• 検索結果がありません。

閉リンクロボットの適応制御における効率的計算アルゴリズムと制御実験

N/A
N/A
Protected

Academic year: 2021

シェア "閉リンクロボットの適応制御における効率的計算アルゴリズムと制御実験"

Copied!
4
0
0

読み込み中.... (全文を見る)

全文

(1)

Title

閉リンクロボットの適応制御における効率的計算アルゴリ

ズムと制御実験( 本文(Fulltext) )

Author(s)

川崎, 晴久; 大岡, 泰仁; 竹村, 信人

Citation

[全国大会講演論文集 : JSME fall annual meeting] vol.[第74期]

p.[449]-[450]

Issue Date

1996

Rights

The Japan Society of Mechanical Engineers (一般社団法人日本

機械学会)

Version

著者最終稿 (author final version) postprint

URL

http://hdl.handle.net/20.500.12099/44357

(2)

閉リンクロボットの適応制御における効率的計算アルゴリズムと制御実験

An Efficient Computational Algorithm of Adaptive Control

for Closed-Loop Robots and Its Experiments

正 川崎 晴久(岐大)

○ 大岡 泰仁(岐大)

竹村 信人(岐大)

Haruhisa KAWASAKI, Gifu University, 1-1, Yanagido, Gifu-si, Gifu

Yasuhito OHKA, Gifu University

Nobuhito TAKEMURA, Gifu University

This paper presents an efficient computational algorithm of model-based adaptive control for closed-loop robots. The algorithm is an extension of the computational algorithm for serial-link robots, which was derived by Kawasaki and Bito. This is implemented to a 6 D.O.F robot with parallel-link mechanism using a 32bit DSP. Experimental results of trajectory control are also shown.

Key Words : closed-loop, adaptive control, computational algorithm, robots 1.はじめに ロボットの高速・高精度化に向け,非線形動特性や負荷 変動に対処すべく,近年モデルベースド適応制御が盛んに 研究されている。しかし,閉リンク機構を含むロボットで は制御則と適応則の計算量は多大となり,多自由度ロボッ トへの適用は困難である。このため,より効率的な計算ア ルゴリズムが求められている。 本報告では,シリアルリンクロボットを対象にした適応 制御計算アルゴリズム(1)を拡張し,閉リンクロボットの効 率的な適応制御計算アルゴリズムを示す。さらに平行リン ク機構を有する 6 自由度ロボットを対象に,1)PD 制御, 2)PID 制御,3)閉リンクを考慮しない適応制御,4)閉リンク を考慮した適応制御の実験比較結果を報告する。 2.適応制御計算アルゴリズム 閉リンクロボットの駆動関節トルクは,閉リンク機構を 適当な関節で仮想的に切断し,木構造開リンク機構を考え (Fig.1),拘束条件を考慮して木構造の関節トルクから計算 する。no自由度の木構造開リンク機構の運動方程式は,関節 変位を qoとすると,

( )

(

)

( )

M q qo o &&o+C q q qo o,& &o o+g qo oo (1) と表される。上式の添字 o は木構造開リンク機構を示す。こ こで,動力学パラメータの線形結合で構成される未知な mo ×1 の定数パラメータベクトルをσoとすると,次の関係式で 表される行列 Yo m n R o o ∈ × を定義できる。

( )

(

)

( ) (

)

M q qo o&&or+C q q qo o,& &o or+g qo o =Y q q q q0& , & , & ,&&o o or orσ τo= or (2)

ここで qorは目標軌道である。いま,肢にベースリンクか ら手先へ順に番号をつけ,肢 i の要素であるリンクには根元 から順にリンク 1i,2i,…と番号をつける。このとき木構造に Branch 2 Link 22 Link 14 Joint 15 Joint 14 Link 13 Joint 22 Link 12 Joint 21 Joint 11 Link 11 Base = Link 01 Branch 1 Joint 12 Link 21 Joint 13 Link 23 Joint 23 Link 15 Branch 3 Branch 4 Branch 5

Fig.1 Virtual tree-structure mechanism ついて以下の関係が成り立つ。 1) 任意の肢s のリンク jsのパラメータは,その肢の先端側 (js<is)の関節トルク

