基于FWA-RFNN的散装物料智能装车系统

Intelligent loading system for bulk materials based on FWA-RFN

  • 摘要: 针对煤矿散装物料装车过程中偏载严重、误差大等问题,提出了一种基于烟花算法(FWA)优化递归模糊神经网络(RFNN)的散装物料智能装车系统。将列车车厢速度的实测值与设定值进行比较,得到偏差作为RFNN控制器的输入,通过RFNN控制器对偏差进行模糊化、动态记忆调节、去模糊化等处理,并利用FWA对RFNN权重进行优化,使RFNN控制器自适应输出修正后的控制参数;依据散装物料装车计量模型,根据各传感器采集的物料质量、物料高度、车厢装载过程中的行驶距离及RFNN控制器输出的控制参数,求得所需调节的牵引电动机频率,进而改变牵引电动机转速,从而调整列车车厢速度,实现散装物料的无偏载装车。实际应用表明,经FWA优化后的RFNN控制器可快速调节车厢速度,且保持速度稳定,满足多车厢分布均衡装载的要求,同时提高了装车精度。

     

    Abstract: In order to solve the problems of serious unbalanced loading and large errors during the loading of bulk materials in coal mines, an intelligent loading system for bulk materials based on the fireworks algorithm (FWA) optimized recursive fuzzy neural network (RFNN) is proposed. By comparing the measured value and the set value of train carriage speed, the deviation is obtained as the input of RFNN controller. The deviation is processed by RFNN controller for fuzzification, dynamic memory adjustment and defuzzification. FWA is used to optimize RFNN weight so that RFNN controller can self-adaptively output the corrected control parameters. The traction motor frequency is obtained by the bulk material loading metering model, the material quality, material height and distance traveled during carriages loading collected by each sensor and the control parameters output from RFNN controller. Furthermore, the traction motor speed is changed so as to adjust the train carriage speed and realize the unbalanced loading of bulk materials. It has been proved that RFNN controller optimized by FWA can quickly adjust the carriages speed and keep the speed stable so as to meet the requirements of distributed and balanced loading of multiple carriages and improve the loading accuracy at the same time.

     

/

返回文章
返回