Fig。 11。  The  goal  position  pi ðkÞ。

Table 6The parameters used in the simulation  environment。

Model Area (m×m) Source  position (m) Q Kx ; Ky rix ; riy Growth rate Initial wind velocity  (m/s)

Farrell’s odor model 100 × 100 (10, 0) 5123。7618 10, 10 2, 3 0。001 (1, 0)

Table 7The parameters used in the decision process。

c1 c2 Sampling  time (s) j (m) k (m)

0。8 0。2 1 40 30

Table 8The parameters used in (17) for the motion  control。

b c a a x vmax (m/s) xmax (rad/s)

0。01 0。5 8 0。5 —0。5 0。8 1。57

Fig。 13。  The prediction errors。 xs   is the real position of the odor   source。

Table 9

The  nine scenarios。

Cases The position of odor source (x; y) The initial wind speed (vx ; vy )

Case 1 (10, 0) (1, 0)

Case 2 (30, 10) (1, 0)

Case 3 (30, —20) (1, 0)

Case 4 (10, 0) (1。5, 0)

Case 5 (30, 10) (1。5, 0)

Case 6 (30, —20) (1。5, 0)

Case 7 (10, 0) (0。8, 0)

Case 8 (30, 10) (0。8, 0)

Case 9 (30, —20) (0。8, 0)

signðwy Þ· jxxi ðkÞ—hxi ðkÞj

j wy j

2 · wx    is the coordinate of the goal position at the y axis if the estimated position is regarded as the origin。

Therefore, in terms of the coordinate system shown in Fig。 11, we have the goal position described by

( xxi   kÞ—hxi ðkÞj

pxiðkÞ¼ hxiðkÞþ j    ð    2

 xxi   kÞ—hxi ðkÞj

jwy j