Study Of Intelligence Technology On Application For Dam Safety Monitoring

The focal and difficult problems of dam safety monitoring are studied in this dissertation. The structure of automatic system for dam safety monitoring and the key problems of dam safety intelligent decision and support system (DSIDSS) are studied mainly. The main contents are as follows.(1) Aiming at more reliability of automatic system for dam safety monitoring, the fieldbus structure and the communication model based on fieldbus are applied, which also improves the flexibility of the system. The structure, communication model and functions of the monitoring network are studied mainly.(2) The pre-disposal methods for dam safety monitoring data are studied to deal with its reality and reasonability. The on-line detection models for monitoring data are established, applying the theory of grey system and abrupt change in process, which solved effectively the problem of on-line detection for monitoring data of automatic systems.(3) The analysis methods for monitoring data are discussed, applying neural network. The self-learning BP network model for seepage data based on self-learning neuron is established. The two steps of learning tactics are proposed, which provided a new method for dam safety monitoring.(4) The uncertainty in DSIDSS is solved which is based on the theory and method of fuzzy measure and fuzzy integral. Applying fuzzy sets and probability theory, the method of dam safety classification and the presentation of safety limits are proposed.(5) The knowledge expression and inference technology are discussed. The knowledge base of DSIDSS is established by applying knowledge meaning networks and fuzzy productive rules. The fuzzy inference method made up for the defects of Bayes inference method. The application shows that it is a reasonable and effectual inference method.(6) Based on the study of former dam safety evaluation systems, and applying computer technology, the dam safety intelligent decision and support system is constructed by studying its platforms, structures, sub-functions and cross technology.