Applications of Artificial Intelligence in Structural Engineering

The civil engineering problems are not repetitive, as the problem definition is always influenced by a number of factors like financial modes, importance of structure and site conditions and so on. Therefore, although the use of computers in structural analysis started almost four decades ago, the profession has not been able to make use of computers fully, especially, for structural design and planning. This is mainly because of problem specific nature, need for logical reasoning, feasibility constraints and use of past experience required in actual design process and planning. Expert systems have capabilities to incorporate some of these requirements for programming a machine for solving a design problem. Artificial Intelligence (AI) is a very versatile and potential technology in the field of computer technology, which enables computer users in various fields to solve problems for which algorithmic approach cannot be formulated and which normally requires human intelligence and expertise. Expert Systems (ESs) and Artificial Neural Networks (ANNs), the best known manifestations of AI, have today gained immense credibility and acceptance in many professional fields. Artificial neural networks are biologically inspired in the sense that neural network configurations and algorithms are usually constructed with the natural counterpart in mind. The tremendous processing power of human brain is basically the result of the massively parallel processing units called neurons. A human brain functions with hundreds of thousands of such biological neurons, which are interconnected by a highly complex network. Every neuron consists of a cell body, axon and dendrites. Dendrites extend from the cell body to the other neurons where they receive signals at a connection point called the synapse. These inputs are communicated to cell body where all such inputs are essentially summed up. If the resulting sum exceeds a specified threshold value, the cell fires and a signal is sent down the axon. Using this model, an artificial neuron is developed which performs the basic characteristics of the biological neuron. ANNs consist of small processing units called nodes, which operate in parallel, and these nodes are densely interconnected by elements called weights. The information to be stored is fed at the input and small values are assigned to the weights. The weights are modified until the output of the network is satisfactory. The artificial neurons can be arranged in a network in a variety of ways by changing the type of connectivity, the number of neurons and the number of layers. In a multi layer arrangement, the input and the output layers are separated by a number of hidden layers.