Page 148 - IJOCTA-15-3
P. 148

E.M. Shaban / IJOCTA, Vol.15, No.3, pp.517-534 (2025)







































            Figure 1. Typical structure of the discrete SDP-PID controller
            Abbreviations: PID, proportional integral derivative; SDP, state-dependent parameter; TF, transfer function.


                The PID structure, shown in Figure 1, is best  stability. These innovative plus gains mirror the
            suited for systems with discrete time SDP-TF of   term SDP-PID+. Given the regulator SDP-PID+
            at most second order (n ≤ 2), and unity sample    structure, as depicted in Figure 2, the NMSS rep-
            time delays (δ = 1). To extend its applicability,  resentation for SDP-PID+ control is given in
            the regulator structure is employed, where the ref-  Equation (7).
            erence signal is set to zero, i.e. r k = 0, making the
            error signal e k = −y k , and treating external dis-          x k = F k x k−1 + g k u k−1
                                                                                                          (7)
            turbances as changes in the reference signal, r k ,           y k = h x k
            11–15,17
                    as depicted in Figure 2. This is possible
            because the external set point does not affect the
            control process . 33,34  This regulator form modifies  Equation (7) enables the direct implementa-
                                                              tion of full SVF control using the measured input
            Equation (4) to the following form in Equation
                                                              and output signals of the controlled system with-
            (6).
                                                              out requiring the design of a state reconstructor. 16
                                                              Here, the k th  sample of the non-minimal feedback
                           m
                        −  P  b j+δ−1 {χ k } z −(j+δ−1)       state vector is defined in Equation (8).
                          j=1
                   e k =                          u k
                                 n                                                                          T
                                P                                                z k e k ∆e k ,
                            1 +    a i {χ k } z −i
                                i=1                     (6)                     |     {z    }              
                                                                             Typical PID feedback states   
                                                                                                           
                                                                                     n ≤ 2
                                                                                                           
                                                                                                           

                        −B χ k , z −1                                                                      
                      =              u k                              e k−2 e k−3 . . . e k−(n−2)  e k−(n−1)  ,  
                                                                                                            
                                                                   
                         A (χ k , z −1 )                              |               {z               }   
                                                              x k =       Extra proportional feedback states if  
                                                                                                           
                                                                                      n>2
                                                                                                           
                                                                                                           
            To accommodate higher-order processes (n > 2)                                                  
                                                                                                           
                                                                    u
                                                                                                            
            and numerator polynomials of order greater than         k−1 u k−2 . . . u k−(m+δ−3)  u k−(m+δ−2) 
                                                                   
                                                                   |                  {z                  }
            unity (m + δ − 1 > 1), the SDP-PID+ approach                     Extra input feedback states   
            introduces additional compensators, called plus                    if m ≥ 1 and/or δ>1
            gains, that decelerate control actions and enhance                                            (8)
                                                           520
   143   144   145   146   147   148   149   150   151   152   153