Research And Development On Temperature Anomaly Detection System Of Media Heated By Microwave Based On Deep Learning

As a new kind of heating technology,microwave heating could replace traditional heating methods for large industrial heating,because it has the advantages of high efficiency,no secondary pollution and rapid heating.Compared with traditional heating methods such as heat transfer,microwave heating is a high efficiency heating method that directly interacting with heating medium.Using this heating method can greatly improve the utilization ratio of energy,so as to achieve the effect of energy conservation and environmental protection,which is of epoch-making significance.But the microwave heating process,the ability of the heating medium to absorb microwaves changes with time and temperature.And the microwave heating process,which involves complex coupling between time-varying electromagnetic field and thermal field,is extremely complicated.At this point,the heated medium may produce local overheating.Worse,it may cause unexpected safety accidents,such as burning and even explosion.Besides,due to the microwave magnetron is limited by industry,there is a nonlinear problem in the output of microwave power.In addition,there are highly instable factors such as high nonlinearity and strong coupling in the heating process.Therefore,it is difficult to build a precise mathematical model of the heating process.In the actual research process,the approximate mathematical model of microwave heating process is proposed.However,there are differences in the actual controlled object and the approximate mechanism model,so the dynamic characteristics of the system can not be accurately reflected,which makes the stability of the system poor and can not meet the expected performance requirements.Aiming at the problem of local overheating in microwave heating process,a method of deep learning feature extraction based on multi-dimensional data is proposed in this paper.The convolution neural network model is constructed to extract the deep data characteristics of the multi-dimensional data in the microwave heating process.Based on the extracted data characteristics,the temperature local superheat anomaly detection model is constructed,the temperature change of the microwave heating process is detected in real time.The microwave heating power is adjusted according to the feedback of the detection information to achieve temperature control.The algorithm consists of convolutional neural networks(CNNs)and unsupervised learning method named Isolation Forest algorithm(IFA).Firstly,CNNs is utilized to extract features of the data collected from a WXD15 S microwave heating system.Then,IFA detects the local overheating.Compared with the algorithm with common model,experiment results show that the proposed algorithm owns better measurement performance and higher precision.