【摘要】 提出一种基于神经网络的仿真优化加速 /求解器模块 ,该模块作为笔者开发的仿真优化系统 (SimOpt)的一个有机组成部分 ,以一种新的基概率神经网络 (AiNet)作为核心引擎 ,通过训练和学习过程替代仿真模型进行结果评估 ,从而大大提高了仿真优化的速度。更多还原

【Abstract】 Although simulation optimization has become an important method for complex system optimal design, it holds some flaws (e.g., high time expensive, non-converse calculation, etc). The paper puts forward a simulation optimization acceleration & calculation module based on neural network, which is an organic component of simulation optimization system (SimOpt system) developed by authors. The module takes a new Probabilistic Neural Networks (AiNet) as the kernel engine, which replaces simulation model as simulation result evaluation after its training and learning, thereby, it improves the speed of simulation optimization greatly.更多还原