| Unstable processes with time delay widely exist in complex chemical manufacturing industries. Their mathematical model of the plant can be equivalent or reduced to first-order and second-order processes with time delay. system instability, which have been the difficulty of engineering practice. On one hand, since there is no balanced state, they are sensitive to the load disturbance and prone to changes in output, thus it is difficult for the system to maintain to a new steady-state process, on the other hand, the lag item, which present in the control loop or feedback loop, resulting in later transporting measuring signals to the controller, thus causing the system overshoot increasing, longer adjustment time and closed-loop control stability reducing. What’s more, most of the actual production processes are complex and dynamic process model deviation will have a greater impact on the design of the controller.In this paper, the twice optimal control theory was used to a class of second order process with time delay system. first optimization establishes a finite-dimensional space time-controlled object model, using linear ITAE criteria to achieve optimal system first optimal control. Then return to the infinite dimensional space, designing the infinite dimensional viewer and controller to complete the second optimization, and give the controller parameters formula. The control system exhibits rapid and smooth, good robustness of parameter perturbations and load disturbance characteristics.Due to the imprecise mathematical model, unobserved state of each connection between the accused and the time lag item, a prediction model approach is added to solve this problem in this paper, making the twice optimal control structure design can directly pick up the variables among the parts, effectively simplify the design process.As a result of the increasing order of the time-sharing model, the amplitude oscillations phenomenon appears in the system during the transition, this paper puts forward the structure pruning aiming at twice optimal method. By increasing the value of local feedback K/to the strengthen the system damping in the middle band, but followed slow down the rising speed inevitably. Correspondingly, the open-loop amplification factor Kcl is also need to increase to ensure the smoother and faster system response process.Finally, simulation examples are illustrated in the case of load disturbances and parameter perturbations to prove the superiority of this method, through before and after structure pruning comparison, as well as other control schemes. The results reveal that the twice optimal control method essentially eliminated controlled object spiral amplitude and phase characteristics of the system in response to rapid and smooth tracking, good robustness; On this basis, after structure pruning, tracking response, stability and robustness have been greatly improved, which possess a high practical value in engineering. |
