基于PSO-BP神经网络的船体分段任务包工时定额模型

为解决船体分段任务包工时定额的计算过度依赖线性公式而忽略工时定额与工艺参数之间的非线性关系的问题,提高工时定额计算的效率和精确度,将PSO-BP神经网络技术应用到船体分段任务包工时定额中。通过对影响船体分段中间产品额定工时的工艺参数进行分析,建立多输入单输出的PSO-BP神经网络模型,并应用实际数据对PSO-BP神经网络进行训练,测试仿真结果与实际值之间的误差在允许范围内。验证结果表明,采用PSO-BP神经网络建立船体分段任务包工时定额模型,能对任务包作业工时进行准确预测。

Man-hour Quota Model of Hull Block Working Package Based on PSO-BP Neural Network

In order to solve the problem of over-reliance on the linear formula to calculate the quota man-hour of hull block working package and neglect the non-linear relationship between man-hour quota and process parameters, at the same time to improve the calculation efficiency and accuracy of man-hour quota. The PSO-BP neural network technology is applied to calculate the man-hour quota of hull block working package, and the PSO-BP neural network model with multi-input and single-output is established by analyzing the process parameters affecting the quota man-hour of the hull block intermediate product. The PSO-BP network is trained with actual data. The error between simulation results and actual values of sample test is within the allowable range. The verification results show that the PSO-BP neural network model can accurately predict the man-hour quota model of hull block working package.