Generally semiconductor manufacturing consists of three consecutive stages. The first one is Wafer Process, which repeats several types of manufacturing process and the total number of the process step reaches several hundreds. The total number of process facilities is nearly 1,000 with several substitutive machines. Therefore it is very difficult to optimize the production schedule. And the second one is Assembly, which is not complicated comparing with Wafer Process, nevertheless, the machine operation is really peculiar. For example, set up time is much longer than processing time in most facilities. Finally the last one is Final Test, whose process frequently changes according to the product quality like production yield, and it is almost impossible to create the optimal production schedule. In our previous researches we applied the well-known cooperative scheduling method, Lagrangian Decomposition and Coordination method, into the actual scale Wafer Process model. Then we have verified that the proposed method can create feasible schedule effectively. In this paper we apply the proposed method into an actual Assembly production line. First we try to confirm the effectiveness of the proposed method in the simple model without considering set up time. Then we evaluate the performance of the proposed method with more complicated model including set up time. Finally we try to verify the effectiveness of the proposed method by comparing with conventional dispatching rule and the feasibility with the improvement of a list scheduling algorithm.

抄録全体を表示

Kanban system has been known to be a superior tool to control complex production and inventory systems operating in the just-in-time (JIT) environment. The kanban system works more effectively at low inventory levels when the consumption rates of parts necessary to produce finished goods are leveled. The parts consumption leveling in JIT systems is called “Heijunka”. In this research, we investigate a multi-stage multi-product production and inventory system in which the production order quantity is exponentially smoothed. The paper presents an efficient algorithm based on simulation to determine the near optimal exponential smoothing factor and safety stock level at each workstation. We use numerical experiments to evaluate the performance of the exponentially smoothed ordering policy in a supply chain as compared to the original kanban system. And also the performance of the two types of optimization methods, global and local optimizations, in the supply chain are compared. We conclude that the locally optimized production smoothing is quite effective as a production and material replenishment ordering policy as an SCM.

抄録全体を表示

Industries need to evaluate productivity and energy consumption concurrently when designing and improving manufacturing systems. Manufacturing system simulation to evaluate productivity was often used. However, manufacturing system simulation to evaluate productivity and energy consumption concurrently was not proposed. The purpose of our research is to establish manufacturing system simulation for productivity and energy consumption. In this paper, first, requirements for manufacturing system simulation are analyzed. Second, functions considering analyzed requirements are proposed. The functions are a state transition model for facilities, a simulation system, and an evaluation system. Third, a manufacturing system simulation environment which is implemented the functions is also proposed. Finally, case studies for a small scale machining line and a middle scale semiconductor manufacturing line are carried out to confirm the efficiency of our proposed simulation.

抄録全体を表示

A planning method for multi-stage logistic system is proposed to reduce total logistics costs and stock-out under uncertain demand. Mathematical model is constructed to decide logistics between different locations and inventory at locations. The proposed method is applied to decide logistics plan in continuous time periods as rolling planning method. The problem involves two types of significant measures: logistics costs in the system and stock-out at shops. In order to reduce total logistics cost, virtual logistics process generated through plural periods is considered in the mathematical model. In addition, inventory control at shops is included to reduce probability of stock-out at all shops in the model. Strategy to control inventory for reduction of probability of stock-out is applied under the condition logistic cost is reduced to execute rolling planning in multi-periods. The proposed method is compared with Pull system as the conventional procedure to evaluate the effectiveness of the method. Further, characteristics of the proposed method are investigated on simple data using probability distribution of demand in practical use.

抄録全体を表示

This paper deals with an alignment device for micro parts, which have a size in the order of 100 micrometers, by using a board stage that is vibrated in the natural frequencies of the various vibration modes. Part alignment is important for manufacturing processes, such as the assembling of small parts and the inspection of parts. Since it is difficult to grip and handle micro objects, a simple and reliable method is required. In this paper, the design concept and guidelines to realize such a device are presented. In addition to the theoretical representation, an example stage was designed for the alignment of the micro parts. It was also analyzed by Finite Element Method (FEM) simulations. Also, the designed vibration stage was fabricated using a stainless steel thin plate. Experimental results showed that the actual vibration is similar to both the theoretical results and the simulation results. As an example of the alignment of the micro parts, some micro-seize electronics parts were tested.

