Bacteria Foraging Optimization Algorithm Improvements And Its Application In The Analysis And Calculation Of Water Safety

The ultimate goal of a comprehensive analysis of water safety is investigating and estimating the security situation of an area and the river basin. According to the analysis of the results, taking effective measures is in order to guarantee to the developments of the region and the river basin.Therefore, The most powerful basis for management of the executive branch is the accurate and reasonable water safety analysis results. However, comprehensive analysis method used to achieve this result is an urgent need to explore.Based on the conventional bacterial foraging algorithm, this thesis proposes three kinds of improved algorithms, immune evolutionary-bacteria foraging optimization algorithm, particle swarm-bacteria foraging optimization algorithm, immune evolutionary-particle swarm-bacteria foraging optimization algorithm;Proposes four kinds of universal formulas(the improved Weber-Fechner, logarithmic function, power function, Logistic)and two kinds of universal models(projection pursuit regression,forward neural network) in the surface water, groundwater, lake eutrophication, water, air, ecology, water safety, water resources and the sustainable use of water resources. By the improved algorithm for optimized formulas and parameters in the model, this thesis has been universal index formulas and models in the the application of these comprehensive evaluation of the nine range. A variety of evaluation formulas and models were be applied to the analysis of instances of different regions,which has been a reliable evaluation results. In summary, the four kinds of formulas and two kinds of models have a scientific and universal usability, which provides a scientific method for the hydrology, environmental management executive departments.To sum up, the content and results of this research are as follows:(1) Immune evolutionary-bacteria foraging optimization algorithm, particle swarm-bacteria foraging optimization algorithm and immune evolutionary-particle swarm-bacteria foraging optimization algorithm were proposed. Selecting eight test functions to test performance of the algorithm, results show that the immune evolutionary-particle swarm-bacteria foraging optimization algorithm (IE-PS-BFOA) compared to the previous two hybrid optimization algorithm has certain advantages in the convergence speed and accuracy.(2) Baseding on the index system and grading standards of the surface water, groundwater, lake eutrophication, water, air, ecology, water safety, water resources and the sustainable use of water resources in the typical areas, finding that we can always set the proper reference values and transformed forms for each index, The evaluation criteria were transformed into [0.0943,0.2742], [0.1388,0.3486], [0.1926,0.4193], [0.2739,0.4822], [0.3648,0.5236] range.(3) This research constructes four kinds of universal formulas(the improved Weber-Fechner, logarithmic function, power function, Logistic)and two kinds of universal models(projection pursuit regression,forward neural network) based on normalized index values in the surface water, groundwater, lake eutrophication, water, air, ecology, water safety, water resources and the sustainable use of water resources. These formulas and models are not restricted with how many number of indicators,which makes that the traditional formulas and models have been greatly simplified. These methods are standardized and unified, more pervasive and universal.