The Analysis Of Learning Process And Research Of Learning Control Strategy

Learning control was first proposed from the generalized intelligent control which contains learning control, cognitive control, associative decision making and so on. As an important research field of intelligent control, learning control was used in more and more industrial control field, such as control theory and education theory. With the development of the science and technology, learning control has been developed great and achieved some good results. Researching learning control can make people study well, live great and optimize the learning resources allocations in work, improve the service configuration, obtain the higher learning efficiency, alleviate the pressure of learning. So, it has good theoretical and practical value to research the learning control. The achievements of this paper are described as follows:Firstly, according to current research situation of learning control, we introduce the basic idea of learning control. Meanwhile, on the basis of giving the basic idea of reinforcement learning, we analyze the learning rules that reinforcement learning on the point of reward-punishment rule. Then we compare reinforcement, supervised learning and unsupervised learning from neural network control.Secondly, we discuss the principle and classification of iterative learning control. And according to the modeling of learning dynamics process, we study function of positive factors of reward-punishment rule on the individual learning process and results by simulating individual learning process using a nonlinear learning system. The simulation shows that the positive incentives for learners in the process of individual learning will make the actual output of learning system approaching towards the desired objective.Thirdly, we research group learning on the foundation of PBL, and analyze the learning process of PBL based on the constructivism theory. Then, we study the analysis of the PBL learning process of the open-loop and closed-loop and give the advantages and disadvantages of them from the control view. On the basis of these analyses, a learning model of PBL complex task by using decentralized control strategy is built. Finally, experiments show that the learning model proposed can simulate the learning process of complex task.Finally, we summarize the individual learning process and the PBL control process, and put forward some directions of study in the further.