τ

iに影響を与えない 2) 肢の先端が連結リンクの場合,その肢より先の肢の影 響を受けるが,結合に関係のない肢の影響は受けない これらの関係を考慮した場合,d 本の肢をもつ木構造におけ るパラメータ係数行列 Yoは次のように定義できる。 Y y y y y y y y y y y y y y y y y y y y y o n n n n n n n n n n n n n n n n n n d d d d d d d d d d d = 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 2 1 1 1 1 1 2 1 2 1 1 2 2 2 2 2 2 2 2 2 2 L L L L O M M O M L M O M L L L L L L O M L M O M L L L O L L L n n n d d d 0 y O M ⎡ ⎣ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎤ ⎦ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ (3) だだし,yiky A B A 0 ik j k i T k k i T k k i T k k i i k i i k i , , , = + ⎧ ⎨ ⎪ ⎩ ⎪ μ μ γ

is connected with and joint is translational is connected with and joint is rotational isn' t connected with

(4) で与えられる。ここで,i i −1 R をΣi−1からΣiへの回転行列, i i −1 p を Σi−1で表したΣi−1原点からΣi原点へのベクトル,ωi をΣiの角速度とし, ρi i i =⎧⎨ ⎩ 0 1 if joint is translational if joint is rotational

[ ]

Φii+ αi×

{ }

[

]

[

{ }

]

(

[

{ }

]

[

{ }

]

)

Γi i or i o i or i o T

= ω q& × ω q& × + ω q& × ω q& × Ω Γ Γ Γ Γ Γ Γ Γ Γ Γ Γ Γ Γ Γ Γ Γ Γ Γ Γ i i i i i i i i i i i i i i i i i i i i i i i i i i i i = − − + − − + − − − − − + ⎡ ⎣ ⎢ ⎢ ⎢ ⎤ ⎦ α α α α α α α α α , , , , , , , , , , , , , , , , , , , , , , , , , , , 1 23 23 2 13 3 12 22 33 13 2 13 1 23 33 11 3 12 12 12 3 11 22 1 23 2 13 ⎥ ⎥ ⎥

著者最終稿 日本機械学会第74回全国大会講演論文集(1)pp.449~450, 1996 掲載

(3)

{ }

(

{ }

)

α α ρ ω ω i i i i i i i i or i i i i i o i or i i or i q q q = + × + × + − − − − − − R R q z R q z z 1 1 1 1 1 1 2 & &

& & , && ,

(

)

(

)

(

{ }

{ }

)

β β ρ ω ω i i i i i i i i i i i or i o i i i i o i or i i or i q q q = + + − × + × + − − − − − − − − R p R q z R q z z 1 1 1 1 1 1 1 1 2 1 Φ & &

& & &&

, , ,

(

)

hi p R h p k k j j i k k k i k k k = = + = − − −

1 1 1 を定義すると k i k i i k k k i μ =1 = −μ 2 1 1 R z R , kγi=kμi×hik+1

[

]

Ai = βi Φi 0 ,Bi=

[

0

[ ]

βi× Ω i

]

となる。これらの式は再帰的に計算(計算式は省略)でき,関 節トルクτorと推定値σ$oの第 i 要素は τor i ikσo k k i nd , = , =

y , σ$&o i, i

(

&o j, &or j,

)

ji T j i P q q = − − =

y 11 (5) で求められる。上式で求めたτorから,閉リンクロボットの 駆動トルクτcは拘束条件を考慮して

(

)

