【摘要】 针对基于经验值的螺旋输送量机理模型计算精度低,无法为混凝土布料重量自动控制系统提供准确设定值问题,基于螺旋式混凝土布料机输送机理,以螺旋输送量智能预报为目标,通过分析输送量影响因素,确定输入神经元数量,并进一步优化网络参数,进而提出了具有3-3-1形式网络结构的BP(Back Propagation)神经网络螺旋输送量预报方法。仿真结果表明,基于BP神经网络的螺旋输送量预报精度明显高于基于经验值的螺旋输送量机理模型计算精度,且计算结果平稳性好,预报的平均相对误差为0.0389,符合误差要求,可用于混凝土布料重量自动控制系统目标值的设定。更多还原

【Abstract】 The calculation accuracy of the spiral conveying mechanism model based on the empirical value is low, so it is impossible to provide the accurate setting value for the automatic control system of the concrete distribution weight. Based on the conveying capacity of spiral concrete distributor, aiming at the intelligent forecast of spiral conveying quantity, the number of input neurons is determined by analyzing the influencing factors of conveying capacity, and the network parameters are further optimized and then the spiral conveying capacity method of BP neural network with 3-3-1 network structure is constructed. The simulation results show that the prediction accuracy of spiral conveying capacity based on BP neural network is obviously higher than that of spiral conveying capacity mechanism model based on empirical value, and the calculated results are stable. The average relative error of the forecast is 0.0389, which meets the error requirements. It can be used to set the target value of the automatic weight control system of concrete distributor.更多还原