Immune Particle Swarm-based ADRC Algorithm And Its Application

Particle swarm optimization algorithm is a random search algorithm which worksby imitating the activities of natural creatures and their swarm intelligence. Becauseit’s simple in structure and rapid in optimization, it has been widely used in scienceand engineering. Large results of study have proved its good comprehensive ability tosearch in solving continuous space and discrete space problems. However people havefound its disadvantages, which other random search algorithms may also have,through practical application, like the inclination to premature convergence and lowconvergence velocity in the later period of evolution. This paper supposes a highefficient immune particle swarm optimization algorithm model. By combining theimmune particle swarm optimization algorithm with the modern active disturbancerejection cybernetics, and based on immune particle swarm, it also renders a methodof active disturbance rejection control of chaos system, and an active disturbancerejection control method of parallel active power filter. This research consists of:1．It offers a new optimization algorithm that can avoid the disadvantages ofparticle swarm optimization algorithm——immune network particle swarmoptimization algorithm. It is an integration of particle swarm optimization algorithm,and the clonal selection mechanism and immunological network theory in artificialimmune system. As the result of test function, the new way can obviously improve themultiformity of particle swarm and the precision of result, and efficiently avoidpremature convergence and low convergence velocity in the later period of evolution.2．Nonlinear chaos system works in a complex way which is quite difficult topredict. Active disturbance rejection control is a method of robust control that canwork without any precise system model, but it’s hard to define the controllerparameters. This research makes use of the strong overall search capability and rapidreal-time reaction of immune particle swarm optimization algorithm, updates theparameter of the controller, applies the result to chaos system control, and builds anactive disturbance rejection controller of chaos system based on the immune particleswarm optimization algorithm. The controller applies to chaos system control very well.3．Parallel active power filter is a nonlinear, multivariable and strongly coupledsystem, but it is difficult to build its precise mathematic model. Active disturbancerejection control is a method of robust control that can work without any precisesystem model, but it’s hard to define the controller parameters. Here we propose anactive disturbance rejection optimizing control algorithm for parallel active powerfilter, and at the same time an overall optimization of the parameters of the activedisturbance rejection controller by improving the immune particle swarm algorithm.Test results show that the control algorithm demonstrated in this paper can not onlywork without any precise system model, effectively inhibit the high frequencyinterference to APF compensation performance, control the active power filter well,but also has good dynamic characteristics and robustness.