(A Study on Evaluations of the Behavior and Operational Skills

for an Excavator Based on the Control Engineering Approach)

D160613

2018 3

1 1

1.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6

2 8

2.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8

2.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

2.2.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

2.2.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.2.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

2.2.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2.2.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2.3 . . . . . . . . . . . . . . . . . . . . . . . . . . 19

2.3.1 . . . . . . . . . . . . . . . . . . . . . . . . . 20

2.3.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

2.3.3 . . . . . . . 22

2.3.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28

2.4.1 PQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

2.4.2 PID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

2.4.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

2.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

2.5.1 PQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

2.5.2 PID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36

2.5.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

2.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

344

3.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

i

3.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

3.2.1 . . . . . . . . . . . . . . . . . . . . 46

3.2.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

3.3.1 . . . . . . . . . . . . . . . . . . . . . . 52

3.3.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

3.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

3.4.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56

3.4.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

3.4.3 L + T . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

3.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

463

4.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

4.2 CMAC-PID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

4.2.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

4.2.2 CMAC-PID . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

4.2.3 CMAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

4.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

4.3.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

4.3.2 CMAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

4.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

4.4.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

4.4.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

4.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

5 90

94

98

100

102

ii

1

1.1

2017 55

3 29 1 [1]

Fig.1.1

55 10

Fig. 1.1: Number of workers for constriction fields [1]

1

[2] [3] 2020

i-Construction

[4]

3D

ICT(Information and Communication Technolory

[5, 6]

i-Construction

3K

Society5.0

[7] ,

[8, 9]

2

3 ICT [10, 11]

ICT

[12]

[13]-[16]

[17]

[18] [19]

3

AI

(i)

(ii)

(iii)

1.2

James Watt

[20]-[22]

[23]

Fig.1.2

4

Fig. 1.2: Block diagram of block diagram

Fig. 1.3: Basic idea of study

[24]-[27]

Fig.1.3

Fig.1.4

5

Fig. 1.4: Hydraulic excavator

1.3

2

3

6

4

PID

5

7

2

2.1

[8, 9]

[28] PQ

PQ

8

PID

[29]

PID

[30] [31].

(1)

(2)

(3)

(4)

(5)

9

[32]

PID

(GMVC:Generalized Minimum Variance Control)

[33] [34]

PQ PID

10

Fig. 2.1: Typical hydraulic circuit for excavator

2.2

T (t)

Qin(t)

Fig.2.1

11

Fig. 2.2: Schematic of a hydraulic system for a swing motion

2.2.1

(1)

0

Fig.2.2

TI(t) ω(t)

TI(t) = IΔω(t)

Ts(2.1)

12

Δ Δ = 1− z−1 TI(t) I

Ts ω(t)

I TI(t) Δω(t)

TI(t)

TI(t) T (t)

2.2.2

T (t) Q(t)

P(t) N η

T (t) = ηP(t)Q(t)

2πN(2.2)

T (t) = CPQP(t)Q(t) (2.3)

CPQ Q(t) Qin(t)

AQ(z−1)Q(t) = z−(k+1)BQ(z−1)Qin(t) + ξ(t) (2.4)

13

⎧⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎩AQ(z−1) := 1 + a1Qz−1

BQ(z−1) := b0Q + b1Qz−1 + · · · + bmQz−m

(2.5)

k z−1 m BQ(z) ξ(t) 0

σ2 aiQ biQ

2.2.3

Pr

Pr

P(t) = Pr (2.6)

(2.6) (2.3)

T (t) = CPQ · Pr · Q(t) (2.7)

(2.4) (2.7)

T (t) = CPQ · Prz−(k+1)BQ(z−1)Qin(t) + ξ(t)

AQ(z−1)(2.8)

14

T (t) Qin(t)

AT (z−1)T (t) = z−(k+1)BT (z−1)Qin(t) + ξ(t) (2.9)

⎧⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎩AT (z−1) := 1 + a1T z−1

BT (z−1) := b0T + b1T z−1 + · · · + bmT z−m

(2.10)

aiT biT

2.2.4

Q(t) P(t) Q(t)

P(t)

Q(t) P(t)

ω(t) Qm(t)

Qm(t) =qm

2πω(t) (2.11)

qm P(t)

15

TI(t)

TI(t) =qm

2πP(t) (2.12)

Q(t)

Q(t) Qm(t)

P(t) =Ts

Δκ(Q(t) − Qm(t)) (2.13)

κ P(t)

Q(t) (2.1) (2.11) (2.12)

Qm(t) =1

Δ

(qm

2π

)2 Ts

IP(t) (2.14)

(2.14) (2.13)

P(t) =1

ΔTsκQ(t) − 1

Δ2κ(qm

2π

)2 T 2s

IP(t)) (2.15)

P(t) Q(t)

(K1 + Δ2)P(t) = K2ΔQ(t) (2.16)

K1 = κ(qm

2π

)2 Ts2

I(2.17)

K2 = κTs (2.18)

16

Q(t) P(t)

P(t) =K2

K1 + Δ2ΔQ(t) (2.19)

2.2.5

P(t) Pr

(2.19) P(t) Q(t)

(2.19) (2.4)

P(t) =K2

K1 + Δ2Δ

(z−(k+1)BQ(z−1)

AQ(z−1)Qin(t) +

ξ(t)AQ(z−1)

)(2.20)

(2.3) (2.4)

T (t) = CPQP(t)(z−(k+1)BQ(z−1)

AQ(z−1)Qin(t) +

ξ(t)AQ(z−1)

)(2.21)

(2.21) (2.20)

AT (z−1)T (t) = Δ(z−(k+1)BT (z−1)Qin(t) + ξ(t)

)2(2.22)

AT (z−1) = A2Q(z−1)(K1 + Δ

2) (2.23)

BT (z−1) = (CPQK2)12 BQ(z−1) (2.24)

