Maximize parallelism and minimize communication overheads are important issues for distributed memory systems. Communication and data redistribution cannot be avoided even when considering global optimization of data ...Maximize parallelism and minimize communication overheads are important issues for distributed memory systems. Communication and data redistribution cannot be avoided even when considering global optimization of data distribution and computation decomposition. A new approach based on loop fusion is presented exploiting pipelining parallelism, thus communication overhead can be hidden and data redistribution can be avoided. This technique exploits pipelining from complex loop structures, which distinguishes itself from traditional pipelining techniques. Ex-periments show that the technique is superior to other optimizations.展开更多
基础数学函数库是计算机系统非常关键的软件模块,然而国产申威平台上的长向量超越函数只能依靠循环调用系统标量函数来间接实现,该方法无法充分发挥申威平台SIMD扩展部件的计算性能。为了有效解决此问题,实现了申威平台基于SIMD扩展部...基础数学函数库是计算机系统非常关键的软件模块,然而国产申威平台上的长向量超越函数只能依靠循环调用系统标量函数来间接实现,该方法无法充分发挥申威平台SIMD扩展部件的计算性能。为了有效解决此问题,实现了申威平台基于SIMD扩展部件底层优化的长向量超越函数,提出了浮点计算融合算法,解决了两分支结构算法难以向量化的问题;提出了基于Estrin算法动态分组的大阶数多项式实现方法,提高了多项式汇编计算的流水性能。这是在国产申威平台上首次实现长向量超越函数库,提供的函数接口包含三角函数、反三角函数、对数函数、指数函数等。实验结果表明,双精度版本最大误差控制在3.5ULP(unit in the last place)以下,单精度版本最大误差控制在0.5ULP以下,该性能与申威平台直接循环调用系统标量函数相比有显著提高,平均加速比为3.71。展开更多
The hybrid flow shop scheduling problem with unrelated parallel machine is a typical NP-hard combinatorial optimization problem, and it exists widely in chemical, manufacturing and pharmaceutical industry. In this wor...The hybrid flow shop scheduling problem with unrelated parallel machine is a typical NP-hard combinatorial optimization problem, and it exists widely in chemical, manufacturing and pharmaceutical industry. In this work, a novel mathematic model for the hybrid flow shop scheduling problem with unrelated parallel machine(HFSPUPM) was proposed. Additionally, an effective hybrid estimation of distribution algorithm was proposed to solve the HFSPUPM, taking advantage of the features in the mathematic model. In the optimization algorithm, a new individual representation method was adopted. The(EDA) structure was used for global search while the teaching learning based optimization(TLBO) strategy was used for local search. Based on the structure of the HFSPUPM, this work presents a series of discrete operations. Simulation results show the effectiveness of the proposed hybrid algorithm compared with other algorithms.展开更多
A heat transfer model of furnace roller cooling process was established based on analysis of furnace roller＇s structure. The complicated model was solved with iteration planning algorithm based on Newton search. The ...A heat transfer model of furnace roller cooling process was established based on analysis of furnace roller＇s structure. The complicated model was solved with iteration planning algorithm based on Newton search. The model is proved logical and credible by comparing calculated results and measured data. Then, the relationship between water flow velocity, inlet water temperature, furnace temperature and roller cross section temperature, outlet water temperature, water temperature rise, cooling water heat absorption was studied. The conclusions and recommendations are mainly as follows： l） Cooling water temperature rise decreases with the increase of water flow velocity, but it has small relationship with inlet water temperature; 2） In order to get little water scale, inlet water temperature should be controlled below 30 ℃. 3） The cooling water flow velocity should be greater than critical velocity. The critical velocity is 0.07 m/s and water flow velocity should be controlled within 0.4-0.8 m/s. Within this velocity range, water cooling efficiency is high and water temperature rise is little. If cooling water velocity increases again, heat loss will increase, leading to energy wasting.展开更多
The central composite process optimization was performed by response surface methodology technique using a design for the treatment of methyltin mercaptide with modified semi-coke. The semi-coke from the coal industry...The central composite process optimization was performed by response surface methodology technique using a design for the treatment of methyltin mercaptide with modified semi-coke. The semi-coke from the coal industry was suitably modified by treating it with phosphoric acid, with a thermal activation process. The objective of the process optimization is to reduce the chemical oxygen demand （COD） and NH4＋-N in the methyltin mercaptide industrial effluent. The process