Lecture人工神经网络用于石油工程解析.pptVIP

Artificial Neural Networks and Reservoir Engineering Applications Dr. Guojian Cheng Department of Petroleum Engineering, West Virginia University School of Computer Science, Xi’an Shiyou University gucheng@, gjcheng.xsyu@ Lecture Topics An Overview of Computational Intelligence Applications for Petroleum Reservoir Engineering Artificial Neural Networks and Reservoir Engineering Applications Support Vector Machines and Reservoir Engineering Applications Fuzzy System and Reservoir Engineering Applications Evolutionary Computation and Reservoir Engineering Applications Swarm Intelligence and Reservoir Engineering Applications Hybrid Intelligence System and Reservoir Engineering Applications Contents Biological Neurons vs. Artificial Neurons Perceptron Neurocomputing ANN basic issues: Architecture, Learning, Generalization and Types Main ANN Learning Algorithms ANN Models and Functions ANN Applications for PE Summary Biological Neurons vs. Artificial Neurons Biological Neurons encode their outputs or activations as a series of brief electrical pulses (i.e. spikes or action potentials). Dendrites are the receptive zones that receive activation from other neurons. Cell body (soma) of the neuron’s processes the incoming activations and converts them into output activations. Axons are transmission lines that send activation to other neurons. Synapses allow weighted transmission of signals (using neurotransmitters) between axons and dendrites to build up large neural networks. Biological Neurons vs. Artificial Neurons The McCulloch-Pitts (MP) neuron is a class of typical artificial neuron. It is also known as a Threshold Logic Unit (TLU): A set of synapses (i.e. connections) brings in activations from other neurons. A processing unit sums the inputs, and then applies a non-linear activation function (i.e. squashing/transfer/threshold function). An output line transmits the result to other neurons. Single Layer Perceptron We can connect any number of MP neurons t