Qin(t) Qin ΔQin(t)

17

AT (z−1)T (t) = Δz−(k+1)BT (z−1)ΔQin(t) + ξ(t) (2.25)

⎧⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎩AT (z−1) := 1 + a′1T z−1 + · · · + a′nT z−n

BT (z−1) := b′0T + b′1T z−1 + · · · + b′mT z−m

(2.26)

n AT (z) a′iT b′iT

Qmax

Qmax

T (t)

ΔQ(t) = ΔQin(t) = 0 (2.27)

P(t)

0 (2.6) (2.19)

18

Fig. 2.3: Block diagram of an event-driven controller

2.3

[30]

(2.9) (2.25)

[32]

Fig.2.3

PID

19

x(t)

P(t) u(t) Qin(t)

y(t) T (t)

2.3.1

(2.9)

PID

Δu(t) = KI(t)e(t) − KP(t)Δy(t) − KD(t)Δ2y(t) (2.28)

e(t) := r(t) − y(t) (2.29)

e(t) r(t)

(2.25)

PID

20

PID

Δ2u(t) = KI(t)Δe(t) + KII(t)e(t) − KP(t)Δ2y(t) − KD(t)Δ3y(t) (2.30)

KP(t),KI(t),KII(t),KD(t)

(2.27)

Δu(t) = 0 (2.31)

(2.30)

u(t)

(2.31)

21

2.3.2

ET

P(t)

ET1

ET2 P(t) 100% P1

ET3 P(t) 100% P2

ET4 P(t) 100% P3

P1 P2

(2.25) P3

P1

P(t) 100%

2.3.3

ET

PID ET P(t)

22

KP(t) =

⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩

KP1(if ET1 ∼ ET2)

KP2− KP1

P2 − P1

(P(t) − P1) + KP1(if ET2 ∼ ET3)

KP2(if ET3 ∼ ET4)

0 (otherwise)

(2.32)

KI(t) =

⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩

KI1(if ET1 ∼ ET2)

KI2− KI1

P2 − P1

(P(t) − P1) + KI1(if ET2 ∼ ET3)

KI2(if ET3 ∼ ET4)

0 (otherwise)

(2.33)

KII(t) =

⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩

KII2

P2 − P1

(P(t) − P1) (if ET2 ∼ ET3)

KII2(if ET3 ∼ ET4)

0 (otherwise)

(2.34)

KD(t) =

⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩

KD1(if ET1 ∼ ET2)

KD2− KD1

P2 − P1

(P(t) − P1) + KD1(if ET2 ∼ ET3)

KD2(if ET3 ∼ ET4)

0 (otherwise)

(2.35)

KP1,KI1,KD1

KP2,KI2,KII2,KD2

23

2.3.4

KP1KI1

KD1KP2

KI2KII2

KD2

KP1KI1

KD1KP2

KI2KII2

KD2

GMVC

PID [34]

GMVC PID

GMVC PID

A(z−1)y(t) = Δid z−(k+1)B(z−1)u(t) +ξ(t)Δ

(2.36)

A(z−1) := 1 + a1z−1 + a2z−2

B(z−1) := b0 + b1z−1 + · · · + bmz−m

⎫⎪⎪⎪⎪⎪⎪⎬⎪⎪⎪⎪⎪⎪⎭(2.37)

(2.37) ai bi id 0 1

id = 0 id = 1

J

J := E[φ2(t + k + 1)] (2.38)

24

φ2(t + k + 1)

φ(t + k + 1) := P(z−1)y(t + k + 1) + λΔ(id+1)u(t) − P(1)r(t) (2.39)

λ

Diophantine

P(z−1) = Δ(id+1)E(z−1)A(z−1) + z−(k+1)F(z−1) (2.40)

⎧⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎩E(z−1) := 1 + e1z−1 + · · · + ekz−k

F(z−1) := f0 + f1z−1 + f2z2 + f3z3 · id

(2.41)

E(z−1) F(z−1) Δ(id+1)A(z−1) P(z−1)

P(z−1)

[36]

Pl(z−1) = 1 + p1z−1 + p2z−2 (2.42)

p1 = −2 exp(− ρ

2μ

)cos

( √4μ−1

2μρ

)

p2 = exp(ρ

μ

)

ρ = Tsσ

μ = 0.25(1 − δ) + 0.51δ

⎫⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎬⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎭

(2.43)

σ δ

σ 60

δ 0 ≤ δ ≤ 2.0

25

δ = 0 Binomial δ = 1.0 Butterworth

(2.36) (2.39) (2.40) t k + 1

φ(t + k + 1|t) := F(z−1)y(t) +(E(z−1)B(z−1) + λ

)Δ(id+1)u(t) − P(1)r(t) + E(z−1)ξ(t + k + 1) (2.44)

t

φ̂(t + k + 1|t) := F(z−1)y(t) +(G(z−1) + λ

)Δ(id+1)u(t) − P(1)r(t) (2.45)

G(z−1)

G(z−1) := E(z−1)B(z−1) (2.46)

= g0 + g1z−1 + + · · · + gngz−ng (2.47)

(ng = m + k)

ng G(z−1) (2.44) (2.45)

φ(t + k + 1|t) = φ̂(t + k + 1|t) + E(z−1)ξ(t + k + 1) (2.48)

φ̂(t + k + 1|t) = 0 (2.49)

(2.38)

26

(2.45) (2.49) GMVC

Δ(id+1)u(t) =Pl(1)

G(z−1) + λr(t) − F(z−1)

G(z−1) + λy(t) (2.50)

u(t) y(t) (2.50)

fi gi

(2.50) id = 0 (2.28)

KP1= − f1 + 2 f2

G(1) + λ(2.51)

KI1=

f0 + f1 + f2

G(1) + λ(2.52)

KD1=

f2

G(1) + λ(2.53)

(2.50) id = 1 (2.30)

KP2=

f2 + 3 f3

G(1) + λ(2.54)