variables considered for process optimization are the semi-coke dosage, adsorption time and effluent pH. The optimized process conditions are identified to be a semi-coke dosage of 80 g/L, adsorption time of 90 min and a pH value of 8.34. The ANOVA results indicate that the adsorbent dosage and pH are the significant parameters, while the adsorption time is insignificant, possibly owing to the large range of adsorption time chosen. The textural characteristics of modified semi-coke were analyzed using scanning electron microscopy and nitrogen adsorption isotherm. The average BET surface area of modified semi-coke is estimated to be 915 mE/g, with the average pore volume of 0.71 cm3/g and a average pore diameter of 3.09 nm, with micropore volume contributing to 52.36%.展开更多
The optimum control strategy and the saving potential of all variable chiller plant under the conditions of changing building cooling load and cooling water supply temperature were investigated. Based on a simulation ...The optimum control strategy and the saving potential of all variable chiller plant under the conditions of changing building cooling load and cooling water supply temperature were investigated. Based on a simulation model of water source chiller plant established in dynamic transient simulation program (TRNSYS),the four-variable quadratic orthogonal regression experiments were carried out by taking cooling load,cooling water supply temperature,cooling water flow rate and chilled water flow rate as variables,and the fitting formulas expressing the relationships between the total energy consumption of chiller plant with the four selected parameters was obtained. With the SAS statistical software and MATHEMATICA mathematical software,the optimal chilled water flow rate and cooling water flow rate which result in the minimum total energy consumption were determined under continuously varying cooling load and cooling water supply temperature. With regard to a chiller plant serving an office building in Shanghai,the total energy consumptions under different control strategies were computed in terms of the forecasting function of cooling load and water source temperature. The results show that applying the optimal control strategy to the chiller plant can bring a saving of 23.27% in power compared with the corresponding conventional variable speed plant,indicating that the optimal control strategy can improve the energy efficiency of chiller plant.展开更多
文摘Maximize parallelism and minimize communication overheads are important issues for distributed memory systems. Communication and data redistribution cannot be avoided even when considering global optimization of data distribution and computation decomposition. A new approach based on loop fusion is presented exploiting pipelining parallelism, thus communication overhead can be hidden and data redistribution can be avoided. This technique exploits pipelining from complex loop structures, which distinguishes itself from traditional pipelining techniques. Ex-periments show that the technique is superior to other optimizations.
文摘基础数学函数库是计算机系统非常关键的软件模块,然而国产申威平台上的长向量超越函数只能依靠循环调用系统标量函数来间接实现,该方法无法充分发挥申威平台SIMD扩展部件的计算性能。为了有效解决此问题,实现了申威平台基于SIMD扩展部件底层优化的长向量超越函数,提出了浮点计算融合算法,解决了两分支结构算法难以向量化的问题;提出了基于Estrin算法动态分组的大阶数多项式实现方法,提高了多项式汇编计算的流水性能。这是在国产申威平台上首次实现长向量超越函数库,提供的函数接口包含三角函数、反三角函数、对数函数、指数函数等。实验结果表明,双精度版本最大误差控制在3.5ULP(unit in the last place)以下,单精度版本最大误差控制在0.5ULP以下,该性能与申威平台直接循环调用系统标量函数相比有显著提高,平均加速比为3.71。
基金Projects(61573144,61773165,61673175,61174040)supported by the National Natural Science Foundation of ChinaProject(222201717006)supported by the Fundamental Research Funds for the Central Universities,China
文摘The hybrid flow shop scheduling problem with unrelated parallel machine is a typical NP-hard combinatorial optimization problem, and it exists widely in chemical, manufacturing and pharmaceutical industry. In this work, a novel mathematic model for the hybrid flow shop scheduling problem with unrelated parallel machine(HFSPUPM) was proposed. Additionally, an effective hybrid estimation of distribution algorithm was proposed to solve the HFSPUPM, taking advantage of the features in the mathematic model. In the optimization algorithm, a new individual representation method was adopted. The(EDA) structure was used for global search while the teaching learning based optimization(TLBO) strategy was used for local search. Based on the structure of the HFSPUPM, this work presents a series of discrete operations. Simulation results show the effectiveness of the proposed hybrid algorithm compared with other algorithms.
