Study On Synthesizing Quantified Post-evaluation Of Expressway Construction Item

The post-evaluation of expressway construction project is an item after theevent. It is an all-round following, tracking, investigating and evaluating forevery economic targets to approve and initiate a project, programmingdesign, engineering construction and operating management in expresswayconstruction with a whole of the stage. It is for us to assimilate the experienceand lesson from the investment of developing expressway project and toensure the decision-making more scientific, enhancing the management leveland investment benefit.Post-evaluation of highway project is always accomplished by calculatingand analyzing a series of index. So it is important and significant to researchand design a set of such index, which reflect the work in every earlier stage,design, construction, operation and economic effect in the highway projectmanagement. In this thesis, based on the achievements of the previousliterature, the post-evaluation system of highway project is studied from fouraspects which including recounting theory foundation, putting forward thenew index, index analysis and synthetically quantitative assessment in thispaper. The main research works in the dissertation are included as follows:1. Based on the conception of highway construction post-evaluation and thefundamental Contents of highway construction post-evaluation reports, themain contents of highway construction post-evaluation including processesevaluation, social economics effect evaluation, objectives sustainabledevelopment evaluation and comprehensive evaluation are studied. Somemethods, indexes and evaluation thinking are proposed to give a supplementand perfection for the study and application of highway construction post-evaluation.2. A new approach based on neural networks and genetic algorithms forgrade to post-evaluation of highway project is established. It is integratedmethod by using non-linearity of neural networks and whole random searchcapability of genetic algorithm, which can search the best appropriate rightweight and framework of neural network by making use of whole searchingcharacteristic of genetic algorithm, which formerly depends on gradsinformation to adjust right weight of network. Lastly, cognition function ofbrain was simulated by artificial nerve network by changing threshold limitvalue of linking point. BP network model based on genetic algorithms wasused to assess indicators value post-evaluation on highway project as historydata be inputting samples of the network. The results are in conformity withcomprehensive index method.3. Through introducing Analytic Hierarch Process, Gray System and MatterElement,Fuzzy Comprehensive Evaluation, a grading model is found inwhich multi-theories are in harmony. As the advantages and disadvantages ofeach theory are complementary, different theories are used at various steps ofcomprehensive post evaluation of highway. Every theories learns from othersstrong points to offset own weakness, so it can bring out the best in eachother.The model also combines merits of the BS and K.J methods together. Tosettle the problem that current conclusion of highway post evaluation ismainly consist of simple results, the evaluating result includes collectivitygrade, white excellent degree, gray bad degree and total gray degree byapplying this model for post-evaluation of highway project, which makes theoutcome of post-evaluation more scientific, impersonal. Certainly, this modelis also compatible to solve the other complex and large scale system such as making decision and forecasting etc.4. The study works in the dissertation are summarized and some questionson this subject which need to be deeply studied are put forward.