KI2= − f1 + 2 f2 + 3 f3

G(1) + λ(2.55)

KII2=

f0 + f1 + f2 + f3

G(1) + λ(2.56)

KD2= − f3

G(1) + λ(2.57)

27

2.4

2.4.1 PQ

100%

PQ Fig.2.4 Fig.2.5

2.4.2 PID

PID Fig.2.6

PID

KP = 3.00 × 10−6, KI = 1.10 × 10−7, KD = 1.00 × 10−7 (2.58)

PID

28

0 2 4 6 8 100

100

200

T(t)

[%] T(t)

r(t)

0 2 4 6 8 100

50

100

Q(t)

[%]

0 2 4 6 8 100

50

100

P(t)

[%]

Fig. 2.4: Control result of PQ controller with ordinary

29

0 2 4 6 8 100

100

200

T(t)

[%] T(t)

r(t)

0 2 4 6 8 100

50

100

Q(t)

[%]

0 2 4 6 8 100

50

100

P(t)

[%]

Fig. 2.5: Control result of PQ controller with cold condition

30

0 2 4 6 8 100

50

100

150

T(t)

[%] T(t)

r(t)

0 2 4 6 8 100

50

100

Q(t)

[%]

0 2 4 6 8 100

50

100

P(t)

[%]

Fig. 2.6: Control result of PID controller with cold condition

31

Table 2.1: PID gains for event-driven controller

Gain RV ON RV OFF

KP 1.97 × 10−6 1.95 × 10−6

KI 3.68 × 10−7 1.10 × 10−7

KII 0 3.89 × 10−10

KD 1.61 × 10−6 2.93 × 10−7

2.4.3

PID GMVC PID

(2.6) (2.9)

–(2.10) PID

2.2[s] 3.8[s]

0.1[s]

PID

Table 2.1

(2.19) ΔQ(t)

P(t) (2.25) –(2.26) 3.9[s] 4.3[s]

1[s]

PID Table 2.1

λ = 0.01

Pr 100% Fig.2.6 P1 P2 P3 97% 70% 40%

Fig.2.7 ET

32

Fig.2.8

PID

PID

33

0 2 4 6 8 100

50

100

150

T(t)

[%] T(t)

r(t)

0 2 4 6 8 100

50

100

Q(t)

[%]

0 2 4 6 8 100

50

100

150

P(t)

[%] P1

P3

P2

ET1 ET2

ET4ET3

Fig. 2.7: Control result of event-driven controller with cold condition

34

0 2 4 6 8 101.9

1.95

2

0 2 4 6 8 100

0.2

0.4

0 2 4 6 8 100

2

410-4

0 2 4 6 8 100

1

2

Fig. 2.8: Control parameter trajectories corresponding to Fig.2.7

35

2.5

2.5.1 PQ

Fig.1.4 20t SK200-9

PQ Fig.2.9

1

2.5.2 PID

PID Fig.2.10

PID

KP = 5.00 × 10−2, KI = 2.50 × 10−3, KD = 6.00 × 10−3 (2.59)

PID

36

0 1 2 3 4 5 6 7 8 9t [s]

0

100

200

T(t)

[%]

T(t)r(t)

0 1 2 3 4 5 6 7 8 9t [s]

0

50

100

Q(t)

[%]

0 1 2 3 4 5 6 7 8 9t [s]

0

50

100

150

P(t)

[%]

Fig. 2.9: Control result using a PQ controller

37

PID

Fig.2.6 PQ

2.5.3

PID

(2.6) (2.26)

P(t)

(2.6)

Pr

PID

0.8[s] 1.2[s]

0.1[s]

λ = 1

(2.19) ΔQ(t)

P(t) 3.9[s]

6.5[s]

1[s] λ = 0.01

PID

38

0 1 2 3 4 5 6 7 8 9t [s]

0

50

100

150

T(t)

[%]

T(t)r(t)

0 1 2 3 4 5 6 7 8 9t [s]

0

50

100

Q(t)

[%]

0 1 2 3 4 5 6 7 8 9t [s]

0

50

100

150

P(t)

[%]

Fig. 2.10: Control result using a PID controller

39

Table 2.2: PID gains included in the event-driven controller

Gain RV ON RV OFF

KP 5.10 × 10−2 0

KI 2.40 × 10−3 1.90 × 10−3

KII 0 2.00 × 10−6

KD 1.60 × 10−2 3.00 × 10−3

Table 2.2 KP2

Pr 100%

Fig.2.10 P1 P2 P3 83% 60% 40%

Fig.2.11 , Fig.2.12 Fig.2.11

ET

Qmax PID

0.5s

40

0 2 4 6 8t[s]

0

50

100

150

T(t)

[%]

T(t)r(t)

0 2 4 6 8t[s]

0

50

100

Q(t)

[%]

0 2 4 6 8t[s]

0

50

100

150

P(t)

[%]

ET1 ET2

ET3 ET4

P2P1

P3

Fig. 2.11: Control result using the event-driven torque controller

41

0 2 4 6 80

0.05

0.1

0 2 4 6 8

t[s]

0

2

4 10-3

0 2 4 6 80

2

4 10-6

0 2 4 6 80

0.01

0.02

Fig. 2.12: Control parameter trajectories corresponding to Fig.2.11

42

2.6

GMVC

PQ

PID

43

3

3.1

44

[37, 38, 39]

[40, 41]

[42]

[43]

(1)

(2)

(3)

(4)

[44, 45]

3 4 [46, 47]

45

+

+

[48]

+

3.2

3.2.1

(i)

Fig.1.4 20t SK200-9

360

46

3

3

(ii)

[51] PID

PID Ziegler and Nichols

47

[52] Chien, Hrones and Reswick [53]

T L K

G(s) =K

1 + T se−Ls (3.1)

L T

L/T 1

[54, 55]

Fig.3.1

[46]

