The prediction unit (105) is provided with a prediction model for the amount of plating liquid adhesion according to the nozzle pressure, the distance between the nozzles (nozzle clearance) and the steel plate speed. Unit of learning (108), according to the deviation using model presumption of bath adhesion and plating equipment from the actual state of attachment, attachment learning bath volume prediction model and the actual state of bath attached to the universal value deviation. The adaptation unit (106) serves as an adaptive quantity to calculate a temporary departure from the predicted value of the plating bath adhesion model and the actual state value of the plating bath amount. Then have attached bath volume prediction model and use it to the preset preset unit calculus (102), focus on the change of the amount of bath attached to the target value, the appropriate results and learning results were used to select preset, computing. The prediction of the model adhesion is predicted by using the proper compensation and the deviation of the actual plating bath adhesion action, so that the amount of plating solution is controlled accurately.
【技术实现步骤摘要】
【技术保护点】
一种镀液附着量控制装置,该镀液附着量控制装置从镀敷设备接收实际状态信号,并且向镀敷设备发送用于控制镀液附着量的控制信号,其中,所述镀敷设备,将连续进给的带钢浸入熔融浸镀液的镀槽中,在提升时从喷嘴吹出高压的气体,去掉不需要的镀液,从而使带钢附着所需厚度的镀液,其特征在于,具备:附着量推定部,该附着量推定部取得镀液附着量预测模型和带钢速度、气体的压力、喷嘴间隙等实际状态,使用所述镀液附着量预测模型,推定镀液附着量,其中,所述镀液附着量预测模型中记述了带钢速度、气体的压力、 喷嘴喷射位置即喷嘴位置中的喷嘴与带钢之间的距离即喷嘴间隙等和该结果附着的镀液附着量的关系;适应演算部,该适应演算部根据所述附着量推定部的输出和在最靠近当前的操作中实际附着的镀液附着量,求出控制模型与实际的镀液附着状况的背离程度,使用 该背离程度,计算用于使镀液附着量预测模型接近目前的镀敷设备的状态的适应量;适应结果存储部,该适应结果存储部存储求出的适应量;学习部,该学习部着眼于镀敷设备的操作状态的类似性,按照各层,学习所述附着量推定部的输出与实际附着的镀 液附着量 ...
【技术特征摘要】
...
还没有人留言评论。发表了对其他浏览者有用的留言会获得科技券。