τc τ T T or = S −1J ここで,S q q = ∈ ⎛ ⎝ ⎜ ⎞ ⎠ ⎟ × ∂ ∂c Rnc nc (6) と計算できる。 Fig.2 に示す平行リンク機構を有する 6 自 由度ロボットにおける Walker のアルゴリズム(2)と上記のア ルゴリズムの計算量を Table.1 に示す。表中,ODP はオリ ジナルの,MDP は最小数のダイナミクパラメータを意味す る。 3.実験 Fig.2 の ロ ボ ッ ト を 対 象 に , 32bit 浮 動 小 数 点 演 算 DSP(TMS320C30)の制御装置を用い,軌道制御の実験をし た。実験では,1)PD 制御,2)PID 制御,3)シリアルリンクロ ボットと見なした従来の計算アルゴリズムによる適応制御, 4)閉リンク機構の適応制御の 4 種類の制御法で軌道誤差の比 較を行った。適応制御の未知パラメータとして,3)はシリア ルリンクロボットの,4)は閉リンクロボットの最小動力学パ ラメータ(3)を使用した。実験条件として,サンプリング時間 は 1.85[ms]とし,各ゲインはそれぞれ試行錯誤により良好な 応 答 と な る よ う に 設 定 し た 。 ま た , 初 期 値 ・ 終 端 値 を qSTART =[ ,0 60 30 0 90 0, , ,− , ]T[deg] , qEND=[30 90 0 60 85 30, , , ,− , ]T [deg]とした。 Fig.3(a),(b)は手先に負荷をつけない場合,Fig.3(c)は負荷 300[g]を取り付けた場合における全関節の軌道誤差ノルムを 表している。図中の AD は適応制御を表し,()内は制御の試 行回数を表す。Fig.3(a)から閉リンク機構の適応制御 1 回目の 最大誤差は,PD 制御に比べて 56%となっているが,シリア ルリンク機構適応制御 1 回目とはほとんど変わらない。しか し Fig.3(b)から制御 10 回目ではシリアルリンク機構適応制御 に比べて 76%となった。さらに Fig.3(c)から手先に負荷を取 り付けた場合,閉リンク機構適応制御 1 回目の最大誤差は, PD 制御の 60%, PID 制御の 62%,シリアルリン branch 2 branch 3 branch 1

Table.1 Number of computations

Parameter Multi. Addi.

Walker ODP 2539 1959 New ODP 2073 1374 Algorithm MDP 1494 1013 0 0.05 0.1 0.15 0.2 0 0.2 0.4 0.6 0.8 Time[s]

Error Norm [rad]

PD

PID Serial(1)

Close(1)

(a) PD, PID, Serial-AD(1), and Close-AD(1) without load.

0 0.05 0.1 0.15 0.2 0 0.2 0.4 0.6 0.8 Time[s]

Error Norm [rad]

Serial(1) Close(1)

Close(10)

Serial(10)

(b) Serial-AD(1,10) and Close-AD(1,10) without load.

0 0.05 0.1 0.15 0.2 0 0.2 0.4 0.6 0.8 Time[s]

Error Norm [rad]

Serial(10)

Close(10) PD

PID

(c) PD, PID, Serial-AD(1), and Close-AD(1) with load. Fig.3 Error norms

ク機構適応制御の 68%となった。 4.結言

本報告では閉リンク機構を含むロボットにおける効率的 Fig.2 6 D.O.F robot with

(4)

な適応制御計算アルゴリズムを提案した。また,6 自由度閉 リンクロボットによる実験を行い,その有効性を示した。

参考文献

(1) H.Kawasaki, et al. , "An Efficient Algorithm for the Model-Based Adaptive Control of Robotic Manipulators", IEEE Trans. on R&A, Vol.12, No.3, 1996 to be appeared. (2) M.W.Walker, "An Efficient Algorithm for the Adaptive

Control of a Manipulator", 5th IEEE Int. Conf. on R&A, 1988. (3) 川崎晴久, 神崎一男, ”マニピュレータモデルにおける最 小動力学パラメータと逆動力学計算法”, 日本ロボット学 会誌, 第 11 巻, 第 1 号, pp.100-110, 1993.

参照

関連したドキュメント

Furthermore, computing the energy efficiency of all servers by the proposed algorithm and Hadoop MapReduce scheduling according to the objective function in our model, we will get

The (GA) performed just the random search ((GA)’s initial population giving the best solution), but even in this case it generated satisfactory results (the gap between the

13 proposed a hybrid multiobjective evolutionary algorithm HMOEA that incorporates various heuristics for local exploitation in the evolutionary search and the concept of

5 used an improved version of particle swarm optimization algorithm in order to solve the economic emissions load dispatch problem for a test system of 6 power generators, for

It is suggested by our method that most of the quadratic algebras for all St¨ ackel equivalence classes of 3D second order quantum superintegrable systems on conformally flat

At the same time, a new multiplicative noise removal algorithm based on fourth-order PDE model is proposed for the restoration of noisy image.. To apply the proposed model for

In order to demonstrate that the CAB algorithm provides a better performance, it has been compared to other optimization approaches such as metaheuristic algorithms Section 4.2

The main idea of computing approximate, rational Krylov subspaces without inversion is to start with a large Krylov subspace and then apply special similarity transformations to H