[47]

48

Fig. 3.1: Scheme of operability evaluation

3.2.2

Fig. 3.2

Fig. 3.3

Fig. 3.4

(3.1) +

Fig. 3.5

Fig. 3.3

49

Fig. 3.2: View of Bucket Motion from Cabin

Fig. 3.3: Block diagram of evaluation for motion of hydraulic excavator

50

Fig. 3.4: Simulator of bucket motion

Fig. 3.5: Block diagram of evaluation for simulation

51

Fig. 3.6: Flow chart of GA

3.3

3.3.1

GA

GA (2) (6)

GA Fig.3.6

(1)

(2)

(3)

(4)

(5)

(6)

GA

••

52

Fig. 3.7: Block diagram of system parameter estimation by GA

Fig. 3.7

(1)

N

T K L

(2)

1 0

J :=1

1 +

n∑t=0

ω(t) {ym(t) − y(t)}2(3.2)

ym(t) y(t)

GA n

53

0 0.5 1 1.5 2 2.5 3 3.5t [s]

0

20

40

60

80

100

120y m

, y [%

]

ymy

63.2%

10%

Fig. 3.8: Weight coefficients of criteria for GA

ω(t)

ω(t) =

⎧⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩

1 (ym(t) > 63.2%)

3 (63.2 ≥ ym(t) ≥ 10%)

5 (ym(t) < 10%)

(3.3)

GA

Fig. 3.8 ym(t)

63.2%

Fig. 3.8 100%

(3)

α%

54

(4)

(3) 2 β%

[49, 50]

Pc = Jmax(Pm, Pn) + |Pm−Pn |4

Pd = Jmax(Pm, Pn) − |Pm−Pn |4

⎫⎪⎪⎪⎪⎪⎪⎬⎪⎪⎪⎪⎪⎪⎭(3.4)

Pc Pd Pm Pn

Jmax(Pm, Pn) Pm, Pn

(5)

γ%

(6)

(2) (5) G

K,T, L

3.3.2

(i) L T

(ii) L T 10

10

55

L–T L + T

(iii) 3 1 5

(iv) GA

3.4

3.4.1

7 Table 3.1

7 T L

Fig. 3.9

3

L T L T

0.4 0.7

56

Table 3.1: Evaluation results of the simulation

No. L T L + T responsiveness

Criterion 0.2 0.25 0.45 3.0

1 0.2 2.0 2.2 2.0

2 0.2 0.1 0.3 4.1

3 0.4 0.5 0.9 2.6

4 0.1 0.5 0.6 3.1

5 0.4 2.0 2.4 1.4

6 0.4 0.2 0.6 3.1

7 0.1 0.125 0.225 4.6

8 0.3 0.15 0.45 3.4

9 0.1 1.0 1.1 2.1

10 0.3 1.5 1.8 1.8

0 0.5 1 1.5 2 2.5 3T [s]

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

L [s

]

2.0

2.6 1.4

4.6

3.4

2.1

1.8

3.0 4.1

3.1

3.1

Evaluation pointCriterionconstant L+T

Fig. 3.9: Simulation result of the responsiveness evaluation on the L-T map

57

3.4.2

5 GA

Table 3.2

5 T L

Fig. 3.10

3

3.4.3 L + T

L T

L T L + T

Fig. 3.11

L + T

S R

S = a log(R) + b (3.5)

S log(R) r Table 3.3

R S

L + T L

T

S

58

Table 3.2: Evaluation results of the excavator operation

No. Estimated Parameters responsiveness

L T L + T

1 0.13 0.14 0.27 4.6

2 0.03 2.60 2.63 1.6

3 0.17 0.58 0.75 2.4

4 0.20 0.27 0.47 3.8

5 0.21 0.18 0.39 4.0

6 0.69 0.27 0.96 1.4

7 0.23 0.32 0.55 3.2

8 0.21 0.27 0.48 3.4

9 0.17 0.21 0.38 4.0

10 0.15 0.17 0.32 4.2

0 0.5 1 1.5 2 2.5 3T [s]

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

L [s

]

4.6

1.6

2.4 3.8

4.0

1.4

3.2 3.4

4.2 3.0

Evaluation pointCriterionconstant L+T

Fig. 3.10: Excavator result of the responsiveness evaluation on the L-T map

59

R Fig. 3.11

[56] S R

60

0 0.5 1 1.5 2 2.5 3L+T [s]

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

5.5

Res

pons

ivin

ess

SimulationExcavatorCriterionSimulation WF

a = -2.72b = 2.51

Fig. 3.11: Result of the responsiveness evaluation on the L+T axis

Table 3.3: Correlation coefficients and P-values of evaluations

Simulation Excavator

r P[%] r P[%]

L 0.39 24.0 0.006 98.5

T -0.93 2.9 × 10−3 -0.72 0.80

L + T -0.97 9.1 × 10−5 -0.88 0.02

61

3.5

GA

62

4

4.1

i-construction [58]

ICT 3

[59] [10, 11]

[60]

63

[61, 62]

PID

CMAC [64, 65, 66]

[67, 68]

CMAC CMAC PID

PID

CMAC

PID

64

4.2 CMAC-PID

CMAC-PID

CMAC-PID

CMAC

CMAC-PID

4.2.1

[63]

Albus [64, 65, 66]

Cerebellar Model Articulation Controller

CMAC

(i) CMAC

CMAC Fig.4.1 CMAC

S →M→A→ P (4.1)

65

�

Fig. 4.1: Structure of cerebellum and CMAC

S M A P

S M

M A P

(4.2)

P = H(S) (4.2)

CMAC

(ii)

[Step1 ] CMAC P̂

[Step2 ] S1 CMAC P1 CMAC

[Step3 ] S1 P̂1

66

[Step4 ]

Wnewh ← Wold

h + gP̂1 − P1

K(4.3)

