Research On Hybrid Immune Intelligent Optimization Algorithms And Its Applications In Complex System Application

Artificial immune system (AIS) has the advantages of solving complex optimization problems, which simulating the behavior of intelligent characteristic such as self-organization, self-learning ability in biological immune system. Modern industrial system is becoming more and more complex while the system modeling, optimization and control of complex system should need high performance algorithms to assist. It is difficult to meet the performance requirements only rely on a single optimization method. Hybrid immune intelligent processing technology not only can provide the effective way for this kind of problem but also is a direction of the development artificial immune system.Inspired by the mechanism of immune system, deeply mining the evolutionary learning mechanisms contained in biological immune system and combining with other intelligent processing method advantages, several hybrid immune intelligent optimization algorithms and its related application was studied in this dissertations. A series of work of this paper was launched from algorithm theory, algorithm design, and performance testing, comparative analysis to the practical application. On the theoretical side, four types of hybrid immune intelligent optimization methods were studied and the performances of the proposed algorithm were confirmed through the simulation experiments. On the aspect of application, the hybrid immune intelligent optimization methods were introduced to provide new practical technology for complex engineering problems such as chaotic system’s active disturbance rejection optimization control and permanent magnet synchronous motor system Multi parameter identification as well as good control performance and satisfactory identification are obtained. The main work can be summarized as follows:1. Considering the competition and cooperation between populations, the thought of Lotka-Volterra in ecology was introduced into the artificial immune algorithm, a competitive cooperative coevolutionary immune-dominant clone selection algorithm (CCCICA) was proposed. The affinity maturation of antibody is enhanced by the local optimization of the immune-dominance, the clone expansion and the adaptive dynamic hyper-hybrid mutation and other factors in the species. The population diversity is evaluated and adjusted by the locus information entropy. All subpopulations share one memory which consists of the dominant representatives of each evolved subpopulation. The high level memory is optimized by using the immune genetic crossover operator. Several best individuals are migrated to subpopulations from the top excellent population based on the predefined condition. Through those operations, information is shared among populations for co-evolution.2. In order to expand the search space of solution, the swarms group is divided into Gather State and Explore State during the search, a novel immune binary-state particle swarm optimization algorithm (IBPSO) is proposed. Elitist learning strategy is applied to the elitist particle to help the jump out of local optimal regions when the search is identified to be in a gather state. This paper propose a concept of explore strategy to encourage particle in a explore state to escape from the local territory. They exhibit a wide range exploration. Moreover, in order to increase the diversity of the population and improve the search capabilities of PSO algorithm, the mechanism of clonal selection and the mechanism of receptor edition are introduced into this algorithm.premature stagnation phenomenon is restrained and redundancy iteration is avoided.3. Integrated with the principle of immune system optimization, the thought of co-evolutionary and particle swarm neighborhood information, an immune coevolutionary particle swarm optimization algorithm model was proposed. In the proposed algorithm, the whole population is divided into two kinds of subpopulations consisting of one elite subpopulation and several normal subpopulations. The best individual of each normal subpopulation will be memorized into the elite subpopulation, during the evolution process. A hybrid method, which creates new individuals by using three different operators, is presented to ensure the diversity of all the subpopulations. Furthermore, a simple adaptive wavelet learning operator is utilized for accelerating the convergence speed of the pbest particles. The improved immune clonal selection operator is employed for optimizing the elite subpopulation while the migration scheme is employed for the information exchange between elite subpopulation and normal subpopulations. The ability of complex problem optimization is improved through this operation.4. A hybrid algorithm integrating the clone selection algorithm with the ant colony algorithm by adaptive fusion (ACALA) based on local optimization search strategy is proposed. In order to increase the diversity of the antibody and improve the search capabilities of ant algorithm, a mechanism of chaotic disturbance is introduced into this algorithm. The operation of clone expansion, immune gene, etc is adopted to enhance the variety of antibody and affinity maturation. The adaptive control parameter is used to achieve the purpose of integrating the clone selection algorithm with the ant colony algorithm organically. Simultaneously, the proposed hybrid algorithm can prevent premature convergence effectively by taking advantage of local optimization search strategy.5.The Immune Binary-State Particle Swarm Optimization Algorithm (IBPSO) is used to optimize the parameters of Active Disturbance Rejection Control.First, we apply the Immune Binary-State Particle Swarm Optimization Algorithm (IBPSO) which possesses the performance of strong global searching ability and fast real-time to optimize the parameters of Active Disturbance Rejection Control. The presented method has been successfully applied to control chaotic system. Furthermore Active Disturbance Rejection Control based on Immune Binary-State Particle Swarm Optimization Algorithm for the chaotic system is constructed. Secondly, Immune Binary-State Particle Swarm Optimization Algorithm (IBPSO) strategy integrated with Active Disturbance Rejection Control and cerebellar model articulation controller (CMAC) combined control is designed for chaotic systems. The ADRC-CMAC is comprised of a cerebellar model articulation controller (CMAC) and ADRC controller. Immune binary-state particle swarm Algorithm is used to online tune the Parameters of the ADRC-CMAC. According to results of discrete chaotic system show that the presented two control method has better control performance and strong robustness.6.A novel Parameter identification approach to PMSM based on Immune Co-evolution Particle Swarm Optimization algorithm (ICPSO) is proposed which using the advantage of ICPSO with large space and fast parallel search capability.Finally, the proposed method is further verified by its application in multi-parameter estimation of permanent magnet synchronous machines, which shows that its performance is much better than other PSOs in simultaneously estimating the machine d-q-axis inductances, stator winding resistance and rotor flux linkage. In addition, it is also effective tracking the varied Parameter.