本文選題：換熱網(wǎng)絡(luò) + 量子粒子群算法��； 參考：《華東理工大學(xué)》2015年碩士論文

【摘要】：自上世紀(jì)全球性的能源危機(jī)以來,世界各國都尤其重視能源的開發(fā)、利用以及對環(huán)境的影響。在能源的使用過程中,換熱網(wǎng)絡(luò)結(jié)構(gòu)與工藝過程的匹配程度影響整個系統(tǒng)的能量綜合利用程度以及投資操作費(fèi)用,因而換熱網(wǎng)絡(luò)優(yōu)化設(shè)計、改造與運(yùn)行對實(shí)施工業(yè)過程節(jié)能降耗具有重要意義。本文基于改進(jìn)量子粒子群算法對換熱網(wǎng)絡(luò)展開全局優(yōu)化研究,并以常壓塔換熱網(wǎng)絡(luò)為例,對網(wǎng)絡(luò)中涉及的流股進(jìn)行局部優(yōu)化。 首先,基于換熱網(wǎng)絡(luò)分級超結(jié)構(gòu)模型,構(gòu)建了有分流換熱網(wǎng)絡(luò)的數(shù)學(xué)模型與無分流換熱網(wǎng)絡(luò)的數(shù)學(xué)模型。基于狼群算法參數(shù)設(shè)置思想對量子粒子群算法參數(shù)設(shè)置進(jìn)行改進(jìn)研究。對有分流且非等溫混合的分級超結(jié)構(gòu)模型,以最小化年綜合費(fèi)用為目標(biāo)函數(shù),采用雙層求解與改進(jìn)量子粒子群優(yōu)化算法相結(jié)合的策略,分別對外層網(wǎng)絡(luò)結(jié)構(gòu)參數(shù)、內(nèi)層分流比和換熱量等工藝變量進(jìn)行全局搜索。與文獻(xiàn)中典型算例進(jìn)行比較驗(yàn)證,表明雙層改進(jìn)量子粒子群算法求解有分流換熱網(wǎng)絡(luò)模型問題具有可行性和有效性。進(jìn)而,提出用單層改進(jìn)量子粒子群算法綜合無分流換熱網(wǎng)絡(luò),即在生成無分流換熱網(wǎng)絡(luò)結(jié)構(gòu)變量的同時,生成換熱負(fù)荷變量。不涉及流股分流的同步綜合模型比有分流的復(fù)雜度與求解時間顯著降低。采用三個典型算例對該方法的有效性進(jìn)行了驗(yàn)證。 最后,對換熱網(wǎng)絡(luò)中的流股進(jìn)行改造研究。采用基于代理模型的全局優(yōu)化方法來優(yōu)化常壓塔的余熱回收過程,在優(yōu)化迭代過程中用kriging代理模型來代替耗時的精確模型評估,優(yōu)化了常壓蒸餾塔中段回流操作條件,提升高品質(zhì)能的利用,而且實(shí)現(xiàn)了能量優(yōu)化并且保證了側(cè)線產(chǎn)品之間的分離精度。
[Abstract]:Since the global energy crisis in the last century, countries all over the world have attached great importance to the development, utilization and environmental impact of energy. In the process of energy use, the matching degree between the heat transfer network structure and the technological process affects the comprehensive utilization of energy and the operation cost of the whole system, so the heat transfer network is optimized. The transformation and operation are of great significance to the implementation of energy saving and consumption reduction in industrial processes. Based on the improved Quantum Particle Swarm Optimization (QPSO) algorithm, the global optimization of the heat exchanger network is carried out in this paper. Taking the atmospheric tower heat exchanger network as an example, the local optimization of the flow streams involved in the network is carried out. Firstly, based on the hierarchical superstructure model of heat transfer network, the mathematical model of the heat exchanger network with shunt and the mathematical model of the heat exchanger network without shunt are constructed. The parameter setting of quantum particle swarm optimization (QPSO) is studied based on the parameter setting idea of werewolf swarm optimization algorithm. For the hierarchical superstructure model with shunt and non-isothermal mixing, the objective function is to minimize the annual synthesis cost, and the strategy of two-layer solution combined with the improved quantum particle swarm optimization algorithm is adopted, and the parameters of the outer layer network structure are obtained, respectively. Internal flow ratio and heat transfer process variables such as global search. The comparison with typical examples in literature shows that the two-layer improved quantum particle swarm optimization (QPSO) algorithm is feasible and effective in solving the model of shunt heat transfer network. Furthermore, a single layer improved quantum particle swarm optimization algorithm is proposed to synthesize the shunt free heat transfer network, that is to say, the heat transfer load variable is generated at the same time as the network structure variable without shunt heat transfer. The complexity and solution time of the synchronous synthesis model without shunt is significantly lower than that of the model with shunt. Three typical examples are used to verify the effectiveness of the method. Finally, the modification of heat transfer network is studied. The global optimization method based on agent model is used to optimize the recovery process of waste heat in atmospheric distillation column. In the optimization iteration process, the kriging agent model is used to replace the accurate evaluation of time-consuming model, and the operation conditions of the middle section of atmospheric distillation column are optimized. Improve the use of high-quality energy, and achieve energy optimization and ensure the separation accuracy of side-line products.
【學(xué)位授予單位】：華東理工大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2015
【分類號】：TQ021.8;TP18

【參考文獻(xiàn)】

相關(guān)博士學(xué)位論文 前3條

1 仇汝臣;原油蒸餾產(chǎn)品質(zhì)量預(yù)測模型化和過程優(yōu)化研究[D];天津大學(xué);2007年

2 高月華;基于kriging代理模型的優(yōu)化設(shè)計方法及其在注塑成型中的應(yīng)用[D];大連理工大學(xué);2009年

3 孫俊;量子行為粒子群優(yōu)化算法研究[D];江南大學(xué);2009年

，

本文編號：1894491