An improved clonal selection algorithm was proposed for parameter identification of chaotic systems. Global exploration and local exploration of the algorithm are effectively balanced by two mechanisms of hypermutation and receptor editing. Convergence quality of the algorithm is improved by introducing strategy of learning from elite antibody. Experiments on 10 optimization problems show that the algorithm has good performance in terms of accuracy, convergence speed and stability. Taking synchronization control of unified chaotic system with unknown parameter as research object, synchronization controller was designed reasonably and stability of the synchronization system was analyzed theoretically. By setting synchronization scale factors, a variety of synchronization methods such as complete synchronization, anti-synchronization and projective synchronization of unified chaotic system are realized. Simulation results show that the method realizes accurate identification of unknown system parameter and effective synchronization control of drive-response system. Feasibility and effectiveness of the method are verified.

Estimation of unknown parameters of chaotic systems is a primary problem in chaos control and synchronization, which could be transformed into an optimization problem with multi-dimensional parameter space by constructing a proper fitness function. A hybrid swarm optimization algorithm combining improved bare bones particle swarm optimization algorithm and improved differential evolution is proposed. Particle position update mechanism, mutation operation, crossover operation, and crossover probability factor design are improved, taking into account both diversity of population and convergence rate of algorithm. On this basis, fusion optimization strategy of bare bones particles swarm optimization algorithm and differential evolution algorithm is discussed. Co-evolution of two algorithms is realized. To test algorithm optimization performance, simulation experiments were carried out with six benchmark functions. It shows that the proposed algorithm has powerful global optimizing ability, more stability, fast convergence speed and higher optimizing precision and so on. Lorenz chaotic system was taken as an example to estimate three unknown system parameters.