Prediction Of Aquaculture Production Of Qingdao City By Using Grey Prediction Models

Aquaculture system is a multi-factors, multi-layers, multi-objective grey system, which consists of many complex relationships. Grey system theory was proposed by Chinese scientist Deng Julong in early 1980s to analyze incomplete information. Grey prediction is performed on the grey prediction model, which forecasts the future based on the past, to determine the future change trend of the system in question, thereby providing the decision makers with the scientific evidence. At the first place, this paper briefly presented the basic concept and main content of the grey system theory. Finally, the grey model theory is used to simulate and forecast the developing trend of Qingdao city's aquaculture. In this part, the freshwater aquaculture production of Qingdao city for the next five years was forecasted using kinds of grey prediction models, so as its mariculture production. And based on the linear optimization theory and methods, the optimum mariculture production and area in 2012 is also predicted.Grey relationship degree is the basis of the relation analysis, which is also the important part of the grey theory. The relations between total output of aquatic products and relative structure factors are studied through grey incidence analysis. Results showed that total aquaculture production, especially mariculture production, and culture area, play the main influence on aquatic products, whereas aquaculture production value has litter effect. It further illustrated that shell fish-aquaculture production and the mode of aquaculture in seawater are the superior elements for the mariculture production factor. Fish aquaculture production and reservoir-aquaculture production have relatively great influence. In conclusion, area of briny breed aquatics and species of aquatic product breed aquatics should be firstly taken into consideration when aquaculture programming is designed in Qingdao city. Rationally managing resources and introducing new breed should also be considered to ensure the sustainable development of aquaculture here.Based on the data (1998-2007) of the mariculture production of Qingdao city, GM(1,1) model, Verhulst model and DGM model were separately applied to forecast total aquaculture production of the next five years. Results indicated that the grey prediction model is more suitable for the forecast which is based on short-term data, because the forecast accuracy in short term is higher than that in long-term. After careful discussion, it has been found that this inaccuracy is caused by the differences of the initial data, not the drawbacks of the model.When the accuracy of the GM(1,1) model fitting does not match the established requirement, we can make use of the residual sequence to revise the GM(1,1) model, raising the accuracy. This paper adopted the moving average method to scale the original data, aiming at avoiding possible large fluctuation about the original data. Compared with the original GM(1,1) model , revised GM(1,1) model enjoys obvious higher accuracy.Based on the statistic data of the aquaculture production from 1998 to 2007 in Qingdao City, this paper forecasted the changes of aquaculture production in this region with GM (1, 1) model,the Verhulst model and the DGM model respectively. Results showed that the aquaculture production in Qingdao City will enjoy gradual increase in the next few years, but the growth rate would not be significant. The seawater aquaculture production has reached the carrying capacity of the water environment in the coast of Qingdao, we therefore can think about introducing new techniques such as open-ocean cage culture to promote the carrying capacity, but its potential to improve is limited. While the changing trend of the initial data appears to be an S curve, modeling fitting accuracy of the Verhulst model and DGM model tend to be higher than that of the GM (1, 1) model. By using the way of grey correlation degree, the initial value and the fitting value of the Verhuls and the DGM model are analyzed comprehensively. It shows that actual values and fitting values with the DGM model have a very higher fitting precision than the Verhulst model. Therefore, it is concluded that the DGM model is more suitable for the forecast of the changes in the coming five years of the seawater aquaculture production in the study area. The forecasting results show that the seawater aquaculture production will be maintained at 819 thousand tons.Based on grey prediction model and linear programming model, in the coming 2012, the freshwater aquaculture production will be between 49.5 thousand tons and 75.5 thousand tons. And the reasonable forecasting culture area in 2012 will be 42.34×103 hectare. The output of marine aquaculture will be between 772.7 thousand tons and 893.9 thousand tons.The grey prediction model has the advantages of few sample data required, easy calculation, and high prediction accuracy in short terms etc. What's more, its predicted result usually complies well with the real situation, which demonstrates the grey prediction model has good feasibility. The grey prediction model is based on the grey theoretical and the application to program aquaculture in Qingdao is based on rigorous mathematics theory. It is therefore concluded that the forecasting results can provide scientific evidences for the future aquaculture planning of Qingdao city. After brief review of its developing history and careful summary on its current status, this paper illustrated the main problems in the aquaculture of Qingdao city and further proposed some positive countermeasures and suggestions for the future study.