal value of the parameter. For the particular problem in which all the coefficients on the control variables are constant, the approximate solution consists with the optimal solution of the usual LQ problem with the parameter fixed to the nominal value. It is shown that the loss caused by the approximation vanishes faster than the average staying time of the parameter at the non-nominal values as the latter is made to approach zero.

抄録全体を表示

When Kalman filters are implemented with microprocessors or signal processors, quantization errors (rounfoff errors and coefficient quantization errors) due to finite wordlength implementation affect the state estimate. This paper studies analysis of quantization errors in Kalman filters and synthesis of minimum quantization error Kalman filter structures.
Infinite wordlength Kalman filters are described by the state and output equations, and filter structures are introduced to Kalman filters by the state transformation, since quantization effects of digital systems highly depend on structures. Finite wordlength Kalman filters are also described by the equations. Roundoff errors and coefficient quantization errors are analyzed for any structures of Kalman filters. The results of error analysis are in agreement with the results of simulation. Minimum quantization error Kalman filters are synthesized by using minimization method of quantization errors in state-space digital filters. Synthesis method of minimum quantization error Kalman filters is very effective to reduce quantization errors.

抄録全体を表示

On the minimal realization from the Markov parameters of a linear dynamical system, beautiful results were obtained by Ho-Kalman and Rissanen. In particular, the Rissanen algorithm is superior in generality, simplicity, recursive structure, and numerical stability. However, it has redundancy. For example, it produces n(n+3)/2 parameters for a strictly proper linear scalar system, while the transfer function of that system has at most 2n parameters.
In this paper, an algorithm for the minimal realization from the Markov parameters of linear constant multivariable system is obtained which produces an observable (controllable) canonical form realization with minimum number of parameters and preserves the superior properties in the Rissanen algorithm. Further, from this algorithm the irreducible matrix fraction description with minimum number of parameters is easily derived.

抄録全体を表示

In this paper, the servomechanism synthesis for nonlinear multivariable systems is considered.
“Servomechanism performance”, here we mean, is to track any step reference signals with zero steady state error while the system is subject to any step external disturbances.
While nonlinear characteristics can be seen in various manners, we concentrate on certain classes of nonlinear systems than to attack the general cases. Following nonlinear systems are treated; systems with input-channel non-linearity, systems with output-channel non-linearity, and each nonlinear characteristic is assumed to be sector-type.
Firstly we present feedback laws that insure the stability of compensated nonlinear systems.
Next, based on these results, derive necessary conditions that the nonlinear systems to be controlled must satisfy to be able to get the desired “servomechanism performance”. Furthermore, the existence condition of an unique equilibrium point are also investigated.

抄録全体を表示

A problem of designing an adaptive tracking system with internal stability is considered by an indirect method for the single-input single-output discrete-time linear system described in Lüders-Narendra canonical state form.
Under the assumptions that system parameters and the state are known, appropriate state-feedback gains and pre-compensator's gains that replace all the system poles by the desired ones and, at the same time, control the output to follow the reference signal generated through a known dynamical system with an unknown initial condition are first characterized in a form easy for computation. In the adaptive design, these system parameters and the state are all replaced by the estimates. The least square estimation scheme is used for parameter estimation and the state is reconstructed via the method of an adaptive observer. It is then established that in the proposed scheme, all the variables remain bounded, the closed-loop poles approach the desired ones and the output tracks the reference trajectory asymptotically. A computer simulation result is also provided to illustrate the tracking behavior to the desired reference trajectory.

抄録全体を表示

In this paper, a design method of a model reference adaptive control system (MRACS) is presented for linear continuous single-input single-output plants in the presence of disturbanes.
It is assumed that the disturbances are expressed as the solutions of linear differential equations with unknown coefficients. Examples of such disturbances are stepwise, sinusoidal and exponential ones, including the sum of them. The proposed MRACS also works as a disturbance accommodating control system.
The boundedness of all signals in the MRACS is proved, which assures the convergence of the output error to zero.
Computer simulations are presented to illustrate the effectiveness of the proposed method.

抄録全体を表示

The present paper gives a scheme for compensating unknown deterministic disturbances in the continuous-time model reference adaptive control systems. The disturbaces cause the variations in tuned parameters of the controller as well as the output errors between the reference outputs and system outputs. The proposed adaptive control system consists of two feedback loops: one performs the model-matching adaptively and the other includes an error feedback controller which can compensate the disturbances, uncertainty in initial conditions of the system and unexactness in the model-matching. It is shown that even if the error feedback controller is employed the condition for the model-matching and the asymptotic stability of the adaptive system can be preserved. Finally, we present the effectiveness of the proposed configuration in numerical simulation results.