Woldh Wnew

h K S1

g

g = 1

[Step ] Step2 Step4

(iii) 2 CMAC

Fig.4.2 2 CMAC 2 CMAC CMAC

CMAC 3 4

S (7, 4) M {C,H,K} {b, g, j}

4 3 1 8

4 CMAC 4

8 3 −4/3

(iv) CMAC

CMAC

67

Fig. 4.2: Example of 2 dimensional CMAC model

(1)

(2)

(3)

(4)

CMAC

(1)

BP

BP

CMAC

68

(2)

CMAC

CMAC

CMAC

CMAC

Fig.4.2 S (7, 4) {C,H,K} {b, g, j}

(7, 4) (6, 4) {C,G,K} {b, g, j}

H G

(3)

CMAC

Fig.4.2 S (7, 4) {C,H,K} {b, g, j}

S (7, 4) (6, 4) {C,G,K} {b, g, j}

H G

S (4, 7)

69

Fig. 4.3: Block diagram of skill based CMAC-PID controller

(4)

CMAC

1

1 1

Fig.4.2 1 1

100

CMAC

48

4.2.2 CMAC-PID

CMAC-PID Fig.4.3

u∗(t)

y∗(t) PID CMAC

70

PID

Δu(t) = KI(t)e(t) − KP(t)Δy∗(t) − KD(t)Δ2y∗(t) (4.4)

e(t) := r(t) − y∗(t) (4.5)

KP(t) KI(t) KD(t) PID e∗(t) Δ

Δ = 1 − z−1

PID CMAC WPID,h(t)

KP(t) =K∑

h=1

WP,h(t)

KI(t) =K∑

h=1

WI,h(t) (4.6)

KD(t) =K∑

h=1

WD,h(t)

K CMAC CMAC 3

CMAC e(t), y∗(t) Δy∗(t) WP,h(t)

WI,h(t) WI,h(t)

4.2.3 CMAC

CMAC Fig.4.4

y∗(t) u(t)

71

Fig. 4.4: Human skill leaning by CMAC-PID

u∗(t) u(t) u∗(t)

WnewP,h (t) = Wold

P,h (t) − g(t) ∂J(t)∂KP(t)

1

K

WnewI,h (t) = Wold

I,h (t) − g(t) ∂J(t)∂KI(t)

1

K(4.7)

WnewD,h (t) = Wold

D,h(t) − g(t) ∂J(t)∂KD(t)

1

K

g(t) J(t)

g(t) =1

c + a · exp(−b|u∗(t) − u(t)|) (4.8)

J(t) :=1

2ε(t)2 (4.9)

ε(t) = u∗(t) − u(t) (4.10)

(4.8) a b c

a u∗(t) u(t) c u∗(t)

72

u(t) b u∗(t) u(t)

a (4.7)

∂J(t)∂KP(t)

=∂J(t)∂ε(t)

∂ε(t)∂u(t)

∂u(t)∂KP(t)

= ε(t)Δy∗(t)

∂J(t)∂KI(t)

=∂J(t)∂ε(t)

∂ε(t)∂u(t)

∂u(t)∂KI(t)

= −ε(t)e(t) (4.11)

∂J(t)∂KD(t)

=∂J(t)∂ε(t)

∂ε(t)∂u(t)

∂u(t)∂KD(t)

= ε(t)Δ2y∗(t)

73

Fig. 4.5: Experimental method

4.3

CAMC-PID

4.3.1

Fig.4.5

SK200-9

90◦ 10

0◦ 90◦

y∗(t) u∗(t)

Ts 200ms

74

0 20 40 60 80 100 120 140-1

0

1

u* (t)

0 20 40 60 80 100 120 140t[s]

0

0.5

1

y* (t)

y*

r

Fig. 4.6: Data of novice operation

0 20 40 60 80 100 120 140-1

-0.5

0

0.5

1

u* (t)

0 20 40 60 80 100 120 140t[s]

0

0.5

1

y* (t)

y*

r

Fig. 4.7: Data of professional operation

75

4.3.2 CMAC

y∗(t) u∗(t)

2 2

1 Fig.4.6 Fig.4.7

y∗(t) 1 u∗(t) 1

y∗(t) u∗(t) (4.4) PID

CMAC

(4.7) -(4.11) CMAC

10 10 CMAC Fig.4.8 Fig.4.9 Fig.4.8

Fig.4.9

76

0 20 40 60 80 100 120 140-1

0

1u(

t)u*(t)u(t)

0 20 40 60 80 100 120 1400

0.5

1

y(t)

y*(t)r(t)

0 20 40 60 80 100 120 1402

2.2

2.4

KP(t)

LearningInitial

0 20 40 60 80 100 120 1400

0.20.40.60.8

KI(t)

0 20 40 60 80 100 120 140t[s]

2.3

2.32

2.34

KD

(t)

Fig. 4.8: Learning result for novice operation

77

0 20 40 60 80 100 120 140-1

0

1u(

t)u*(t)u(t)

0 20 40 60 80 100 120 1400

0.5

1

y(t)

y*(t)r(t)

0 20 40 60 80 100 120 1402

2.2

2.4

KP(t)

LearningInitial

0 20 40 60 80 100 120 1400

0.20.40.60.8

KI(t)

0 20 40 60 80 100 120 140t[s]

4.36

4.38

4.4

KD

(t)

Fig. 4.9: Learning result for professional operation

78

4.4

4.4.1

PID Fig.4.8 Fig.4.9

Fig.4.10 Fig.4.11

71 77 52 58

6

PID

[69] PID

PID PID

u(t) = KPe(t) + KI

∫ t

0

e(τ)dτ + KDde(t)

dt(4.12)

(4.12) P e(t)

I

PID KP

KP

KI

KI

KP KD

KD

KD

PID

79

52 53 54 55 56 57 580

0.5

1

u(t)

u*(t)u(t)