抄録全体を表示

The estimation of steering characteristics such as steer feeling by using only computer simulation technology is extremely difficult. In this paper, to solve this problem, we consider the reproduction of behavior of steering wheel and the control method with developed computer simulation and steering simulator when the road surface condition changed suddenly. Considering the sudden change of the road surface condition to be disturbance, the disturbance compensator for suppression control is designed. In addition, we propose a controller design procedure for electric power steering (EPS) systems to develop the control method efficiently. First, the human model of driver for the computer simulation of EPS is constructed from the investigation of vehicle dynamic behavior by the steering simulator test or actual car test. Next, we design outline of the disturbance compensator by the computer simulation using the human model. The disturbance compensator is also designed in detail with experiments of the steering simulator, and it will be designed finally by experiments of the actual car.

抄録全体を表示

This study proposes a new control method for under-actuated planar satellites having body-fixed thrusters that can generate unilateral and constant forces. The proposed method is applicable when the thrusters are not balanced, which means a rotational motion is inevitably induced whenever the thrusters are turned on. By applying the method, the position and attitude of any planar satellite can be controlled to arbitrary target states so long as a controllability condition is satisfied. The control logic is based on a switching control using plural manifolds. First, the satellite is transferred to an initial manifold, and then it moves to another manifold while decreasing its velocity and approaching the target states. The ON/OFF timing of the thrusters to transfer among the manifolds is explicitly specified in this paper. A numerical simulation validates the proposed control logic.

抄録全体を表示

A surveillance system has been required to gather information after large-scale disasters on stricken areas safely and quickly. To collect the information, it is useful to use a blimp, because of its low-sky availability, safety and long flight. A blimp flying at low altitude can capture high quality images offering 3D imagery in disasters. This paper proposes a new path following control method for outdoor blimp robots under windy conditions to collect detailed disaster information. The method consists of path following control in the wind coordinate on the horizontal plane and altitude control on the longitudinal plane. And this method minimizes the lateral errors arising from the wind disturbances considering path regeneration. Some simulations and experiments for a 12m class blimp are performed to confirm the usefulness of the proposed method.

抄録全体を表示

This paper proposes a level control of a carrier platform on a cart with an active vibration reducer of a 3-degree-of-freedom parallel link type. The carrier platform is supported with three links, which are connected to each linear actuator. The pitch angle, the roll angle and the vertical displacement of the carrier platform are controlled by these linear actuators. Even when the cart runs on uneven grounds, the carrier platform is always kept levelly and the vertical displacement of the platform is set to zero with the active vibration reducer. A method to measure the vertical displacement of the cart by using cheap sensors is proposed. The usefulness of the proposed control system is demonstrated through experimental results.

抄録全体を表示

Quadruped animals switch gait patterns depending on speed to minimize energy consumption. This is similar to the phenomenon that excited natural vibration modes are switched depending on excitation frequency in a multi-degree-of-freedom system. Therefore, in this paper, it is assumed that quadruped animals move by using the natural vibration of their own musculoskeletal systems. To verify this idea, we made an experimental apparatus that has a simple structure modeled after the actual musculoskeletal system of a quadruped animal. The apparatus consists of a body, a shoulder, a waist, and four legs. Each leg is excited by a DC motor fixed to the shoulder or waist. First, it was clarified that the apparatus has natural vibration modes similar to the gait patterns (trot, pace, and gallop) of quadruped animals by free vibration tests. Next, the apparatus was driven by forced vibration in which four DC motors were excited with the phase angle of gait patterns. The energy consumption of the motors was minimized when the excitation frequency corresponded to the natural frequency. Finally, the apparatus was driven by self-excited vibration generated by decentralized control in which four DC motors controlled locally are used. The vibration modes were similar to trot, pace, and gallop patterns. The frequencies of self-excited vibration corresponded to the natural frequencies.

抄録全体を表示

In the precision measurement and positioning fields, e.g. the semiconductor exposure machine, an air type anti-vibration apparatus is widely implemented. The above machine must continuously operate because of a high price. However, when an earthquake occurs, in almost case, the anti-vibration apparatus becomes shutdown in order to avoid mechanical contact. In order to operate continuously, we previously proposed a switching control scheme by using the changeover concerning floor vibration feedforward and/or feedback parameters. In order to be applicable for various earthquakes, we propose a modified switching control scheme such that the acceleration, displacement, and two kinds of SI values are simultaneously utilized for both switching and recovering. Moreover, for massive earthquakes shutdown switch is prepared. Simulated results using earthquake examples show the effective of proposed scheme.