抄録全体を表示

In spite the fact that adaptive control is nothing but an exact model matching under the condition that the plant parameters are unknown. there have been very few researches which have been carried out along the line of extending exact model matching to unknown plant. The usual exact model matching developed by Wolovich is not easy to be extended to adaptive control because the input dynamics of the resulting system depend on plant parameters. In this paper, a new exact model matching with the input dynamics which do not depend on plant parameters are presented, and is extended to the construction of adaptive control system in the very natural way. The structure of the resulting adaptive control system is essentially the same as the one of Narendra et al., but it needs the concept of neither positive real function nor augmented error signal, and is very plain and understandable. It is proven easily that the plant input signal converges to that which is employed in the exact model matching, and hence the total stability is assured. Also, the problem of degree difference of the plant does not cause any trouble at all.

抄録全体を表示

The anomaly diagnosis for a system is important to assure the safety operation of the system and the improvement of an availability.
This paper discusses the optimal sensor location for the linear dynamical system, whose anomaly model is assumed as a dynamics step, based on the Kullback's information criterion such that the anomaly is detected as quickly and as accurately as possible. The proposed method is also shown to be reasonable from the viewpoint of the generalized likelihood ratio (GLR) detection method.
Furthermore, the proposed sensor location is applied to the anomaly diagnosis of actuator and plant and shown that the sensor location is also effective for the anomaly diagnosis whereas it is synthesized based on the dynamics step anomaly model.

抄録全体を表示

Several efforts have been aimed at automatic transcription of instrument tone. This paper deals with a transcription method of sung song using digital signal processing techniques. The fundamental function of transcription is the measurement of tonality (key signature) and metrics (time signature). The tonality measurement is based on peak frequency identification and the metrical feature is obtained from segmentatoin of the vocal sound. The peak frequency is identified by the interpolation of its complex spectrum, which has been recently proposed by the authors. Unstable peak frequency is reduced by median filtering and pitch drift is compensated by histogram techniques. The segmentation is carried out by power dip, key change, and variation of harmonics ratio. The key scale is determined by comparing the tonality distribution with every major scale and minor scale. The metric unit is given by correlating the segmentation result with the standard rhythm.
Experimental resuits show that two Japanese nursery rhymes have been stably transcribed into acceptable notes. This system can be served as a supporting facility for the ethnomusicology.

抄録全体を表示

This paper describes an analysis of the temperature characteristic of piezoresistive semiconductor pressure sensors and its improvement. In this analysis the non-linear temperature dependence of thermal expansion of silicon is considered to calculate the thermal stress in the silicon diaphragm. The temperature characteristic of the sensor is then calculated considering the non-linearity of piezoresistive coefficients for temperature. This analysis reveals that the temperature characteristic strongly depends on the difference in thermal expansion between the silicon diaphragm and its mounting die. The appropriate dimension of the mounting die which minimizes this thermal stress is calculated and a trial pressure sensor is manufactured. The trial pressure sensor with a compensating circuit consisting of thermisters and resisters shows a good temperature characteristic; the zero point change and the span change are less than±1% for temperature range -40∼120°C.

抄録全体を表示

The difference equation approach is mostly used in the computer control of heating temperature of billet in the furnace. This approach is not suitable for the on-line prediction calculation as it requires a vast amount of recursive computation. The form of the difference equation used in this approach does not make it possible to directly apply various modern control strategies to the control aspect.
To avoid these problems it is desired to have a linear constant equation which relates the temperature of billet as output and the temperature of furnace as input. Theoretical derivation is impossible, because thermal constants are nonlinear and heat transfer coefficient is not constant over the whole furnace.
A new algorithm is proposed in this paper. The furnace is divided into several zones in which a linear constant input-output equation can be obtained. Intervals are determined not apriori but from the operating data. The boundary of intervals is fixed according to the behaviour of recursive estimates of parameters.
The proposed algorithm has been applied to the existing furnace with satisfactory results. The predicted temperature of billet agree with the observed data with good accuracy.

抄録全体を表示

A path planner and an execution system are proposed for autonomous vehicle control.
The planner creates a near shortest path avoiding obstacles that are represented by combinations of circles and line segments on a two dimensional map. For realizing real time execution, path search procedures employ a heuristic pruning strategies in selecting a node to expand and in generating successor nodes. Nodes are selected for expansion in order, according to a cost assigned to each node. The cost is mainly evaluated by approximating a path length from the initial node to the goal node. In order to expand a node and to generate successor nodes, a specific search procedure is activated that finds positions avoiding obstacles in the direction of the goal, and creates successor nodes corresponding to the positions.
The execution system, utilizing an ultrasonic range finder equipped to the vehicle performs a plan repair against unknown obstacles when echoes from the obstacles are observed. The plan repair is conducted by a map edition and replanning in such a way that new circles representing the echoes are added to the map.
Obstacle avoidance tests with a vehicle controlled by microprocessors demonstrate the utility of heuristics just outlined.

抄録全体を表示