The position measurement methods for robotics vehicles navigation have been proposed by emitting various laser beams into various fixed retroreflectors. Authors thought up a position measurement method, emitting only one laser beam into a retroreflector, through introduction of a reference axis. The method is based on the measurement of an angle between the laser beam and the reference axis. To construct the reference axis through the standard point on the site, a double corner-cube is proposed here. It is shown that the angle between tracking beam and the reference axis can be measured continuously with utilizing the characteristic of the double corner-cube while tracking the standard point by laser beam.
This paper considers a single machine scheduling problem in just-in-time production environment under the constraints of prespecified ready and due times. That is, every job is available for processing on the ready time of each job and cannot be processed before the time. Also, every job must be completed before or just on the due time of each job, and no tardy jobs are allowed. An optimization algorithm for the scheduling problem based on the Branch-and-Bound method is proposed to effectively determine the optimum solution so as to minimize the total weighted earliness with respect to the due times as objective function. Although the computational time through the proposed algorithm increases exponentially to solve problems by increasing the number of jobs, the branching operation and strong lower bound for the proposed algorithm will enhance the computational efficiency. The effectiveness of the proposed algorithm will be also shown by some simulation results.
For the free-flying space robots, normally, the total momenta of the robot are written to obtain the desired kinematic model. In this paper, an approach based on the concept of the Primary Body (PB) is presented to derive a set of generalized expressions for the total momenta of the free-flying space robots, consisting of a serial manipulator mounted on a free-base, e.g., a satellite. The generalized expressions help in deciding upon a suitable model. Thus, the choice of the satellite as the PB leads to the definition of the well-known Generalized Jacobian Matrix, which is important when the motion of the satellite and the end-effector are to be calculated together. However, to control the end-effector only, the concept of PB suggests that the end-effector, in contrary to the satellite, should be selected as the PB that will give more efficient algorithm.
In the theoretical analysis of genetic algorithms, WST (Walsh-Schema Transform) is a well-known method. WST is useful when we analyze genetic algorithms from the viewpoint of schema. However, WST can be used only when the gene is binary : the gene consists of only two types of symbols. We propose WST-like method called OSD (Orthogonal Schema Decomposition) which can be used also for the non-binary gene. In this paper, the definition and the meaning of OSD, the fast algorithm for performing OSD, the relation between OSD and WST and some usages of OSD are shown.
This paper deals with a new manufacturing system for single batch production that will be a typical style in the next generation. Manufacturing data are divided into three types to reduce recalculation efforts against unpredicted events in the factory. They are product oriented data, facility oriented data and product and facility consideration data. The roles of process and operation planning, which have not been clearly defined yet, are distinguished according to the data separation.Due to this concept, a method to generate assembly plans is developed. First, representations of the plans and the manufacturing data for assembly are discussed. Then, a way to find all possible assembly plans with the divided manufacturing data is developed. A computational result shows the advantages of our method over the existing ones.
One of the most essential requirements in the flexible automation is to increase the product flexibility with reducing the heavy load of CAD/CAM systems. The keywords to realize the requirement seem intelligent and distributed information processing. This paper proposes the feasible solution to realize the intelligent and distributed production systems in order to increase the flexibility especially against the errors and failures of the factory floor. The flexible cell assignment to execute the given jobs and the dynamic scheduling to re-schedule quickly when required are the important functions for the factory management. The former is realized by generating flexible sets of machining features in CAD, and non-deterministic cell assignment according to the functionality/ performance and real-time status of cells. The latter is realized by performing operation planning in the cell controllers with use of the real-time simulation and dynamic scheduling system. An example reveals feasibility and availability of the proposed control/ management architecture of the production system.
