Factory automation is advanced in industrial field, and then higher speed drive and performance are desired for an AGV (automated guided vehicle). A new steering control of AGV by fuzzy control has been proposed instead of the PI control. However, it is necessary much time to investigate the rules and to adjust the scaling factors for excellent performance in the fuzzy control. Thus, this is the matter to be solved.
In this paper, a new steering control of AGV based on the neural network using the back propargation method is proposed. The good steering control results by the fuzzy control is adopted for the teaching signal of neural network. At the first step, the effect of the number of learning and the learning errors on the steering control results are discussed by the computer simulation using the AGV model. Further, the ability of generalization in the turning radius and the traveling speed is also investigated. It becomes clear that the AGV can travel along the traveling course provided that the neural network learns both of the right and left turning at the maximum traveling speed and the minimum turning radius. Then, it is proved by the experiment using the AGV built as a trial that the proposed steering control method is very available.

抄録全体を表示

A new method of optimal selection of cutting length of steel bars etc. using mutually connected Hopfield type neural networks is proposed. When each raw bar of various length is cut to several finished bars of required length, the combination of cutting length in a bar should be determined to keep the amount of scraps produced minimum as well as to maintain the balance among the products in accordance with the customer's order. The proposed neural networks can determine the optimal combination of length by searching the networks using energy function. The configuration of networks, the definition of network energy function and the derivation of the network constants, such as weight and threshold, are explained, along with the results of simulation studies.

抄録全体を表示

A method available for early detection of abnormality in an oil-filled transformer is discribed, in which four gas sensors having different characteristics and neural network are used to identify gas species (H₂, CH₄, C₂H₄, C₂H₂ and mixture of two species). In order to improve the selectivity of gas sensors, the time response patterns induced by changing sensor temperature, and the stationary sensor output is identified by neural network. Furthermore, the mixture ratio of gases is derived by using the stationary sensor output in response to the changing sensor temperature.
Gas species are well discriminated, and the mixture ratio derived from sensor output agrees well with the measurement by gas chromatography. Therefore, it is confirmed that our method is applicable to the transformer diagnostic technology.

抄録全体を表示

In excavating tunnels for subways or power cables, shield tunneling machines are used. The paper proposes a detection system, for the machines, which detects anomalous objects ahead, to prevent the cutters of the machines from getting damages and also to prevent the construction delay caused by the accident. The system uses one transmitting and several receiving transducers (both sonic) located on the cutter plane of the machines for getting information on the anomalous objects. To enable the system to be used on-line, the propagation from the input electrical signal to the output one through sonic wave is modeled by series combination of two second-order damped oscillation elements and one dead-time element.
Finally, the validity of the proposed system is demonstrated with experiments in an anechoic water tank and at a construction site.

抄録全体を表示

This paper proposes cooperative path control and its application to linear X-Y table. Usually, X axis servo system and Y axis servo system are designed independently for X-Y table. Therefore, when frequency responses of these control systems are largely different as compered with frequency components of desired signals, large path error is often caused. In this paper, area error is newly introduced as an index of path error and imaginary desired signal that minimizes the area error is used instead of usual desired signal. This cooperative path control system includes path error feedback loop and it is essential path control system. Then, it is very effective for path control of X-Y table or robot manipulator. This method is applied to linear X-Y table and effectiveness is confirmed by experiment.

抄録全体を表示

This paper presents a method to design a neural network for coin recognition by a genetic algorithm (GA). The GA specifies an architecture of neural network, but does not train the network. The back-propagation (BP) method trains the network. After training it by the BP, the GA varies the architecture of the network to fit the environment, which is to achieve a 100% recognition accuracy and to make the network small in size. The network reduced by the GA and the BP is further decreased by using the BP with forgetting of weight. The object of this paper is to design a smaller neural network for hardware implementation of coin recognition system. Results by computer simulation show the effectiveness of the method to variably rotated coin recognition problem.

抄録全体を表示

This paper reports a method of dynamic simulation parameter tuning for a coal-fired power plant using genetic algorithms (GA). GA is a search algorithm based on the mechanics of natural selection and natural genetics. GA is one of the effective methods for optimization problems and it requires a formalization of problems and a fitness function definition. Because the dynamic simulation can be repeated and the error between real data and simulation results is defined by a fitness function, it is reasonable to consider genetic operations. In order to satisfy the environmental laws and regulations, the control of NO_x emission is important in the field of power plant. We have, therefore, made a NO_x analytical model and used the model for the dynamic simulation at the change of loads, the kind of coal, and so on. In this paper, the optimal parameters for the simulation are determined by GA.

抄録全体を表示

This paper presents a scheduling method using a genetic algorithm for LSI assembling processes. The new method uses a new mutation named block exchange for reducing the number of exchanging tools for assembling machines. The new mutation directly handles schedules on the gantt chart. A performance index which considers both the total operating time and the number of tool exchange is introduced. The proposed scheduling method can handle multiple assembling processes with such constraints as minimum number of exchanging tools and/or minimum operating time.
Simulations using actual data of LSI assembling processes are done to verify the feasibility of the new method, and the results show that the new scheduling method is highly potential for field use.

抄録全体を表示

Recently, distributed systems have prevailed according to computer's downsizing and groupware functions such as electronic mail have been used.
In engineering departments, EOA (Engineering Office Automation) including CAE, CAD/CAM, CASE becomes a important subject for having common informations.
In U. S., concurrent engineering has been applied to improve productivity in DICE project, Boeing Co., H. P. Co., Raytheon Co. and so on.
This paper discusses concurrent engineering, groupware functions and their effect on Engineering Office System (EOS) and also proposes a unified multimedia environment in EOS which enables real time electronic conferences combined with non-real time electronic conferences on a regular WAN system. It assumes traffic based on operation analysis results of survery within our company and defines network architecture methods using narrow band WAN and broad band LAN systems and discribes their characteristics.

