CAJ | 학술논문

为解决灰色神经网络模型中参数不易确定的问题，提出改进的粒子群（IPSO ）算法，寻找灰色神经网络的参数最优解。在寻优过程中引入一个速度阈值，当粒子的飞行速度小于给定阈值时，给粒子施加一个加速动量，重新初始化粒子速度及位置。用训练好的网络预测两个地区的忙时话务量，与未改进的粒子群算法优化灰色神经网络（PSO‐GNN ）模型、灰色神经网络（GNN）模型、反向传播神经网络（BPNN）模型的预测结果比较，比较结果表明，改进的粒子群算法优化灰色神经网络（IPSO‐GNN）模型提高了预测结果的精度。
위해결회색신경망락모형중삼수불역학정적문제，제출개진적입자군（IPSO ）산법，심조회색신경망락적삼수최우해。재심우과정중인입일개속도역치，당입자적비행속도소우급정역치시，급입자시가일개가속동량，중신초시화입자속도급위치。용훈련호적망락예측량개지구적망시화무량，여미개진적입자군산법우화회색신경망락（PSO‐GNN ）모형、회색신경망락（GNN）모형、반향전파신경망락（BPNN）모형적예측결과비교，비교결과표명，개진적입자군산법우화회색신경망락（IPSO‐GNN）모형제고료예측결과적정도。
To solve the problem that the parameters in grey neural network are difficult to determine ,the improved particle swarm algorithm was employed to search the optimums by the introduction of a threshold of velocity .When the particle velocity was less than the threshold ,an accelerated momentum was applied on the particles to reinitialize the particle velocity and posi‐tion .The proposed approach was used to predict the busy telephone traffic of two regions ,and the forecasting results were com‐pared with those of GNN ,PSO‐GNN and BPNN .The results show the high prediction accuracy of the proposed algorithm .