Study On Dominating Functional Zones Based On Immune Genetic Algorithm

To implement Scientific Outlook on Development, transform the ways to achieve economic growth and promote the harmonious development between human and nature, asignificant measure of promoting dominating functional zones has been created in the eleventh five-year plan. With inadequate per-capita natural resources, heavy environment pressure and the growing incoordination between population, economy and environment,there is great difference between different regions in development. In the suggestion of the CPC Central Committee about making the eleventh five year plan, the government put forward the plan of marking out the national land into four types of dominating functionalzones, namely zones optimized to develop, zones important to develop, zones restricted to develop and zones forbidden to develop, to standardize the land development order and form a reasonable land development structure.The core issue of the dominating functional zones planning is how to balance the relationship between the strategic distribution, the development speed, the consumption of resources and the environment costs. Traditional planning solutions and measures have built us good foundations both in theory and in the method aspects. How to choose a better solution to make sure the dominating functional zones are feasible and scientific is to solve. The developed countries have formed separate district planning system and have played active roles in optimizing the space structures, reducing the gap between different regions and solve the regional imbalance, and in China series of district planning have also accumulated a mass of experiences in the theory and practice. There has been a relative comprehensive theory system of priority zones planning. In this case, how to examine the formed theory system is an innovative work, which needs the support of effective analysis method and technology.In this paper, a further study for the dominating functional zones planning has been made. Social economic development data have been widely collected, from which category indexes that can reflect the economic development are selected. With these indexes according to the similarity, the priority zones planning solutions are optimized through intelligent algorithm, which affects influences of human factors and can be analyzed statically and dynamically in the same time. It is significant for realizing the dominating functional zones scientifically, forming a sound space development structures and keeping in sustainable development.The problem of dominating functional zones planning is resolved as a gathering problem hereby. First of all, various indexes that can reflect the economy development status are chosen from regional economy development data. According to different orientations of the different functional zones, the principles of evaluation and the choices of index system differ:1. Zones optimized to develop: give prominence to the quality of development and the full effect.2 Zones important to develop: give prominence to the comprehensive development effect of economic growth, population gathering and urbanization, etc.3. Zones restricted to develop: give prominence to the harmony of economic growth, environment protection and social development.4. Zones forbidden to develop: give prominence to the effects of ecological environment protection.Economy growth, ecological environment protection and harmonious social development, these three aims and principles are regarded as the starting point. In the process of an index system beginning, some index data can be obtained directly from the yearbook or other files. By giving proper quantitative computing systems to the data can not get comprehensive indexes and some difficult-to-quantification data, three first grade indexes, twelve second grade indexes and 70 third grade indexes are designed, from which indexes that can approximately reflect the regional economic development status are separated and abstracted as the influencing factors. By quantizing and weighed comprehensively evaluating the chosen factors to turn the problem into the clustering form in math, the crowding problem model of dominating functional zones partition is given.As for clustering problem in data mining, it is a classifying process by the differences between data of various samples, the differences are the distances between various sample data. In this paper euclidean distances are adopted, by the value of which the performance the classification results are indicated. Optimized algorithm is adopted to optimize the classification result so that the differences between the same regions are smallest and the differences between different zones are biggest. It provides a huge selection space to the solution of the problem by turning it into an optimized model to find the best classification method.The basic genetic algorithm seeks for an optimal resolution by imitating the evolution process in the real world. In the algorithm implementation the binary coding replaces the genetic coding in the reality. In the process of the algorithm, the gene intersections and mutations are iterative processes at random and without introduction, which allows the degradation phenomena while provides the individual the opportunity to evolute. In the aspect of imitating the ability to deal with business, that of the algorithm is far from enough. To learn the biological intelligence, in addition to combine it with the evolution computing and to form an algorithm according to improved evolution theory can enhance the centralperformances and resolve practical problems better. In this paper, the improved genetic algorithm is used to resolve the added optimizing problem. The immune genetic algorithm applies the concept and the theory of immunity in the field of life science to the genetic algorithm. On the basis of maintaining the original features, it makes use of the immunity system to reduce the degradation in the process of execution and improve the convergence of the algorithm. The immune process of the algorithm implements by three steps:1. Elect advanced individual set of iterative population2. Extract binary codes of vaccine from advanced individual set3. Randomly choose immune individual in the population to inject the vaccineThe improved immune genetic algorithm resolves problems by seeking for good chromosome from the gene embedded in the original chromosome. By encoding the decoding of the crowding problem, it creates a function of individual fitness level with the objective function of the clustering problem and seeking for the best classification by incessantly select individuals of better class effect. By using the immune operator, the degradation phenomena in the development process can be overcome and the convergence of the algorithm can be speeded up in the application of clustering so that it will not become of low efficiency and local optimum.The parameters of the algorithm are chosen by incessantly doing experiments on classification effects. After computing and analyzing for many times, a set of optimum parameters are obtained, which are applied to resolve the clustering problem model of the dominating functional zones planning. By iterative optimized computing, the division result in accord with the reality can be obtained, which truly reflected the differences between various regions. The result of clustering analyzing is of significant scientific value and can be used as a valid place for regional governments to make a long-term economic development plan.In the end of the paper, the result of optimized computing is analyzed and summed up. The influence degree to the result by the index selected in the computing is analyzed, which provide reference to the selection of more representative indexes according to the fundamental realities of China. In the mean time, it shows that there are still improving areasfor the classification effect of iterative optimized algorithm, which can provide reference for the further improving and optimizing of the algorithm in future.