52 53 54 55 56 57 58

0

0.5

1

y(t)

y*(t)r(t)

52 53 54 55 56 57 582

2.2

2.4

KP(t)

LearningInitial

52 53 54 55 56 57 580

0.20.40.60.8

KI(t)

52 53 54 55 56 57 58t[s]

2.3

2.32

2.34

KD

(t)

Fig. 4.10: Extracted data of Novice operation

80

71 72 73 74 75 76 770

0.5

1

u(t)

u*(t)u(t)

71 72 73 74 75 76 77

0

0.5

1

y(t)

y*(t)r(t)

71 72 73 74 75 76 772

2.2

2.4

KP(t)

LearningInitial

71 72 73 74 75 76 770

0.20.40.60.8

KI(t)

71 72 73 74 75 76 77t[s]

4.36

4.38

4.4

KD

(t)

Fig. 4.11: Extracted data of professional operation

81

Table 4.1: Behavior of each action of PID controller

(P) (I) (D)

PID

Table 4.1

Fig.4.11

Fig.4.10

PID

PID

56.5

0

0

0 ON/OFF

82

Fig.4.10

4.4.2

Fig.4.10 Fig.4.11 ΔKP(t) ΔKI(t) ΔKD(t)

Fig.4.12 Fig.4.13

Fig.4.14

Fig.4.15

Table 4.2

83

Table 4.2: Standard deviation of ΔKP, ΔKI & ΔKD

Operator Professional 1 Professional 2 Novice 1 Novice 2

ΔKP 0.017 0.019 0.032 0.041

ΔKI 0.051 0.060 0.063 0.075

ΔKD 0.003 0.006 0.009 0.009

84

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

P

0

10

20

30

40

50

Num

ber P

-0.4 -0.2 0 0.2 0.4

I

0

10

20

30

40

Num

ber I

-0.05 0 0.05

D

0

10

20

30

Num

ber D

Fig. 4.12: Distribution chart of ΔKP, ΔKI & ΔKD for professional 1

85

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

P

0

10

20

30

40

50

Num

ber P

-0.4 -0.2 0 0.2 0.4

I

0

10

20

30

40

Num

ber I

-0.05 0 0.05

D

0

10

20

30

Num

ber D

Fig. 4.13: Distribution chart of ΔKP, ΔKI & ΔKD for novice 1

86

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

P

0

10

20

30

40

50

Num

ber P

-0.4 -0.2 0 0.2 0.4

I

0

10

20

30

40

Num

ber I

-0.05 0 0.05

D

0

10

20

30

Num

ber D

Fig. 4.14: Distribution chart of ΔKP, ΔKI & ΔKD for professional 2

87

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

P

0

10

20

30

40

50

Num

ber P

-0.4 -0.2 0 0.2 0.4

I

0

10

20

30

40

Num

ber I

-0.05 0 0.05

D

0

10

20

30

Num

ber D

Fig. 4.15: Distribution chart of ΔKP, ΔKI & ΔKD for novice 2

88

4.5

CMAC-PID

PID

PID PID

PID

PID

89

5

90

2

PID

3

91

4

CMAC-PID

CMAC-PID CMAC

PID

CMAC

PID

ON/OFF

92

93

[1] 28 3 2 ,

http://www.mlit.go.jp/common/001121700.pdf, (2017 12 13

[2] , (2012)

[3] , , 2009

10 , pp. 78–95 (2009)

[4] i-Construction ICT

http://www.mlit.go.jp/common/001118343.pdf, (2017 12 13

[5] i-Construction : i-Construction (

) , , Vol.68, No.5, pp. 11–15(2016)

[6] i-Construction ( (1)) , , Vol.53,

No.1, pp. 1–7 (2017)

[7] i-Construction ,

http://www.mlit.go.jp/common/001206546.pdf ,(2017 12 13

[8] CMI :

Vol.68, No.7, pp.86–90 (2017)

[9] ( ) ,

, Vol.67, No.1, pp.10–14 (2015)

[10] ICT ,

, Vol.47, No.2, pp.74–77 (2016)

[11] ICT ICT Vol.55, No.6,

pp.523–526 2016

[12] ICT

( ) , Vol.67, No.12,

pp.16–20 2015

[13] ,

Vol.102, No.965, pp.233-234 (1999)

[14] : ,

. , Vol.36, pp.245–250 (2006)

[15] , ( :

) , R&D , Vol.57, No.1, pp.62–65 (2007)

[16] , ,

, , Vol. 48, No. 12, pp.34–39 (2012)

94

[17] Y.Sakaida, D.Chugo, H.Yamamot, H.Asama: ”The Analysis of Excavator Operation by Skill-

ful Operator”, Proc. of SICE Annual Conference, tokyo, pp.538–542 (2008)

[18] , , : ( 1 ,

) C , Vol.67, No.653, pp.201–208 (2001)

[19] , , : ( 2 ,

) C ,

Vol.68, No.666, pp.547–554 (2002)

[20] S. Bennett ”A history of control engineering”, Inst of Engineering & Technology (1986)

[21] , , (1998)

[22] , (1988)

[23] MATLAB/Simulink ,

(2001)

[24] , Vol. 3, No. 4, pp. 266–274 (1967)

[25] , (1987)

[26] , (2005)

[27] , , , , Vol.

