New Nonlinear Discrete-Time Models as Applied to Attitude Control of a Miniature Aerial Robot
| Project/Area Number |
17560202
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Dynamics/Control
|
|---|
| Research Institution | University of Tsukuba |
|---|
Principal Investigator |
HORI Noriyuki University of Tsukuba, Graduate School of Systems and Information Engineering, Professor (70312824)
|
|---|
| Project Period (FY) |
2005 – 2007
|
| Project Status |
Completed (Fiscal Year 2007)
|
| Budget Amount *help |
¥3,380,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥180,000)
Fiscal Year 2007: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2006: ¥900,000 (Direct Cost: ¥900,000)
Fiscal Year 2005: ¥1,700,000 (Direct Cost: ¥1,700,000)
|
|---|
| Keywords | Exact Discretization / Delta-Form Discrete-Time Modeling / Nonlinear Digital Control / Air-jet Flow-Rate Control / Miniature Aerial Vehicle / ハイブリッドモデル / 姿勢制御 / 小型無人飛翔ロボット / 浮揚ロボット / 離散時間モデリング |
|---|
| Research Abstract |
1. A number of nonlinear variable transformations have been proposed for linearizing exactly Logistic, Bernoulli, and scalar and matrix Riccati differential equations with constant parameters, which govern dynamics of nonlinear systems. Based on the linearized systems, exact discretization methods have also been presented. Conventional transformations usually convert the nonlinear equations into unstable linear systems, whereas the proposed method can always lead to stable linear systems, which is desirable for control purposes. Based on the insight obtained, matrix Riccati equations with constant parameters have also been linearized and discretized exactly.
2. For nonlinear systems that have not been discretized exactly, parameter identification methods in delta-form have been by proposed by taking a close look at the discrete-time integral gain. This gain is constant for linear systems, but a function of states and inputs for nonlinear systems. By choosing the form of this gain properly, first order systems and van der Pol systems have been used to validate the proposed method under the ideal situation where there is no noise based on simulations. They indicate that the accuracy of the identified model can be much higher than the ordinary method.
3. A 2D and a 3D miniature unmanned aerial vehicle, powered by jetting compressed air, have been designed and build. Altitude, roll, and pitch controls were performed by adjusting the air flow-rate, using servo-valves, to the on-board nozzles. These vehicles were modeled from experimental data and their controllers designed based on partial-model-matching method. These controllers were implemented digitally. With a commonly used Tustin's method, the sampling rate needed was 1 kHz, whereas the experimented controllers were successful at the rate of 50 Hz.
|
Report
(4 results)
Research Products
(26 results)
