Design Of Multi Sensor Signal Detection Error Compensation And Control System For Canned Motor Pump

Canned motor pump,as a completely sealed pump,can be used to transport inflammable,perishable,explosive,volatile,highly toxic and other special carriers,which is widely used in industrial production and life.The traditional canned motor pump fault detection is mainly based on the TRG table,and judges whether there is a fault in the pump according to the area pointed by the pointer of the table.This method is relatively mechanical and it is difficult to determine the specific cause of the fault.In order to better meet the intelligent development of canned motor pumps,this paper studies the information transmission and intelligent control of canned motor pumps,and designs an FPGA-based multi-sensor intelligent detection and control system for canned motor pumps,and uses the ability of neural network nonlinear fitting,Compensate the error and improve the robustness of the system.The main research contents are as follows:(1)Select the sensor types and main components,analyze the fault detection requirements of the canned motor pump,and add a multi-sensor signal acquisition module to the traditional canned motor pump.A hardware platform for multi-sensor signal acquisition of the canned motor pump is built.According to the functional requirements,the module circuit is designed around the FPGA main control chip,and the module function is verified through board-level debugging.(2)Write logic algorithms such as FPGA signal sampling and serial communication,take Pt100 platinum resistance temperature sensor as an example,introduce its collection and communication process,and finally realize the transmission and communication of multi-sensor signals.It can be simple and intuitive in the Qt5.0-based host computer interface Read the parameter information of multiple sensors in real time.(3)Perform error compensation for multi-sensor acquisition signals.The main method is to calibrate the multi-sensor and correct the non-linearity of the output signal.The fully connected neural network using stochastic gradient descent as the optimization algorithm performs error compensation on the data,and ultimately greatly improves the detection accuracy of each sensor signal.Through the system design of this subject,the multi-sensor data acquisition has high accuracy and efficiency,which can effectively detect the data in the pump in real time and ensure the stable operation of the equipment,which is of great significance to the intelligent development of canned motor pump.