Font Size: a A A

Application Research Of Intelligent Algorithm In The Water Conservancy Project

Posted on:2014-08-07Degree:MasterType:Thesis
Country:ChinaCandidate:C L HanFull Text:PDF
GTID:2268330425952921Subject:Agricultural Soil and Water Engineering
Abstract/Summary:PDF Full Text Request
With the high-speed development of national economy and science and technology, inour country, numerous and more and more large concrete structures of building of waterconservancy projects have been built. After a large number of hydraulic structures built,the engineering maintenance has become the main task of the management staff. How toprotect large span aqueduct, bridges and other important buildings safe operation; How toreal-time dynamically monitor the structural health and assess the damage and provide areference for maintenance have become the question which the modern water conservancypeople should consider. The water conservancy project operation management work isinvolved in a wide range and very multifarious, including risk forecast control, structuralhealth monitoring, structural damage evaluation. In view of this, during the operation ofwater conservancy projects, this paper took the two aspects of the wind speed forecastingin wind vibration control and the structural damage evaluation of concrete cracks diseasesare as the research object. Selecting the two aspects as the research object is based on thefollowing situation:(1) Predictive control in operation of the hydraulic structure, its premise is predicted.Give full play to the value of buildings, requires to predict structure risk incurred duringthe process of running, the load forecasting is an important aspect of structure control, andthe wind load is the structure load forecasting focus and difficult control. Every year due tohigh winds led to the destruction of many buildings in China, only with accurate windspeed forecast, to control them effectively, thus the prediction accuracy of wind speed putforward higher requirements. Due to wind speed has the characteristics of randomness,volatility, and indirect, traditional wind speed forecasting methods (such as the continuousmethod and time series method, etc.) of the prediction effect is difficult to satisfactory.(2) Hydraulic structures in use process, as the growth of the use of time, willinevitably produce the problem such as aging, damage of the structures, thus structuremaintenance has become an important work. Based on whole life cycle cost considerations,it requires, first of all, to understand the damage degree of buildings, and then determinewhether the building need to be fixed. Thus, structure damage evaluation is an important part of the process. But because of numerous water conservancy project building, humanresource is limited, grass-roots management staff of professional quality of restrictedfactors, it is difficult to identify the damage degree of buildings. Again because our countryis in such an earthquake-prone area, due to the earthquake caused the destruction of thebuilding is very big every year, after the earthquake to some important projects, such asaqueduct, Bridges, etc., have to quickly determine the damage degree of its structure,provides the important basis for the repair reinforcement, and make it return to normalwater supply and transportation as soon as possible. How to find a simple and practicalmethod to solve this problem? Traditional structural damage diagnosis are mostly based onthe modal analysis, and based on this diagnostic method is difficult to achieve the goal ofrapid diagnosis of structure damage degree.Author in the study of a large number of literature at home and abroad, in view of thecurrent wind speed prediction precision is low and the problem of low efficiency ofconcrete crack damage detection, this article adopts the technology of intelligent algorithmas the main means to solve the above problems, specific as follows:(1) Wind speed by intelligent algorithm of artificial neural network model (BP andRBF model) and SVM model to predict, and comparing with traditional ARMA model,instance to prove: intelligent algorithm is better than traditional method;(2) In this paper, using the correlation between concrete cracks and structural damage,based on digital image processing technology, the application of intelligence algorithm(TAM neural network and PSO-SVM model) model to evaluate the damage degree ofconcrete slabs cracks, to evaluate the results as a grass-roots staff to repair the structure ofthe important reference. Instance to prove: intelligent identification algorithm is simple andquick, only need for image acquisition cracks of the structure, and then import theautomatic identification system, then gives structural damage degree of the evaluationresults.Based on the actual wind speed data were collected and concrete slab crack imagedata, for wind speed prediction and the crack damage evaluation model of intelligentalgorithm are studied, and the wind speed prediction model and the fracture effectevaluation model is analyzed and compared. The specific contents of the study are asfollows:(1) Today’s popular main intelligent algorithm are introduced, and its basic principle;(2) Introduces the model of wind speed forecasting methods: traditional ARMAmodel and intelligent algorithm (BP models and RBF model, the SVM model), and thewind speed prediction model is established;(3) After the comparative analysis of wind speed forecasting results SVM wind speedforecasting model not only forecast precision is high and relatively stable, and on this basisput forward the improvement direction of the model; (4) Cracks are introduced related technologies of image processing algorithm and thepattern recognition of intelligence algorithm (TAM neural network and PSO-SVMmodel);(5) Cracks were extracted based on binary image feature vector, the TAM and thePSO-SVM evaluation model was established, through the comparative analysis concludesthat TAM diagnosis model has stronger robustness and higher correct recognition rate;(6) Summary of the related work of this paper and its shortcomings.
Keywords/Search Tags:Intelligent algorithm, wind speed forecasting, multi-step prediction, digitalimage processing, TAM neural network, PSO-SVM model, concrete crack, damage assessment(diagnosis)
PDF Full Text Request
Related items