A static output feedback control having positive definite gains is known to give a certain robust stability to a class of collocated space structures. This paper studies an optimal design for static output feedback gains which minimize the H_∞ norm so that the closed-loop system has the prescribed disturbance attenuation capability. For this purpose, two synthesis methods via linear matrix inequalities are proposed. Then, considering structural robustness, the construction of a local feedback control system is investigated. System features include a set of local controllers with a simple structure and the high fault tolerance for preserving closed-loop stability. The validity is numerically shown using a simple uniform beam model.

抄録全体を表示

The fixed thrust optimal-fuzzy combined formation control method is mainly studied for spacecraft formation flying in circular orbits based on two-body relative dynamics. The linear Hill’s equation is used in relative dynamics, and the circular formation in three dimensions is designed. In fuzzy control, relative position, velocity and acceleration errors are selected as fuzzy variables and the multiplication and weighted average principle are introduced into fuzzy inference so as to maximize the reduction of the coupling effect. And since J2 perturbation is known, it will be taken into account effectively in fuzzy control method. In time-fuel optimal control, the simplified relative error model is used and an optimal controller is provided through the minimum principle. Thus, this paper combines both controls above to form the optimal-fuzzy formation control. Optimal control is adopted when relative position errors reach a certain threshold, while fuzzy control works for the small errors lower than that threshold. Finally, to take circular formation as an example, the simulation results demonstrate that not only does the optimal-fuzzy combined control system have a good response performance, but also excels in fuel consumption.

抄録全体を表示

The fundamental characteristics of simplified clustered aerospike nozzles are analyzed by the computational fluid dynamics approach. Several types of aerospike nozzles, having 6, 12 and 24 inner nozzle modules with the same area ratio and total flow rate, are examined. The interactions of the exhaust flows from the neighboring modules create shock waves, which produce high-pressure regions on the nozzle surface. The base regions of the cluster-type aerospike nozzles are not influenced by the clustering of the modules and show features similar to axisymmetric-type aerospike nozzles. The base pressure is nearly equal to the environmental pressure when the pressure ratio is low and suddenly attains a constant pressure when the pressure ratio is high. The computed results show that the number of modules influence the thrust performance, and the major reason for the decrease in thrust performance with a smaller number of modules is due to thrust loss in the ramp region.

抄録全体を表示

The detailed aerodynamics of a shrouded tail rotor in hover has been numerically studied using a parallel inviscid flow solver on unstructured meshes. The numerical method is based on a cell-centered finite-volume discretization and an implicit Gauss-Seidel time integration. The calculation was made for a single blade by imposing a periodic boundary condition between adjacent rotor blades. The grid periodicity was also imposed at the periodic boundary planes to avoid numerical inaccuracy resulting from solution interpolation. The results were compared with available experimental data and those from a disk vortex theory for validation. It was found that realistic three-dimensional modeling is important for the prediction of detailed aerodynamics of shrouded rotors including the tip clearance gap flow.

抄録全体を表示

System identification techniques with adaptive gyroscopic damper systems were proposed by applying the capability to change the moment of inertia around its gimbal axis, under a concept of self-identification with adaptive structure systems. The techniques to identify variations of modal parameters were proposed by sensitivity analyses of the natural frequency and the frequency response amplitude with respect to the modal parameters. Further technique to identify the modal parameters was proposed by extending the formulation with sensitivity of the frequency response amplitude with respect to the moment of inertia. To demonstrate the feasibility of each technique, numerical examples with cantilevered beam models were treated. The relations between the errors and non-dimensional system parameters, which include variations of modal parameters, the moment of inertia and the excitation frequency, were investigated numerically. The results indicated that the errors were sensitive to the relation of the system parameters and had singular points where the moment of inertia was close to optimum one to vibration suppression, and the excitation frequency was near the resonant frequency of the beam or the gimbal. The errors at the singular points were reduced by adjusting the moment of inertia or considering the excitation frequency.

抄録全体を表示

The US GPS and the planed Quasi Zenith Satellite System (QZSS) will enhance the capability of quickly resolving the integer cycle carrier phase ambiguities in precise differential positioning. The paper describes numerical analysis of the positioning performance of combined GPS-QZSS system. The current ambiguity resolution method is employed to resolve all cycle ambiguities of combined GPS-QZSS system. The performance of ambiguity resolution on a short baseline of 1 km in Tokyo is analyzed for different scenarios of the present GPS and combined future GPS-QZSS system. It is also analyzed in the Asian cities of Seoul, Beijing and Shanghai. The ambiguity fix percentage is adopted here for evaluating the performance of ambiguity resolution. It indicates that increasing the number of satellites has a benefit on the capability of getting quick ambiguity resolution. It also indicates that the ambiguity fix percentage with both GPS and QZSS is much higher than that with only GPS on the short-baseline operation.

抄録全体を表示

An airship has usually two or three ballonets in its envelope in which air is contained. Its buoyancy and attitude control is performed by changing the air content of each ballonet. It is said that ballonet slosh may influence an airship’s stability or ride quality. However, no quantitative treatment has been performed so far to investigate this phenomenon. In this paper the coupled equations of an airship longitudinal motion are formulated by modeling the ballonets as cylindrical containers. Some numerical calculations are performed for a 25 m class airship and it has been shown that the ballonet slosh may become a design issue when the shape of the ballonet is thinner or when the ballonet size becomes larger.

抄録全体を表示

The objective of the present study is to demonstrate performances of Evolutionary Algorithms (EAs) and conventional gradient-based methods for finding Pareto fronts. The multiobjective optimization algorithms are applied to analytical test problems as well as to the real-world problems of a compressor design. The comparison results clearly indicate the superiority of EAs in finding tradeoffs.

抄録全体を表示

Thrust and torque generated by a model rotary wing were measured at an ultra-low Reynolds number, Re=4×10³, for various aspect ratios with and without linear blade twist. The measured characteristics were compared with those calculated by the method which is well known to be effective for analyzing a rotary wing at a high Reynolds number. The method combines annular momentum theory and blade element theory. This calculation method can give the quantitative explanation of the effects of the aspect ratio and of linear blade twist on the characteristics of the rotary wings. The calculation results also indicate that the present calculation method has the capability of giving an accurate quantitative estimation of rotary wing performance with the blade aspect ratio larger than 10, operating at the ultra-low Reynolds number.

抄録全体を表示

The rest-to-rest maneuver problem of a flexible space structure is a two-point boundary value problem (TPBVP) and is solved by some gradient methods. If TPBVP is strongly restricted by constraints, TBVP becomes an ill-defined problem, and the solution meeting all constraints cannot be obtained. However, reasonable suboptimal solutions are often needed since real plant systems are necessary to be controlled. In order to obtain such suboptimal solutions, we have developed a modified version of the hierarchy gradient method by installing fuzzy decision logic. Constraints are classified into non-fuzzy constraints and fuzzy constraints according to their priorities. Fuzzy constraints having a trade-off relationship with each other are compromised reasonably by fuzzy decision logic. The usefulness of the proposed method is numerically and experimentally verified by applying it to the rest-to-rest slew maneuver problem of a flexible space structure, where fuzzy constraints are final time, sensitivity of residual vibration energy with respect to the structure frequency uncertainty and maximum bending moment at the root of the flexible appendage.

抄録全体を表示