Font Size: a A A

Early Warning System Construction And Application For Citrus Production And Sale In China

Posted on:2016-12-31Degree:DoctorType:Dissertation
Country:ChinaCandidate:W XiongFull Text:PDF
GTID:1109330461491159Subject:Agricultural Economics and Management
Abstract/Summary:PDF Full Text Request
China is a large citrus producing and consuming country in the world. Since 1978, the citrus industry has developed rapidly. Now it become one of the pillar industries of rural economy in Fujian, Zhejiang, Guangdong, Jiangxi, Hunan, Hubei, Guangxi, Sichuan and Chongqing. The citrus industry has made important contribution to meet the demand of domestic and foreign consumers, increase the income of farmers and torsion the trade deficit. At the end of 2013, the annual output of citrus in China has increased to 33209400 tons, the acreage has increased to 2422200 hectares, ranking first in the world. However, with the increase of citrus production, we also found that the price of citrus usually fluctuates, and the emergencies such as snow damage and frozen injury and citrus fly exposed the fragile and unstable in the citrus production and sales. There is no doubt that citrus producers often suffer large economic losses from these events. In order to balance the citrus production and sales, improve the producers’ income, make the price stability, and ensure the citrus industry developing healthy and orderly. It’s very important for us to establish effective early warning system of citrus production and sales on the basis of advanced early warning system of agricultural products in developed countries.With the citrus production and sale as our research object, we conduct the research from the following aspects:firstly, we constructed the early warning index system of citrus production and sale according to the actual situation of China’s citrus industry and the development characteristics of the macro economy. Secondly, we constructed early warning system of citrus production and sale by using statistical method, entropy method and the BP neural network. Thirdly, we established a risk pre-warning system of citrus emergencies by using case-based reasoning method and scenario analysis method. Finally, we preliminary designed the early warning software of citrus production and sale, and realized the intelligent forecasting and early-warning. The specific research contents and conclusions are as follows:(1)The situation and the influence factors of citrus production and sale. According to the historical data of citrus production and sale in china, we found the following matters:citrus production is high, acreage is large, but yield level is low because of small scale decentralized management and the backward cultivation technology and orchard management. The citrus supply structure is not reasonable with concentrating of citrus mature period, so we must optimize the citrus variety structure. Citrus postharvest handling and processing are still at a low level, and industrialization level is not high. There is no high brand awareness citrus products, at the same time proportion of citrus export is small, most citrus sell on domestic market. In addition, emergency are likely to affect the citrus production and sale because of asymmetric information and the weakness of risk awareness. After understanding the situation and the influence factors of citrus production and sales, on the one hand we can develop targeted policies and measures to eliminate the imbalance between production and sale.On the other hand, we can realize the first link of early warning, lay the foundation for the construction of following early warning system.(2) The establishment of early warning index system of citrus production and sale. Classification and construction of early warning index system is an important part, it can clear the leading, synchronization and lagging index and its’effects of time delay.so it is the basis and a key step to realize the early warning system. Firstly, we identified the citrus producer price growth as alarm indicators and analyzed the causes of alarm from natural sources, exogenous and endogenous sources. Secondly, we set 16 warning indicators. Finally, we utilized time difference correlation analysis, cluster analysis and peak-valley method to make classification of warning indicators about leading, synchronous and lagging properties. We got eight leading indicators, five synchronous indicators and three lagging indicators. These indicators have a good forecasting and monitoring function of production and sale. We can monitor these indicators changing and estimate citrus production and sales alarm occurring according to the range of these indicators.