Page 45 - Read Online
P. 45

Page 12 of 27                   Wang et al. Intell Robot 2023;3(4):538-64  I http://dx.doi.org/10.20517/ir.2023.30


                                                                                          
                                   
               tiveness with time,    will decrease with the search step. Therefore, the definition of    is as follows:
                                                                                            
                                                                2  
                                                              −
                                                                =                                      (21)

               4.2.2. Index of uncertainty
               The index of uncertainty       reflects the focus on the grid cells with a high existence probability of the target
               and not being visited for a long time, which is defined as:



                                                                                               
                                                                                                
                                                
                                 1  ∑ ∑           ∏       ∏    [              (             )] 
                                                                                                
                                                                                     −
                              (      ) =            ·             1 −     −      (     +   |      )· 1 −    (     +   |      )  (22)
                                                                                         
                                                                                               
                                                                 
                                                   
                                                              
                                      =1   =1            ∈Ω       (     +   |      )         ∈         
                                                                                               
               where    (     +   |      ) is the estimation of certainty before the detection occurs at time      +   during virtual search
                      −
                          
               processes, and it is equivalent to the certainty of the previous step with the updating method, as shown in
               Equations (14) and (15).
               4.2.3. Index of detection response
               The index of detection response       reflects the focus on the grid cells with previously discovered targets, which
               is defined as:
                                                                                       
                                                                                       
                                        1  ∑ ∑              ∏        [                   ] 
                                                 
                                                                           −
                                      (      ) =           ·           1 −         (     +   |      )·         (23)
                                                                                       
                                                               
                                                          
                                             =1   =1            =        ∈Ω       (     +   |      )  
                                                                                       
               where    −  (     +   |      ) is the estimation of response value before the detection occurs at time      +   during virtual
                          
               search processes, and it is equivalent to the response value of the previous step with the updating method, as
               defined by Equations (17) and (18).
               4.2.4. Index of motion cost
               The index of motion cost       reflects the penalty for UAV motion cost, which is defined as:



                                                                                                 ] 
                                        
                               1  ∑   1                             −1 [    +1  (              ) 
                                                               ∑
                                                                                                  
                                    
                            (      ) =           · C (      (     −1 ),       (      |      )) +           · C       (     +  −1 |      ),       (     +   |      )  (24)
                                                                                                 
                                    =1                          =1                               
               where C = {        } is the motion cost matrix, in which         represents the UAV motion cost with the previous
               flight direction    and the next flight direction   .
               Based on the indices above, the objective function in Equation (19) is defined as follows:




                                  J(  (      ), U(      |      )) =             (      ) +             (      ) +             (      ) +             (      )  (25)


               where      ,      ,      , and       are the weighting coefficients of each index.
   40   41   42   43   44   45   46   47   48   49   50