Real-Coded Based Immune Mind Evolutionary Algorithm And Its Application In PID Controller Parameters Tuning

PID control is the most common control method used in process control nowadays because of its advantages such as single algorithm, easy adjustment. In the actual applications, many controlled process are high nonlinear, time-variation and lag. In these cases, conventional PID controller can not obtain satisfied control performance. So the PID parameters tuning technology was used to solve the problem of how to select PID parameters.There are many methods for PID parameter tuning, but most of them are difficult to realize or have not very good tuning result and it is often not satisfactory when to compromise the accuracy and speed or its application field is limited.Since there are many advantages of PID control and the tuning of PID parameters is a more difficult problem, many researchers began seeking some artificial intelligence approach to solve the problem of PID parameters tuning. Immune Mind Evolutionary Algorithm (IMEA) is based on the Mind Evolutionary Algorithms (MEA) theory and clonal selection algorithm theory in artificial immune system. It inherited the merits of evolutionary mechamsm of Mind Evolutionary Algorithms such as groups evolutionary strategy and integrated good features of clonal selection algorithm such as self-learning, memory development, antibody diversity, adaptive regulation. So Immune Mind Evolutionary Algorithm has faster convergence and higher accuracy in optimization. For these features, IMEA have theory advantage in the PID parameters tuning.In this paper, based on the theory of IMEA, the real-coded form of IMEA was researched and the definition of clone, variation, recombination and selection operator in real-coded form was proposed. Then the optimize performance of the IMEA was tested by three different special function. On the basis of IMEA theory, this paper applied real-coded based IMEA to the tuning of PID controller parameters and given the specific definition of four operators in PID parameters tuning. The PID controller based on real-coded IMEA was designed and through simulation shows that this control algorithm has better control quality, faster convergence speed and higher control accuracy than traditional PID tuning methods, the PID controller parameters tunned by GA and tuning PID controller parameters tunned by MEA.For the advantages of PID controller based on real-coded IMEA, this paper realized this PID tuning algorithm under embedded real-time operating systemμC/OS- II on PC. Based on those researches, the hardware realization of this controller was studied. First, theμC/OS- II was transplanted into S3C2410A chip which has ARM920T core. Second, the software and hardware realization of parameters PID controller based on real-coded IMEA was studied in S3C2410A embedded development platform. Software, five tasks that monitoring tasks, tuning algorithm tasks, data acquisition, data processing tasks and display tasks were designed and the coordinated running of tasks was completed by the mechanism of message queues. AsμC/OS- II is a sufficient stability and security embedded real-time operating system and the procedures adopted monitoring task to monitor system running, the system stability is very high. Hardware, LCD and keyboard were used as human-machine interface, serial port to communicate with PC. The experiment in embedded development platform shows that the control algorithm has a good work performance and stability. So this controller has high application value and as the foundation of IMEA to use in actual engineering application.