47 No. 1, pp.36–40 (2001)

[28] , , (1999)

[29] , , , ,

PID , 14

USB , pp.82–85 (2015)

[30] T.Yamamoto, K.Takao and T.Yamada, ”Design of a Data-Driven PID Controller”, IEEE Trans.

on Control System Technology, Vol.17, No.1, pp.29–39 (2009)

[31] T.Kinoshita, Y.Oshima, K.Koiwai, T.Yamamoto, T.Nanjo, Y.Yamazaki and

Y.Fujimoto ”Design of a Data-Driven Control System for a Hydraulic Excavator”,

Proceedings of The 2016 International Conference on Artificial Life and Robotics, Okinawa,

pp.393–396 (2016)

[32] , , -

-

CVIM 2006(51(2006-CVIM-154)) pp.159–166 (2006)

[33] D. W. Clarke and P. J. Gawthrop: ”Self-Tuning Control”, Proc. IEE, 126D-6,pp.633–640

(1979)

[34] , , PID ,

C, Vol.132, No.6, pp.873–878 (2012)

[35] , , (1996)

[36] , PID

, , Vol.11, No.1, pp.1–9 (1998)

95

[37] , ,

Vol.61, No.7, pp. 97–103 (2007)

[38] Y.K.Chang, J.R. Hwang: ”A new approach to improve the vehicle ride comfort by electroen-

cephalogram data”, International Journal of Vehicle Design, Vol.57, Issue2-3, pp. 106–123

(2011)

[39]

Vol. 32 No. 10 p. 851–854 (2014)

[40] T.Wang, Q.Wang, T.Lin: ”Improvement of boom control performance for hybrid hydraulic

excavator with potential energy recovery”, Automation in Construction , Vol.30, pp. 161–169

(2013)

[41] A.Nishiyama, M.Moteki, K.Fujino, T.Hashimoto: ”Operation analysis of hydraulic excavator

with different operation system”, Proc. 33rd International Symposium on Automation and

Robotics in Construction, Auburn, pp. 487–495 (2016)

[42] H.Stone, R.Bleibaum, H.A.Thomas: ”Sensory Evaluation Practices”, Academic Press (2004)

[43] M.C.Meilgaard, B.T.Carr G.V.Civille: ”Sensory Evaluation Techniques, Fourth Edition”,

CRC Press. (2006)

[44] , , :

Vol. 79 No. 806 pp. 3327–3334 (2013)

[45] , , , 3

Vol. 34 No. 1 p.

41–52 (2010)

[46] : , , Vol.32, No.1 pp.7–14

(2012)

[47] :

, NHK (2013)

[48] D.E.Goldberg: ”Genetic Algorithms in search, optimization, and machine learning”, Addison-

Wesley Publishing Company (1989)

[49] GA PD

D Vol.123, No.10, pp.1097–1103

(2003)

[50] PID

C Vol.131, No.4, pp.794–799 (2011)

[51] PID , (1992)

[52] J.G.Ziegler and N.B.Nichols: ”Optimum Settings for Automatic Controllers”, Trans. of

ASME, Vol. 64 pp. 759–768 (1942)

[53] K.L.Chien, J.A.Hrones and J.B.Reswick: ”On the Automatic Control of Generalized Passive

Systems”, Trans. ASME, Vol. 74 pp. 175–185 (1952)

96

[54]

- C Vol.134, No.10, pp.1537–1542

(2014)

[55] -

C Vol.135, No.1, pp.66–72 (2015)

[56] Fechner, G.T ”Elements of Psychophysics ”, New York: Holt, Rinehart and Winston (1966)

[57] IT RT –

Vol.52, No.1, pp.16–19 2010

[58] ICT

Vol.55, No.6, pp.470–476 2016

[59]

Vol.55, No.6, pp.511–514 2016

[60] :

:

, 29 , Vol.29

, pp. 217-222 (2002)

[61] M.Kato, T.Yamamoto and S.Fujisawa : ”A Skill-Based PID Controller Using Artificial Neu-

ral Networks”, Proc. of Int. Conf. on Computational Intelligence for Modeling, Control and

Automation, pp.702-705 (2005)

[62] T.Yamamoto, S.Mori and A.Sakaguchi : ”Data-Driven Skill-Based PID Control of a Pilot-

Scale Helicopter Model”, Int. J. of Innovative Computing Information and Control, Vol.4,

No.12, pp.3349–3358 (2008)

[63] , (2015)

[64] J.S.Albus : ”A New Approach to Manipulator Control Cerebellar Articulation Control

(CMAC)”, Trans. on ASME, J. of Dynamics Systems, Measurement, and Control, Vol.97,

No.9, 220-227 (1975)

[65] J.S.Albus : ”Data Storage in The Control Cerebellar Articulation Control (CMAC)”, Trans. on

ASME, J. of Dynamics Systems, Measurement, and Control, Vol.97, No.9, 228-233 (1975)

[66] J.S.Albus : ”Brains, Behavior and Robotics”, McGraw-Hill, (1981)

[67] T.Yamamoto, R.Kurozumi and S.Fujisawa : ”A Design of CMAC Based Intelligent PID Con-

trollers”, Artificial Neural Networks and Neural Information Processing, Lecture Notes in

Computer Science, Vol.2714, pp.471-478 (2003)

[68] K.Koiwai, K.Kawada and T.Yamamoto : ”Design and Experimental Evaluation of an Intel-

ligent PID Controller using CMAC”, Proc. of IEEE Int. Conf. on Networking, Sensing and

Control, pp. 740-745 (2009)

[69] Vol.42, No.4,

pp.330-334 2003

97

A.

[1] Kazushige Koiwai, Liao Yuntao, Toru Yamamoto, Takao Nanjo, Yoichiro Yamazaki, Yoshi-

aki Fujimoto Feature Extraction for Excavator Operation skill using CMAC Journal ofRobotics and Mechatronics Vol.28 No.5 pp. 715-721 2016

[2] Yasuhito Oshima, Takuya Kinoshita, Kazushige Koiwai, Toru Yamamoto, Takao Nanjo, Yoichiro

Yamazaki, Yoshiaki Fujimoto Data-Driven Torque Controller for a Hydraulic Excavator

Journal of Robotics and Mechatronics Vol.28 No.5 pp. 752-758 2016

[3]

Vol.54 No.2, pp. 261-268 2018

[4]

Vol.138 No.5 (2018 5

B.[1] Kazushige Koiwai, Yuntao Liao, Toru Yamamoto, Takao Nanjo, Yoichiro Yamazaki, Yoshiaki