抄録全体を表示

Bottom-up processes have received much attention in unsupervised and developmental learning research domain. In contrast, effectiveness of top-down deeming on acquisition of adaptive behavior is discussed in this paper. Successful experience in the past, or a skill that could be expected to be reused successfully in a novel environment is stored in memory. Then abstract environment recognition via geometric transformation invariance is introduced to measure the reproducibility of executed skill in a novel environment. Additionally, reproducibility of skill in the environment is utilized to make up intrinsic motivation that drives the agent to active conceptualization of search space. It enables the agent to relativize current skill execution robustly in diverse environments. Useful characteristics of top-down deeming process are implemented on reinforcement learning and discussed through simulation experiments in grid world. The results demonstrate acceleration of learning progress by active conceptualization of environment. Additionally, it is shown by experiments for scaled environment that subjective anticipation could bring in consistent strategy of exploration and exploitation. Eligibility trace is also introduced for skill utility problem and it is shown that the traces regarding actions and skills could preserve learning performance for diverse skill settings.

抄録全体を表示

Traditional capacitance sensor sheets usually have a three-layered structure, with a dielectric layer sandwiched by two electrode layers. Each electrode layer has a number of parallel ribbon-like electrodes. The electrodes on the two electrode layers are oriented orthogonally and each crossing area of two crossing electrodes makes up a capacitance sensor cell on the sheet. It is well known that compatibility between measuring precision and resolution is difficult since decreasing the width of the electrodes is inevitable to obtain a high resolution, however, this may lead to reduction of the area of the sensor cells, and as a result, lead to a low S/N ratio. To overcome this problem, a new multilayered structure and related signal processing approach were proposed. This new structure stacks two or more sensor sheets with shifts in position. Both a high precision and a high resolution can be obtained by combining the signals of the stacked sensor sheets. Trial production was made and the effect was confirmed.

抄録全体を表示

Recently, HILS(Hardware in the Loop Simulation) has been utilized in the industrial world for the improvement of the product quality. HILS system requires the real time simulation for the MBD model. Fujikawa proposed a dynamic response analysis technique using the domain decomposition method(DDM) connecting with the elastic elements. This technique ensures the numerical stability and the accuracy of the numerical calculation by introducing the iteration into the constant external force method in each time step. This paper proposes an improved domain decomposition method(IDDM) for more fast simulation of the actual flexible multibody dynamics. In the Fujikawa's technique, the predicted displacement was assumed to be the displacement of the connecting point. In the IDDM, Taylor developing is employed to the predicted displacement in order to reduce the iteration number. This paper also proposes the fast simulation technique of the IDDM for coupled system which consists of three subsystems with different frequency characteristics by employing different time step for each subsystem. It is clarified that the computing time using the IDDM can be reduced on the application for the dynamic simulation of the actual digging operations of the hydraulic excavator.

抄録全体を表示

Impact vibration of cantilever beams with collision wall, that is set with almost no clearance, was treated as an example of impact vibration of strongly unsymmetrical system. Vibrations in the wide excitation frequency range from near 1st natural frequency to near 2nd natural frequency were investigated. The followings were made clear from the numerical analysis: (1) Near 2nd natural frequency, the internal resonance between the 1/7-th sub-harmonic vibration of 1st mode and the primary resonance of 2nd mode is occurred. Because the vibration has a feature of the 1/4-th sub-harmonic vibration, vibration amplitudes become large at the excitation frequencies near 1/4, 2/4 and 3/4 times of 2nd natural frequency. (2) The even order sub-harmonic vibrations of 1st mode occur. (3) The amplitude of the vibrations mentioned at (2) becomes large, if its frequency is close to those of the vibrations mentioned at (1) occur. It is supposed that the energy of the primary resonances of 2nd mode flows into the even order sub-harmonic vibrations of 1st mode through the odd order sub-harmonic vibrations of 1st mode.

抄録全体を表示

As a model for soundproof structures in car bodies, we deal with double walls composed of a porous material sandwiched between a damped steel plate and a cover plate. To analyze vibration transmission between the double walls in the low frequency range, we apply a numerical method previously proposed by the authors using three-dimensional finite element method. In our numerical method, we select displacement vectors as common unknown variables for the soundproof structures including porous materials, viscoelastic materials and elastic materials. We solve under coupled condition. Further, approximate expressions of modal loss factor for the soundproof structures are derived using asymptotic method. Using these techniques, eigenvalue analysis and frequency response analysis are performed. Especially, we focus on the influences of beads in the damped steel plate on vibration transmission from the steel plate to the cover plate. The porous material on the top of the bead in the steel plate is removed locally corresponding to the height of the bead. Because of short distance between the steel plate and the cover plate in this region, vibration transmission through this region increased.

