Neural network is based on neurons.
Neuron is a biological model based on nerve cells of biological nervous system. When people study the biological nervous system to explore the mechanism of artificial intelligence, neurons are mathematized, thus generating the mathematical model of neurons.
A large number of neurons with the same morphology are connected together from the beginning to form a neural network. Neural network is a highly nonlinear dynamic system. Although the structure and function of each neuron are not complicated, the dynamic behavior of neural network is very complicated. Therefore, neural network can be used to express various phenomena in the real physical world.
Neural network model is described based on the mathematical model of neurons. Artificial Neural Network) S is a description of the first-order characteristics of human brain system. Simply put, it is a mathematical model. Neural network model is represented by network topology, node characteristics and learning rules. The great attraction of neural network to people mainly lies in the following points:
1. Parallel distributed processing.
2. High robustness and fault tolerance.
3. Distributed storage and learning ability.
4. It can completely approximate the complex nonlinear relationship.
In the field of control research, the control problem of uncertain systems has always been one of the central topics in control theory research, but this problem has not been effectively solved. Using the learning ability of neural network, the characteristics of the system can be automatically learned in the process of controlling uncertain systems, so as to automatically adapt to the changes of the system characteristics with time and achieve the optimal control of the system; Obviously this is a very exciting intention and method.
At present, there are dozens of artificial neural network models, and the typical neural network models that are widely used include BP neural network, Hopfield network, ART network and Kohonen network. Learning is one of the most important and remarkable features of neural networks. In the development of neural network, the study of learning algorithm plays a very important role. At present, the neural network models proposed by people are all corresponding to learning algorithms. Therefore, sometimes people don't pray for a strict definition or distinction between models and algorithms. Some models can have multiple algorithms. And some algorithms can be used in many models. In the neural network, learning and training the pattern samples provided by the external environment and storing this pattern are called perceptrons; The ability to adapt to the external environment and automatically extract the changing characteristics of the external environment is called a cognitive device. In learning, neural networks are generally divided into two types: teachers and no teachers. Perceptron has teacher's signal learning, while perceptron has no teacher's signal learning. In Bp network, Hopfield network, ART network, Kohonen network and other major neural networks; Bp network and Hopfield network need teachers' signals to learn. Art network and Khonone network can learn signaling without teachers 49[]. The so-called teacher signal is the pattern sample signal provided by the outside world in neural network learning.