25-09-2010, 03:44 PM
[attachment=4389]
Genetic Algorithm for Identification of Time Delay Systems from Step Responses
presented by:
Gang-Wook Shin, Young-Joo Song, Tae-Bong Lee, and Hong-Kyoo Choi
abstraction
Abstract:
In this paper, a real-coded genetic algorithm is proposed for identification of time delay systems from step responses. FOPDT (First-Order Plus Dead-Time) and SOPDT (Second- Order Plus Dead-Time) systems, which are the most useful processes in this field, but are difficult for system identification because of a long dead-time problem and a model mismatch problem. Genetic algorithms have been successfully applied to a variety of complex optimization problems where other techniques have often failed. Thus, the modified crossover operator of a real-code genetic algorithm is proposed to effectively search the system parameters. The proposed method, using a real-coding genetic algorithm, shows better performance characteristics when compared to the usual area-based identification method and the directed identification method that uses step responses.
INTRODUCTION
A major characteristic of industrial systems is that there are various constraints, including long dead times and time constants, multi-variable, nonlinear, and non-minimum phase systems An effective measurement method for industrial systems is needed for system identification of long-range dead-time systems among system constraints. The performance and stability of time-delay systems are influenced by dead time. To solve the constraints, the smith predictor, which compensates for dead time and a matched model, was proposed. In a good model case, the merits of the smith predictor include the ability of improving the performance and of disregarding the dead time caused by the characteristics of a closed-loop system. But, to apply the smith predictor effectively, the model must be matched with a real system. If the model is mismatched with the system, the smith predictor will be difficult to apply in the real world. Thus, system identification is very important in time-delay systems.
The systems for identification could be divided into linear and nonlinear types. However, this study considers two types of linear models: FOPDT (First Order Plus Dead Time) system and SOPDT (Second Order Plus Dead Time) system which are mainly used in the process industry. For system identification methods of real world industrial systems, the input signal can have significant influence on identification results. Generally, test signals include pseudo-random binary sequences pulses, steps, ramps, and sinusoidal functions. However, the step test of all these tests is the simplest, needs little equipment, and can even be performed manually.
Also, because a step test can be easily implemented on programmable logic controllers (PLC) or distribute control systems (DCS), this study used the step test function. In this paper, we reviewed usual identification method, which are area-based identification and direct identification for time-delay systems. As well, we proposed an identification method using genetic algorithms that is better than the conventional identification methods for FOPDT and SOPDT systems.
Genetic Algorithm for Identification of Time Delay Systems from Step Responses
presented by:
Gang-Wook Shin, Young-Joo Song, Tae-Bong Lee, and Hong-Kyoo Choi
abstraction
Abstract:
In this paper, a real-coded genetic algorithm is proposed for identification of time delay systems from step responses. FOPDT (First-Order Plus Dead-Time) and SOPDT (Second- Order Plus Dead-Time) systems, which are the most useful processes in this field, but are difficult for system identification because of a long dead-time problem and a model mismatch problem. Genetic algorithms have been successfully applied to a variety of complex optimization problems where other techniques have often failed. Thus, the modified crossover operator of a real-code genetic algorithm is proposed to effectively search the system parameters. The proposed method, using a real-coding genetic algorithm, shows better performance characteristics when compared to the usual area-based identification method and the directed identification method that uses step responses.
INTRODUCTION
A major characteristic of industrial systems is that there are various constraints, including long dead times and time constants, multi-variable, nonlinear, and non-minimum phase systems An effective measurement method for industrial systems is needed for system identification of long-range dead-time systems among system constraints. The performance and stability of time-delay systems are influenced by dead time. To solve the constraints, the smith predictor, which compensates for dead time and a matched model, was proposed. In a good model case, the merits of the smith predictor include the ability of improving the performance and of disregarding the dead time caused by the characteristics of a closed-loop system. But, to apply the smith predictor effectively, the model must be matched with a real system. If the model is mismatched with the system, the smith predictor will be difficult to apply in the real world. Thus, system identification is very important in time-delay systems.
The systems for identification could be divided into linear and nonlinear types. However, this study considers two types of linear models: FOPDT (First Order Plus Dead Time) system and SOPDT (Second Order Plus Dead Time) system which are mainly used in the process industry. For system identification methods of real world industrial systems, the input signal can have significant influence on identification results. Generally, test signals include pseudo-random binary sequences pulses, steps, ramps, and sinusoidal functions. However, the step test of all these tests is the simplest, needs little equipment, and can even be performed manually.
Also, because a step test can be easily implemented on programmable logic controllers (PLC) or distribute control systems (DCS), this study used the step test function. In this paper, we reviewed usual identification method, which are area-based identification and direct identification for time-delay systems. As well, we proposed an identification method using genetic algorithms that is better than the conventional identification methods for FOPDT and SOPDT systems.