Wave Neural Networks Control And Application Based On Immune Algorithm

The characteristics of a new optimizing algorithm-immune algorithm were discussed. The author presented a model reference adaptive control (MRAC) based on wave neural networks (WNN) optimized by immune algorithm. Then a composite control approach was brought forward which combined wave neural networks (WNN) with digital phase locked-loop (DPLL) to improve performances of induction heating. Simulation results show that the means is effective and feasible.The main works done by author are as follows:1. The author presented a new kind of wave neural networks model reference adaptive control systems based on analyzing some kinds of neural network control model.2. Contrasting the performances of genetic algorithms and immune algorithms, the author put forward an improved immune algorithm and used it to optimize the configuration of the wave neural network.3. The DPLL completed in software was advanced based on comparing the merits and disadvantages of analog PLL with that of DPLL.4. On the promise of holding the control means of PLL, introducing the control ways of WNN optimized by immune algorithms into the control of induction heating power, the author brought forward a multiple control strategies, that is, combined a wave neural networks which optimized by means of immune algorithms with digital phase locked loop circuit. Simulation results show that it is feasible and effective.