An autonomous navigation method based on TV vision system on board a vehicle is under investigation. Our method is to use fluorescent lamp arrays on the ceiling as a lighthouse to control the vehicle. The object of this paper is to propose a method which controls the vehicle to turn at 90 degrees and to show the vehicle motion on the specified course in a virtual factory. This paper consists of three parts : (1) First, the basic principle of turning the vehicle is explained. Next, experiments are carried out to examine this turning method using the practical vehicle. Furthermore, a method to specify the turning direction is suggested. (2) In the next section, the vehicle motion in the perpendicular direction to fluorescent lamps on the ceiling is outlined. (3) In the last section, the vehicle motion on the specified course in the factory is shown by using computer simulation.
It is pointed out that the locomotion rhythm plays an important role in generation of the periodic motion such as walking successfully. This paper is concerned with a control strategy for generating the regular locomotion rhythm. The method has two features : one is that the locomotion rhythm is determined by specifying the reference trajectory and some time data on it, and then the trajectory following servo realizes the walking motion according to the determined rhythm. The other is that the kicking motion is utilized for generating the regular rhythm so that the kinetic energy loss caused by the collision between the swing leg and a floor is compensated. From the experimental results of our new biped locomotion robot SR-4, it is confirmed that the proposed method is useful for generating the regular locomotion rhythm.
This paper presents a method for designing a set of rules for a general knowledge-based controller. For this purpose, we first introduce a representation and analysis scheme called influence diagram from a field of the decision theory, and provide a generic model for a sensor-based control. We show that a rationally-designed control rule set can be obtained by performing probabilistic reasoning and by compiling the solutions obtained from this model. Next, as a practical application, we apply the method to the design of the knowledge-based control system for a wastewater treatment (i. e., sewage disposal) plant. We attemp to explicate the expert operator's cognitive map concerning with this plant by influence diagrams and show that our method derives a set of control rules that is rationally designed for establishing the coordinate control of multiple processes. Finally, the derived control rules are applied to an actual plant for the sewage disposal and the control result is shown.
A new approach using an analog type Hopfield neural network to solve line balancing problems for manufacturing planning is proposed. The energy function of the network to evaluate solutions is composed of three terms ; (A) an operation should be processed at one and only one workstation, (B) the precedence relationship between two operations should be satisfied, and (C) the cycle-time of operations should be minimized. It is shown from the computer simulation that the network can successfully solve the line balancing problems. Furthermore, several techniques to adjust the network parameters are applied for the improvement.
A new functionally partitioned scheduling system has been developed to improve efficiency for cast-to-roll scheduling in steel sheet manufacturing. It is equipped with a two-stage algorithm, namely a, macro-scheduling and a micro-scheduling algorithm, which can solve a multiobjective assigning and sequencing problem in several tens of seconds. To achieve better efficiency of scheduling operation, it was divided into a human component and a computer component. The latter of which was further divided into a workstation component and a host computer component. This functionally partitioned but strategically integrated system, which can create a rather reasonable schedule, has helped operators a great deal and has contributed to drastic reduction of planning time.
This paper describes an application of Case-Based Reasoning (CBR) system for a design problem of Nb-Ti superconducting wire. The developed system has two distinctive features : (1) Basic similarity degree between cases is derived using fuzzy matching. (2) Additionally, another similarity degree is defined considering the basic similarity degree and case modifying knowledge base. This similarity degree, which is named extended similarity degree, is the only criterion when selecting a past case similar to the given problem. Therefore, the case selecting function depends on the case modifying knowledge base. Moreover, it is shown that the case defining function is also effected by the case modifying knowledge base.
This paper deals with a scheduling and re-scheduling method for flexible manufacturing systems (FMS) using automatically guided vehicles (AGVs) and input-output buffers. The purpose is to solve coordination problems of AGVs transferring jobs between machines so as to minimize idle time of each machine and the total job processing time without collision of AGVs, and quick generation of another substituting effective schedule in the case of troubles of AGVs. A hierachical timed Petri net is applied to modeling and analysis of systems. Two priority functions are defined to effectively determine job processing precedence relationships among machining orders and AGV assignment. Some scheduling simulation results are also shown in this paper.