Optimization Of Converter Repair Project Based On Genetic Algorithm

The network planning technology application in engineering project management has been widely used. It tend to optimize single or two goals, But it is more difficult to achieve optimization when deal with multiple targets such as time, cost, quality etc. In the selection of network planning optimization, the targets must be discrete. It must be one-to-one correspondence in multiple objective optimization, so more work must be done when start to solve. For the limitations in structure of the model, it can not completely cover the optimal solution scheme.Multi-objective genetic algorithm is developed in recent years, it is an effective method in the solution of multi-objective optimization and non-inferior. Because of its high efficiency, practical features, so there are more and more academic attentions.This paper is based on the practice in Shanghai Baosteel steelmaking converter project. It put forward the optimizing structure which is based on a multi-objective genetic algorithm, auxiliary network plan and curve fitting methods etc. to build the multi-objective optimization model in converter repairing project management. Considering genetic search can be done in continuous interval, when design the model code, only every component's time limits are judged as coding factors----to deal with the problem that you have to build one-to-one correspondence discrete model when you use multiple coding factor to code. It can reduce the subjective factors'interference to optimal solution. In the known arrangements for the duration of each sub-project, you can solve the corresponding critical path and its corresponding total duration according to the network planning technique. After analyzed the key problem of furnace repair project management, you can get the mutual influence among the time limit , cost and quality, and corresponding equation can also be built through the curve fitting, with the addition of every component's duration, you can get the project cost and quality objectives. Taking time limit, cost and quality as fitness function, transforming the furnace repair project management into a multi-objective optimization problem, using the NSGA-II in multi-objective genetic algorithm to do the genetic Iterative optimization to the model, you can finally get a series of optimization programs which will be provided to the decision makers.This paper analyzed the interactions among the key elements affecting the furnace repair project management, and it reached the relationship and equations between the interactions. In a single target, it can be more accurate to get other targets. And it successfully transforms the furnace repair project management into a multi-objective optimization model. It uses the NSGA-II in multi-objective genetic algorithm to do the genetic Iterative optimization. The NSGA - II algorithm utilizes a non dominated classification method , which is not only approximating the convergence rate, but also maintaining the diversity. NSGA-II algorithm can solve any number of targets, it is a good platform for adding the resource balance in the future.This paper is based on Matlab language to build the model which is practiced in the key projects of one converter furnace repair project, and it get a series of optimal solutions. It is proved that the proposed model has a certain effect and reference value in solving the furnace repair problems and also similar problems.