現今病毒突變的疾病不斷的衍生,例如SARS、禽流感及H1N1,這些疾病,不僅造成人們生活上的影響更可能造成人們死亡。由於新型病毒的高致死率及高傳染率,也使得衛生政策決策當局將流感疫苗、抗病毒藥劑及傳染阻絕列為新型流感三大防疫策略,除了衛生宣導外,其中,施打疫苗是預防方法中最常使用到的。 本研究主要台北市為例,並以人工智慧法探討具時效性傳染病疫苗施打之最佳化問題,其三種啟發式演算法分別為免疫演算法(Immune Algorithm, IA)、遺傳演算法(Genetic Algorithm, GA)及粒子群最佳化(Particle Swarm Optimization, PSO),希望以此三種方法做為求解工具,討論當疫情爆發時如何施打其疫苗,並使探討時間區間中整體的生病人數為最低。 本研究探討在不同之條件設定下,例如:不同的疫苗時效性、傳染率、接觸數、治療率與疫苗數量對於疫苗施打之效用疫分析在不同情況下應如何施打疫苗來降低生病人數。數值結果顯示,本研究所使用的三種啟發式演算法,在不同設定下之問題中,皆可找到一個有效的方案來降低生病人數。
Nowadays, there are a number of varied diseases appearing frequentely, such as SARS, the avian influenza and H1N1 influenza virus. These have impacts on society like the new disease with a high mortality rate and infection rate. Therefore, the Public Health Bureau has to promote influenza vaccine, antiviral agents and denial of infection. There are three major prevention strategies to control new strain of influenza virus. Especially, vaccination is the most commonly used methods for health promotion. In this thesis, three heuristic algorithms, including an immune algorithm (IA), genetic algorithm (GA) and particle swarm optimization (PSO), are used to explore the best time-sensitive infectious disease vaccination. The numerical results of a case of Taipei showed that the proposed approaches can effectively solve optimization problems. The results can be useful for the government to distribute the limited vaccine injections when the disease outbreaks. In this thesis, different set of conditions, such as different vaccines timeliness, the rate of infection, contact number, and a number of treatment rates, are considered and experimented. Numerical results showed that three heuristic algorithms which used in specific studies can efficiently minimize a number of ill people.