Fujimoto : A Consideration on Feature Extraction for Operation Skill Based on Control

Engineering Approach , Proceedings of the 2016 International Conference on Artificial Lifeand Robotics, pp.389-392, Okinawa, January 2016

[2] Takuya Kinoshita, Yasuhito Oshima, Kazushige Koiwai, Toru Yamamoto, Takao Nanjo, Yoichiro

Yamazaki, Yoshiaki Fujimoto : Design of a Data-Driven Control System for a Hydraulic Ex-

cavator , Proceedings of the 2016 International Conference on Artificial Life and Robotics,

pp.393-396, Okinawa, January 2016

[3] Kazushige Koiwai, Liao Yuntao, Toru Yamamoto, Takao Nanjo, Yoichiro Yamazaki, Yoshiaki

Fujimoto : Human Skill Evaluation Based on Control Engineering Approach , Proceedings

98

of the 13th IFAC/IFIP/IFORS/IEA Symposium on Analysis, Design, and Evaluation of Human-Machine Systems, FriITrack4.2., Kyoto, September 2016

[4] Kazushige Koiwai, Ryunosuke Miyazaki, Toru Yamamoto, Koji Ueda, Yoichiro Yamazaki:

A Consideration on Quantitative Evaluation of Machine Responsiveness Based on Control

Engineering Approach , Proceedings of The SICE Annual Conference 2017, pp.531- 534,

Kanazawa, September 2017

[5] Kazushige Koiwai, Ryunosuke Miyazaki, Toru Yamamoto, Koji Ueda, Yoichiro Yamazaki: A

Consideration on Responsiveness Evaluation for an Excavator Based on Control Engineering

Approach , Proceedings of IEEE International Conference on Systems, Man, and Cybernetics2017, pp. 467- 470, Banff, October 2017

99

C.

[1] , PID C

Vol.136 No.5 pp.730–73 (2016)

[2] Design of a Data-Oriented Control System

for Walking Support C Vol.136 No.7

pp. 1013-101 (2016)

[3] Design of a Data-Oriented Evolutionary Controller

C Vol.137 No.7 pp. 914–91 (2017)

[4] Yuntao Liao, Takuya Kinoshita, Kazushige Koiwai, Toru Yamamoto Design of a Performance-

Driven CMAC PID Controller The IEICE Transactions on Fundamentals Vol.E100-A,

No.12, pp. 2963–2970 (2017)

[5] Kazushige Koiwai, Toru Yamamoto, Koji Ueda, Yoichiro Yamazaki Feature Extraction for

Digging Operation of Excavator Based on Data-Driven Skill-Based PID Controller Journal

of Robotics, Networking and Artificial Life Vol. 4, No. 3, pp. 191-194 (2018)

[6] Hiromu Imaji, Kazushige Koiwai, Toru Yamamoto, Koji Ueda, Yoichiro Yamazaki Human

Skill Quantification for Excavator Operation using Random Forest Journal of Robotics,

Networking and Artificial Life Vol. 4, No. 3, pp. 195-200 (2018)

[7] Yuntao Liao, Kazushige Koiwai, Toru Yamamoto Design and Implementation of a Hierarchical-

Clustering CMAC PID Controller Asian Journal of Control (To appear)

D.[1] Kazushige Koiwai, Toru Yamamoto, Takao Nanjo, Yoichiro Yamazaki, Yoshiaki Fujimoto:

Data-Driven Human Skill Evaluation for Excavator Operation , Proc. of 2016 IEEE In-ternational Conference on Advanced Intelligent Mechatronics, pp. 482- 487, Banff, July 2016

100

[2] Kazushige Koiwai, Toru Yamamoto, Koji Ueda, Yoichiro Yamazaki: Feature Extraction for

Digging Operation of Excavator , Proc. of the 2017 International Conference on Artificial Lifeand Robotics, pp.425- 428,, Miyazaki, January 2017

[3] Hiromu Imaji, Kazushige Koiwai, Toru Yamamoto, Koji Ueda, Yoichiro Yamazaki: Human

Skill Quantification for Excavator Operation using Random Forest , Proc. of the 2017 Inter-national Conference on Artificial Life and Robotics, pp.437- 440, Miyazaki, January 2017

[4] Yuntao Liao, Kazushige Koiwai, Toru Yamamoto: Design of a Hierarchical-Clustering CMAC-

PID Controller , Proc. of the 2017 International Joint Conference on Neural Networks,

pp.1333-1338, Anchorage, May 2017

[5] Qiuhao Fu, Kazushige Koiwai, Toru Yamamoto: Design of a Data-Oriented Evolutionary

Controller for a Nonlinear System , Proc. of the 6th International Symposium on AdvancedControl of Industrial Processes, pp.406-411, Taipei, May 2017

[6] Yuntao Liao, Takuya Kinoshita, Kazushige Koiwai, Toru Yamamoto: Design of a Performance-

Driven PID Controller for a Nonlinear System , Proc. of the 6th International Symposium onAdvanced Control of Industrial Processes, pp. 529-534, Taipei, May 2017

[7] Yuntao Liao, Takuya Kinoshita, Kazushige Koiwai, Toru Yamamoto: Design of a Performance-

Driven PID Controller Using Hierarchical-Clustering Based CMAC , Proc. of the 49th ISCIEInternational Symposium on Stochastic Systems Theory and Its Applications, pp.56-57, Hi-

roshima, November 2017

[8] Ryota Sekizuka, Kazushige Koiwai, Seiji Saiki, Yoichiro Yamazaki, Toshio Tsuji, and Yuichi

Kurita: A Virtual Training System of a Hydraulic Excavator Using a Remote Controlled

Excavator with Augmented Reality , Proc. of the 2017 IEEE/SICE International Symposiumon System Integration, Taiwan, December 2017

101

102

103