The aim of this diploma thesis is to describe localization and calculation load identification of dynamic systems using piezoelectric sensors. Finding methods, which would allow us to evaluate loads on simple systems is the key to their application in the structural health monitoring of more complex systems. A theory necessary for understanding and application in aerospace, civil engineering, automobile industry, and train traffic of presented methods is given in the first part of the thesis. In these applications, the methods of wave propagation and different types of neural network methods are used to evaluate load identification. It is possible to evaluate loads by using a time reverse method, a method based on signal deconvolution, and a method based on a voltage amplitude ratio of the piezoelectric sensor. In the next part, the methods are described, the suitable place for gluing of a sensor, and the number of sensors for using method is given. These methods were verified and compared to a simple experimental system. In the following part, the model of the piezoelectric sensor is presented. It is possible to use the model for calculating voltage output from the strain. For methods verification, the problem of train passage in a specific place of the railway is chosen. The speed of the train and its load on the railway was calculated by using these methods.