抄録全体を表示

Recently, it has been demanded for the railway companies to construct extremely high soundproof walls by such reason that many very high residences have been constructed along railway lines. However, the higher the soundproof wall is, the greater the load and moment acting on the walls by natural wind become, and it is hardly possible to install such extremely high soundproof wall without reinforcing the viaduct drastically. Therefore we have developed a new type of soundproof plate. One edge of the plate is equipped with a rotary axis and all or a part of the other three edges are fixed by a magnetic attracting force. Accordingly it has a high soundproof ability in the ordinary state; on the other hand, when the wind load amounts to a limit of keeping safety of the viaduct, the panel is opened and the wind load is reduced sufficiently. We have made clear the performance of the new plate by mechanical analysis and experiments.

抄録全体を表示

Our proposed actuator was applied for a spin-stand tester which evaluated high areal density magnetic recording. Since, the actuator characteristic was a second order lag element that had large amplitude at a resonance frequency. However, the peak value of the resonant frequency is very high because the mechanism uses elastic deformation. As a result, a good performance cannot be guaranteed for a perturbation induced by setting a load mass. Therefore, we have attached a damper made from VEM (visco elastic material) and a restriction board on the lever mechanism to reduce the resonant peak value. In this research, the characteristic changes induced by the damper are modeled. Further, by using the model, the influence of the damper on the transient response of the precise positioning mechanism is discussed. Moreover, a controller which improves the transient response is designed and the effectiveness of the controller is illustrated by experiments.

抄録全体を表示

In the field of vibration control, Direct Acceleration, Velocity and Displacement FeedBack (DAVDFB) is desirable according to the skyhook control theory. We have proposed a first-order differential type seismic displacement sensor which can simultaneously detect absolute acceleration, velocity and displacement signals. Due to low detection sensitivity and narrow bandwidth, however, its control performance is not enough for practical use in industrial scenes. This paper presents a high sensitive method by increasing the number of coil turns, because back electromotive force of a calibration coil is utilized as the velocity signal. Sensitivity calibration and practical control tests are conducted as to ascertain the efficacy using an anti-vibration apparatus.

抄録全体を表示

This note presents an application of reinforcement learning to caterpillar robot locomotion. An excellent advantage of reinforcement learning is that an action can be acquired using only a simple reward. In our previous work, the reward was a forward distance measured using a sensor. This reward was completely an “Objective Reward.” On the other hand, this study uses the reward given by the human's subjective judgment, which is defined as “Subjective Reward.” The main purpose of this study is to compare its performance between the “Objective Reward” obtained from the sensor and the “Subjective Reward” given by a human teacher. The results show that the “Subjective Reward” can give better results than that of the “Objective Reward”, because the “Subjective Reward” has more information than the “Objective Reward”. On the other hand, this note discusses the good teacher who gives an excellent “Subjective Reward”

抄録全体を表示

The present paper describes reduction of work piece shift of helical gears with two number of teeth and large helix angle during form rolling. In-feed form rolling using round dies was employed to form such gears. Two different kinematical models, that is, a friction roller pair and a helical gear pair, were considered to reduce work piece shift. Using these models, work piece shift could be explained as a differential motion of translation caused by the different rotation speeds of work piece in initial and finish form rolling stages. Round dies should be designed so that the die tips can accurately divide work piece root circumference on which the initial threads have formed into desired number of teeth. A series of form rolling using the round dies was performed. As a result, the round dies reduced a work piece shift, and brought high form accuracy of helical gears.

抄録全体を表示

Recently, mirror-like surface on a mold and a die was required for adding high quality on the most products. Particularly, as the processing of the mirror-like surface on the mold or the die with small and complex shape is difficult, the processing is performed a very long time by polishing of the skilled worker. Automatic polishing for the products is very difficult. Therefore the polishing with high efficiency on a small mold or die with complex shape was developed and evaluated. The three polishing tools were firstly developed for the experiment; the first polishing tool is the tool for the high speed polishing on the area in the die, the second tool is the tool using a contour line for the middle speed polishing on the line in the die and the third tool is the tool using a small ball (less than 1 mm diameter) for polishing on the point in the die with small and complex shape. The optimum tool for each size or shape in a die and its optimum polishing condition are investigated for high efficiency in several experiments. The polishing tool is installed in the spindle of the CNC milling machine. This system was evaluated for high speed polishing the mirror-like surface on the small parts with complex shape. It is concluded from the results that the three tools and its polishing process were effective for processing the mirror-like on the surface of the small parts with complex shape, and optimum polishing condition was also cleared.