基金Project(2010CB630800) supported by the National Basic Research Program of China
文摘A heat transfer model of furnace roller cooling process was established based on analysis of furnace roller＇s structure. The complicated model was solved with iteration planning algorithm based on Newton search. The model is proved logical and credible by comparing calculated results and measured data. Then, the relationship between water flow velocity, inlet water temperature, furnace temperature and roller cross section temperature, outlet water temperature, water temperature rise, cooling water heat absorption was studied. The conclusions and recommendations are mainly as follows： l） Cooling water temperature rise decreases with the increase of water flow velocity, but it has small relationship with inlet water temperature; 2） In order to get little water scale, inlet water temperature should be controlled below 30 ℃. 3） The cooling water flow velocity should be greater than critical velocity. The critical velocity is 0.07 m/s and water flow velocity should be controlled within 0.4-0.8 m/s. Within this velocity range, water cooling efficiency is high and water temperature rise is little. If cooling water velocity increases again, heat loss will increase, leading to energy wasting.
基金Projects(5114703,51004059/E041601)supported by the National Natural Science Foundation of China
文摘The central composite process optimization was performed by response surface methodology technique using a design for the treatment of methyltin mercaptide with modified semi-coke. The semi-coke from the coal industry was suitably modified by treating it with phosphoric acid, with a thermal activation process. The objective of the process optimization is to reduce the chemical oxygen demand （COD） and NH4＋-N in the methyltin mercaptide industrial effluent. The process variables considered for process optimization are the semi-coke dosage, adsorption time and effluent pH. The optimized process conditions are identified to be a semi-coke dosage of 80 g/L, adsorption time of 90 min and a pH value of 8.34. The ANOVA results indicate that the adsorbent dosage and pH are the significant parameters, while the adsorption time is insignificant, possibly owing to the large range of adsorption time chosen. The textural characteristics of modified semi-coke were analyzed using scanning electron microscopy and nitrogen adsorption isotherm. The average BET surface area of modified semi-coke is estimated to be 915 mE/g, with the average pore volume of 0.71 cm3/g and a average pore diameter of 3.09 nm, with micropore volume contributing to 52.36%.
基金Project(G-0805-10156) supported by US Energy Foundation
文摘The optimum control strategy and the saving potential of all variable chiller plant under the conditions of changing building cooling load and cooling water supply temperature were investigated. Based on a simulation model of water source chiller plant established in dynamic transient simulation program (TRNSYS),the four-variable quadratic orthogonal regression experiments were carried out by taking cooling load,cooling water supply temperature,cooling water flow rate and chilled water flow rate as variables,and the fitting formulas expressing the relationships between the total energy consumption of chiller plant with the four selected parameters was obtained. With the SAS statistical software and MATHEMATICA mathematical software,the optimal chilled water flow rate and cooling water flow rate which result in the minimum total energy consumption were determined under continuously varying cooling load and cooling water supply temperature. With regard to a chiller plant serving an office building in Shanghai,the total energy consumptions under different control strategies were computed in terms of the forecasting function of cooling load and water source temperature. The results show that applying the optimal control strategy to the chiller plant can bring a saving of 23.27% in power compared with the corresponding conventional variable speed plant,indicating that the optimal control strategy can improve the energy efficiency of chiller plant.