抄録全体を表示

Optimization of utility energy system operations at a paper and pulp manufacturing plant isstudied. The system consists of three boilers and two turbine-generators. The plant discharges a large amount of black liquid which is used as a fuel in the system as well as heavy oil. Steam energy consumed at the plant is produced by this system, while electric energy consumed at the plant is not only generated from the system but also purchased from an electric power company. The goal of the optimization is to minimize the amounts of heavy oil consumed in the system subject to restrictions of black liquid consumption, steam and electric energy requirements and ranges of system operations. Actual operations of the system is analized using the data obtained on every 3 hours. It is found from this analysis that energy characteristic of every unit contained in the system can be represented by a linearized model. This leads to an application of the linear programming to the optimization of the system operations. The results of this study show that in order to optimize the system operations, in general, a unit with higher efficiency is operated with priority to other units, while some inefficient operations are found subject to restrictions of the system operations. Some proposals are thus given; one is to enlarge the capacity of storage tank of the black liquid and the other is to reduce the purchased electric energy under some special operating conditions. It is concluded that a large amount of heavy oil can be saved by these optimized operations of the utility energy system.

抄録全体を表示

This paper describes quite a new type of ion drag micromotor. In this micromotor driving torque is delivered by an ionic current in liquid, though in usual motors generated by electronic current. The motor utilizes an ion drag force by drifting n-butyl alcohol ions accelerated by the applied field between electrodes. Applying special electrodes which are covered by evaporated SiO thin films on its half side and a n-butyl alcohl as an ion seed in an insulating liquid, we could realized a rotating motor with a DC power source without any switching circuit which is dispensable to many usual motors. The ionic current and the rotational speed of the experimental motor were nearly proportional to the applied field strength respectively. The maximum rotational speed of 40 rps was obtained at the applied field of 3×10⁶V/m. The rotational speed and the output torque of the motor were analized by the ion drag mechanism. The experimental results for rotational speed agreed well with the calculated values.

抄録全体を表示

Each kind of photovoltaic modules has its own way of response to each wavelength of solar radiation. Therefore, it is necessary to evaluate the spectral solar radiation in the area where the photovoltaic modules actually are used in order to optimize a photovoltaic power system.
Today, when simulating the output from photovoltaic modules, the efficiency of modules is regarded as constant in general. In fact, the efficiency changes with spectral solar radiation and other factors.
We have set up the photovoltaic modules, amorphous and poly crystalline, in our campus in downtown Tokyo and have been measuring the spectral solar radiation and the maximum power from those modules from March 1991 to February 1992.
This paper reports the changes of the spectral solar radiation by weather, time of a day and the seasons and the effects of these changes on the output from the photovoltaic modules to suggest a method of designing a photovoltaic power system.

抄録全体を表示

Uninterruptible power supplies (UPSs) as power sources in computer systems must be offer high quality and high reliable power to their system. To realize such UPSs, inverter controllers in the UPSs must be digitalized with micro processors. And, it is necessary to obtain high sampling frequency and satisfactory compensation for the processor operating time lag and disturbances in the control system. Besides, simple control algorithm is demanded due to application of the digital controllers to low cost UPSs. Then, a control method using full-order observer in the current minor loop is proposed. In this method, to shorten the algorithm, disturbances are assumed as a state valuable, v_D(n) where v_D(n+1)=v_D(n), and the observer is established in the minor loop. To increase the compensation for the time lag and the disturbances, the observer estimates the time-lagahead current and disturbance, and the observer carries out dead beat operation. Experiments demonstrates that this control method provides excellent characteristics of the inverter for the UPSs.

抄録全体を表示

We propose a robust force control strategy based on observers which compensate disturbance and parameter variations of motor and environment. We present two methods: one is a 1-observer system (using force-based disturbance observer), the other is a 2-observer system (using position-based disturbance observer and environment observer). The 1-observer system is simpler than the 2-observer system and does not need position sensor. However, by using 2-observer system we can determine the sensitivity to actuator disturbance and environment disturbance independently, and can achieve higher precision control. Second derivative of force can be controlled because the nominalized motor and environment are realized by the observers. Robust and fast force control is realized by this strategy. Moreover, an adaptive control is introduced to obtain better performance for wide parameter variation. The parameters of the force controller and the observers are adjusted by an identified stiffness obtained from a least square estimation. By using the identified value of stiffness as a nominal value, the response of the force becomes closer to the force command. We also examine influence of resolution of the position sensor on the performance of the observers. We show the effectiveness of the proposed method by simulations and experiments.

抄録全体を表示

Magnetic Signal Transmitting Method (MST) aims at transmitting the information of underground by radio to work efficiently in drilling work, investigating the nature of the soil.
MST is a system that makes used of its own boring rod for signal transmitting.
A concept of MST is the same of a transformer, but it needs the original analysis method, because its flux path is very long, structure is open flux path, and receiving side makes used of magnetic characteristics of low magnetic field.
In the present paper, we explained the fundamental structure of MST and the experimental model, and led an experimental expression of output voltage in receiving coil. Deciding the parameter of the experimental expression, we measured relative permeability in low magnetic field and considered an influence by the depth of penetration of magnetic flux. Moreover we measured the output voltage of receiving coil, verified the experimental expression, and searched the theoretical expression of transmitting distance to calculate the transmitting distance out of experimental expression, and then searched the signal transmitting distance when it applied to boring rod.

抄録全体を表示