Skip to main content
. 2014 Aug 4;14(8):14106–14130. doi: 10.3390/s140814106

Algorithm 1 The proposed algorithm (see Table 1 for symbols meaning). Details on weighted comparisons are given in Equations (10)(13).

Input: ωn−1, pn−1, history
Output: pn
1: COPComputeCOP (pn−1, ωn−1)
2:
p¯IV Fnposition ofmax{COP}
3: nLinear Prediction(history)
4:
PIV FP(p¯IV Fn)
5: if PIV F > confidence level then
6: pnp¯IV Fn
7: else
8: δ ← 0.95
9: repeat
10:   for all p : Cn(p) ≥ δ × max{COP} do
11:    if maxscore < ψ(p) then
12:     maxscoreψ(p)
13:    end if
14:    if maxTM < Tn(p) then
15:     maxTMTn(p)
16:      TMnp
17:    end if
18:    if maxP < P (p) ∨ maxPScore < ψ(p) then
19:     maxPP (p)
20:      Pnp
21:    end if
22:   end for
23:   if WeightedComparison_P_TM (Pn,TMn) then
24:     pnPn
25:   else
26:    if W eightedComparison_IV F _T M (p¯IV Fn,TMn) then
27:      pnp¯IV Fn
28:    else
29:      pnTMn
30:    end if
31:   end if
32:   δδ − 0.2
33: until maxscoreδ ≤ 0
34: end if