摘要
本文根据模糊线性规划和神经网络的特点,导出了用神经网络求解模糊线性规划的方法;考虑到神经网络电路中的放大器单元和电阻元件客观上存在误差,提出了用误差补偿法减小这些误差对神经网络输出精度的影响。计算机仿真和电路实验都表明了这种方法的有效性。
A method for solving fuzzy linear pregramming with neural network is described and some experimental results are given. The method for error compensation is presented to minimize the effect of errors of amplifiers and connecting resistors.
关键词
神经网络
模糊数学
线性规划
Neural network, Fuzz mathematics, Linear programming