抄録全体を表示

In magnetic hard disk drives, it is important to evaluate head wear durability in near contact regime because the head-disk spacing has to be reduced to ～1nm to achieve a high density recording above 1 Tb/in². This paper presents an evaluation method of the replenishment speed of a mobile lubricant scar caused by asperity contacts. The replenishment speed of sub-monolayer lubricant into a removed scar was analyzed by using the two-dimensional long wave equation on the basis of continuum mechanics. Then, a generic evaluation chart to estimate the removed lube scar width with 50% replenishment after one disk rotation was proposed in relation to the thickness and viscosity of the lubricant. Next, the effective lubricant viscosity of an actual thin lubricant layer was identified by comparing the measured lubricant replenishment process and its numerical simulation. From the analysis of contact pressure and width for roughness and defect asperities of disk surface in addition to the above results, the minimum possible head-disk spacing with 50% lube replenishment for reliable read-write operation and critical asperity contact conditions resulting in head wear and medium scratch are discussed.

抄録全体を表示

This paper evaluates CO2 emissions generated during the manufacturing phase of global production network and discusses requirements analysis of green sustainable policies of global production network to prevent carbon leakage arising from transnational production. It calculates CO2 emissions from electricity used in manufacturing, emissions during transport from mining to manufacturing countries and emissions during transport from manufacturing countries to market. A case study of electric vehicle (EV) production includes outcomes calculated CO2 emissions in the manufacturing phase and transportation phase: for production in China, India, Thailand, Germany, Japan and Mexico to market in Japan, German and Singapore. It is illustrated that CO2 emissions during manufacture of EVs vary significantly, depending on the carbon intensity from electricity generation and each country's technical efficiency, and transport has a minor influence on emissions.

抄録全体を表示

Recently, a new academic field, ”service engineering” has been very actively investigated. However, there are few effective methods and tools to simulate and evaluate services designed based on the concept of service engineering. In the past, the authors proposed a service flow simulation method using scene transition nets (STNs) which is a graphic modeling and simulation method for discrete-continuous hybrid systems. However, this method does not consider how to model and simulate complex service flows including multiple layered, parallel, and interrupted structures and it is difficult to construct STN models of such complex systems using the existing STN concepts. In this paper, the authors propose a new STN modeling method using“multi-aspect STN modeling”concepts in order to provide easy methods for modeling of such complex services. The experimental results for the modeling and simulation of a nursing service demonstrate the effectiveness of the proposed method.

抄録全体を表示

This paper proposes a new method called Adaptive Range Differential Evolution (ARDE). The basic idea can be obtained from the Adaptive Range Particle Swarm Optimization (ARPSO). In the ARDE, the active search domain range consists of the mean and the standard deviation of each design variable. Also the best position is included in this search range. Therefore, the active search domain range is newly defined by utilizing the mean and the standard deviation of each design variable, and the best position. The detailed procedure of the new active search domain range is described in this paper. To keep the best position for the new active search domain range will lead to the wide search range. Of course, the newly active search domain range will shrink through the search iteration. As the result, a highly accurate global minimum can be found. The effectiveness and validity of the ARDE are examined through typical benchmark problems. It could be found through benchmark problems that the ARDE can find a global minimum with the small number of function evaluations in comparison with basic Differential Evolution.

抄録全体を表示

Human beings feel sound in various ways, depending on the environments. For instance, one hears car's engine sound while driving the car in the way different from while hearing it coming from outdoors inside the house. In this paper, the analysis of brain activation to various sounds was conducted using near-infrared spectroscopy. It turned out that the listening to music and uncomfortable sound decreased oxyHb in the prefrontal cortex. This reaction can be considered the response to the concentration on the sound. It also tuned out that there is a viewpoint that the decrease of oxyHb in a brain does not necessarily mean the the decline in the neural activity but the brain activation reaction. It is expected to qualify audio information extracted from these two regions and to serve the plan of sound design.

抄録全体を表示