Study On Maintenance And Strengthening Decision For Deteriorating Bridges

Deficiency of carrying capacity, aging and damage are common problems for deteriorating bridges all over the world. Insufficiency of bridge maintenance funds is increasingly remarkable in many countries in contrast to the fast increasing of bridge maintenance requirement. As a result, maintenance decision has become a major problem for highway management departments. Especially, the road maintenance capital is seriously insufficient in our country at the stage of rapid economic development. Accordingly, it is very necessary and urgent to find good approaches to rationally utilize limited capital.The objective of maintenance and strengthening decision for deteriorating bridges is to coordinate the contradiction of safety and economy in bridge maintenance, that is, to obtain the best maintenance results under very limited resources so as to ensure bridge safety, service function and service life, ultimately, to supply decision support of maintenance and strengthening for bridge management departments. Based on the purpose mentioned above, the author researched problems of bridge maintenance decision from several aspects. The major contents and investigation results of dissertation are given as follows:(1) Based on the linear deteriorating model for bridge decks considered initial state, annual deterioration rate and maintenance effect, a single objective maintenance optimization model and a multi-objective maintenance optimization model of bridge decks are respectively built. The object of the former is to search after the optimal maintenance strategy with least costs and satisfying restriction of average deterioration degree. The object of the latter is to minimize maintenance cost and average deterioration degree simultaneously. Single objective genetic algorithm is used to resolve the problem of maintenance cost minimizing.(2) A maintenance strategy optimization method for bridge decks based on Niched Pareto Genetic Algorithm (NPGA) is presented in this dissertation. Owing to the adoption of fitness sharing, this method keeps good diversity in population all along during the evolution course and supplies decision-making personnel a set of feasible strategies to select according to maintenance engineering. Consequently, the selection scope of maintenance strategies is extended. The method presented here can avoid the defect of failing to consider other influence factors in single objective optimization and overcome some disadvantages of conventional solutions for multi-objective optimization. In this way, the maintenance decision will be more scientific and rational. NPGA is able to search for many optimal solutions along the tradeoff curve and has higher efficiency of search than single objective genetic algorithm. NPGA excels multi-objective genetic algorithm in uniformity and diversity of solutions. But, the appropriate values of tournament size and niche radius must be determinated by sensitivity analysis.(3) A maintenance strategy optimization method for bridge decks based on interactive multi-objective genetic algorithm is presented in this dissertation, which guides genetic search by inputting inaccurate preference information and leads population evolving into optimal comprehensive performance of multiple objects. Fitness sharing technique is introduced to improve the diversity of population and obtain the set of solutions satisfying decision-maker’s preference ultimately. This method compromises on limited maintenance resources and good effect. It is able to consider decision-maker’s preference information relative to engineering requirement and get a set of satisfying maintenance schemes. Consequently, the difficulty of selecting optimal maintenance strategies is reduced. During the course of optimization, the values of two parameters, namely penalty coefficient and interval generations that interactive information is input, need to be determinated cautiously. Too large or too small penalty coefficient is not favorable to search for optimal solutions. If the interactive information is input after too many generations interval, it will fail to lead evolution direction; contrarily, if the interactive information is input too frequently, it will waste manpower and increase calculation difficulty, moreover, increase probability of making mistakes.(4) In the course of optimal selection of bridge strengthening scheme, AHP method is used to calculation the subjective weight of evaluation indexes, information entropy method is used to calculation objective weight. Nonlinear programming method is adopted to calculate combination weight, in the way, subjectivity effectively will be reduced and objective information will be fully utilized. As a result, the robustness of analysis result is improved and decision will be more scientific and rational. Simplified ELECTRE method is adopted for the first time in optimal selection of bridge strengthening scheme. Simplified ELECTRE method can get priority order considered multi-factor comprehensive influence and it is more simplifier than ELECTRE II method in the evaluation process, so, Simplified ELECTRE is more suitable to be programmed.(5) Aiming at the problems of inconsistency of evaluation results from different methods and theory study deviating from practical application in optimal selection of bridge strengthening schemes, a decision support system for optimal selection of strengthening scheme for deteriorating bridge is programmed based on MATLAB. The software firstly evaluates strengthening schemes using different comprehensive evaluation methods, and then combines the inconsistent results. A practical example is provided to show that the software can solve the nonconforming problem of evaluation results in optimal selection of strengthening schemes for deteriorating bridge, reduce work intensity of comprehensive evaluation, improve work efficiency and bring convenience for bridge management department.