论文部分内容阅读
换热网络本身严重的非线性导致了局部极值大量存在而极易陷入局部最优陷阱无法跳出,优化策略产生的领域搜索能力的好坏直接决定了全局搜索和跳出局部最优的能力。收敛精度不仅直接影响着优化程序的效率和求解精度,同时收敛精度的不同取值带来的“误差累积”导致基于原理论的优化方向发生改变。因为计算机以有限位的浮点数来表示实数从而产生舍入误差或者在计算过程中出现有效位数失落,导致在有整型变量优化的换热网络中对有无换热器(0-1)判断中产生不同的结果。通过基于共轭性原理提供优化方向的Powell法来具体分析不同收敛精度对优化方向的影响,将权重较大的精度们在10SP1算例中组合试验分析,特定换热器下具体观察优化路径的改变。
The serious non-linearity of the heat exchanger network leads to the presence of a large number of local extremes and can easily fall into the local optimum trap. The search capability of the field produced by the optimization strategy directly determines the ability of the global search and the local optimum. The accuracy of the convergence not only directly affects the efficiency and accuracy of the optimization program, but also the “error accumulation” caused by the different values of the convergence accuracy leads to the change of the optimization direction based on the original theory. Because the computer to a finite number of floating-point numbers to represent the real number resulting in rounding error or the number of significant digits in the calculation of the loss, resulting in integer variable optimization of the heat exchange network with or without heat exchanger (0-1) Judgments produce different results. Powell method based on the conjugate principle to provide the direction of optimization is used to analyze the influence of different convergence accuracy on the optimization direction. The more accurate weights are combined with the experimental analysis in the 10SP1 example, the specific observation under the specific heat exchanger change.