演化性計算在可計算一般均衡建模中之應用：遺傳程式，遺傳規畫與類神經網路之結合

Abstract

在本計畫中,三種綠色導向的能源政策之經濟與環境因素,將在可計算一般均衡模型,H-CCGE結合了以生物學為基礎的尋找技術下,作模擬與分析。這三種能源政策是以三種不同的轉移型態來設計。其目的是要看從一個高汙染化的經濟體過渡到低汙染經濟體的轉移過程是該以何種方式來進行,才是對整體經濟體有較大的利益。這三種轉移過程分成:緩慢轉移、平滑轉移與快速轉移。這樣的模擬結果可以幫助我們瞭解一個具有爭議性的問題,也就是該以何種速率來實行低汙染能源政策,使得我們可以降低對高汙染能源的信賴。 本計畫所採用的分析工具是結合了以生物學為基礎的尋找技術之可計算一般均衡模型,同時亦舉出了高效率計算對公共政策設計的潛在貢獻。

In this project, the economic and environmental consequences of three green-oriented energy policies are simulated and analyzed based on a computable general equilibrium mode, H-CCGE hybridized with biologically-based search techniques. The three policies are designed in accordance with three different types of transition from a high-polluted economy to low-polluted economy, namely, a slow transition, a smooth transition, and a fast transition. The simulation results of these three different transition plans can help answer the question frequently debated among policy makers, i.e., at what rate the supply of low-polluted energies should be increased so that the reliance on high-polluted energies can be reduced. The analytical tool used in this project, i.e., the computable general equilibrium model hybridized the biologically-based search techniques, also illustrates the potential contribution of high-performance computing to the design of public policy.