ABSTRACT
The text in this paper is for visual purpose only. No rights can be taken from this.
After the deeply analysis of American AASHTO-Ware (or PONTIS), European BriME etc, a degradation prediction and maintenance optimization strategy system for bridge life cycle is designed. The system takes full advantage of mass dynamic data of bridge structure from daily, regular, and special inspections or load tests, and real-time health monitoring. The data are converted by a set of fuzzy membership functions such as Sigmoid, normal distribution etc., then, they are unified and classified to five grades. Hierarchical analysis method is applied to divide bridge into deck system, upper structure, substructure and full bridge. Each structure consists of some components or members. Fuzzy integration (FI) approach can calculate the output of each component or layer based on the input and weights. The risk evaluation vectors of the component or layer can been derived using the principle of maximum degree of fuzzy membership. A fuzzy BP neural network can been auto-trained by the pair of samples from FI method. In final, the BP can give the risk grades of each component, layer or entire bridge, automatically.
In according to bridge structure and parameters, the module library of deterioration and repair effect is established including liner, nonlinear, Markov etc., for instance, reliability has following formula. () β ( t ) = β 0 − [ α 1 + α ′ ( t − t 1 ) ] ( t − t 1 ) λ F ( t 1 ) + ∑ i = 1 n β i ( t ) https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/eq171.tif"/>
here deterioration rate α' = 0, and deterioration index λ = 1 as linear formula; α' ˃ 0 and λ ˃ 1 as nonlinear formula; Σβi (t) represents summary of reliability effect by the n repair procedures.
The library is the base of the economic maintenance optimization analysis of bridge structure life cycle. By taking the inspiration from the investment strategy theory in economics, a formula for dynamic maintenance cost is introduced to describe digitally four factors whose are restricted each other: maintenance cost, the life of bridge, bridge condition, and reliability of bridge in the form of monetary income. Maximum maintenance time and frequency can be calculated by deterioration module. The architecture of the dynamic management system. https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/fig329_1.tif"/>
Repair cost current value as a standard can make the difference of maintenance plans. Its formula is () C p = ∑ k = 1 L C k ( 1 + i ) t 1 + ( k − 1 ) t p https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/eq172.tif"/>
where t 1 is the first repair time; tp : the repair interval; Ck : the k-th repair cost; i: the annual discount rate.
Finally, optimal control methods of multiple targets are used to get an optimal repair plan for a bridge. The system is programmed by pure Java software technology and runs on window NT environment.