(3) The comprehensive early warning analysis of citrus production and sale. Early warning analysis is a core part of early warning system. On the basis of the early warning indicator system, we used statistical method, diffusion index and composite index, the BP neural network model to alert. Firstly, the early warning statistical method used mainly to set up warning limit and warning degree of warning indicators for forecasting the situation of production and sale. The warning limit and degree of annual data were set up by combining with majority principle, minority principle, median principle, price principle and expert judgment. The warning limit and degree of monthly data were set up by using of value at risk (VaR). According to the warning interval of warning degree of the corresponding, we set up the early warning signal lamp system. Then the warning limit and degree of leading indicators were set up by using the mapping method. Secondly, on the basis of the early warning indicators system of citrus production and sale, this paper utilized climate index method for the early warning of citrus. These indexes include leading, synchronization, lagging diffusion index and composite index. Thirdly, on the basis of the early warning indicators system of citrus production and sale, this paper utilized BP neural network model for the early warning of citrus, the prediction results fitted the actual situation of citrus production and sale accurately. The orange monthly price data as sample, through testing and screening, we construct a suitable combination of orange market price forecasting and early warning model. Then we forecast of the orange market price of short-term by using it and the prediction error of the combination model are very small, it can improve the precision of forecasting and early warning.(4) The early warning system of citrus emergency risk. It can effectively reduce the damage to a minimum caused by emergency risk on all aspects of marketing and avoid the citrus farmers, consumers, dealers and the citrus industry suffered a serious blow by constructing the early warning system of citrus emergency risk timely and reasonable. This paper set up a scenario-case of citrus emergency warning system according to the scenario analysis and case-based reasoning technology, the main conclusions are as follows:First, according to citrus production and sale chain risk characteristics and the common causative factors of emergency, it can be roughly divided into production emergency risk, demand emergency risk and operation emergency risk. Second, according to the scenario analysis and case-based reasoning technology, it can be divided into three steps including the description of the scene, the scene retrieval and scene matching and adjustment. The warning process include risk identification system, risk data collection system and risk assessment and alert. Third, according to the production and marketing chain management of citrus is still in the more primitive and backward state we puts forward corresponding measures such as strengthen the establishment of effective early warning system, early warning information collection system, information sharing platform and emergency plan.(5)The design and implementation of citrus early warning software. Based on early warning analysis of citrus production and sale, this paper make the early warning theoretical analysis method and model into software by using technology of software, provide the early warning information for the user. The software is designed by the most Popular Windows platform application development environment of Visual Studio2010 and Matlab2012a mixed programming. The software have many modules including data input, statistics, alarm, early warning, data query and return. With easy operation principle, construction principle, object driven principle of generality, scalability and maintainability principles for design principles to achieve flexible and high efficient, convenient and efficient management of citrus production and sales information; the full realization of all kinds of warning algorithm, the software has good scalability; simple data add, delete; the powerful data query, data import and export functions.In the end, this paper suggest that we should improve the database of citrus production and sales, enhance the analytic technique of early warning, regulate the information inquiring channel and application of the warning results.In this paper, there are three innovations as follows:First, this paper utilized Time difference correlation analysis, Cluster analysis and Peak-valley method to make classification of warning index about leading, synchronous and lagging properties. The combined methods make the results reliable and avoid sidedness and inaccuracy by using any method of these three.Second, in the analysis of statistical method, the value at risk (VaR) used mainly to set up warning limit and warning degree of warning indicators and signal lamp for forecasting the situation of production and sale. It provides a new method to set the warning limit and warning degree. In the analysis of early warning models, this paper construct the combination of orange market price forecasting and early warning model predict the time series model according to the minimum sum of error square optimal rules, based on the results of intelligent prediction model and time series model. The combination model improves the effect and robustness of forecast and early warning of citrus production and sale. In addition, this paper designed early warning software of citrus production and sales in the foundation of early warning system, laid the foundation for the further application of it.Third, this paper constructed citrus production and sale emergency warning system using scenario analysis and case-based reasoning technology for the first time.
Keywords/Search Tags:Citrus Production and Sales, Early Warning Index System, Early Warning of Statistical Method, Early Warning Index, Early Warning Model, Emergencies, Early- Warning Software
PDF Full Text Request
Related items