| With the increasing automation of industrial production,traditional static weighing can no longer meet the needs of mass production,and dynamic checkweighers have emerged in this context.The dynamic checkweigher can automatically check the weight of packaged goods.It is usually located at the end of the automated quantitative packaging production line to ensure that the weight of the goods is within the specified range,and any packaging outside the tolerance range will automatically be off-line.The ⅩⅢ(1)dynamic checkweigher studied in this thesis can weigh 400 packaged goods per minute.It can also be used with other additional equipment,such as metal detectors and X-ray machines,to check other attributes of the package.The efficiency of the static weighing method has been greatly improved.However,due to mechanical vibration interference,electromagnetic interference,and various random impact interferences during the acquisition of dynamic weighing signals,it is difficult for the dynamic checkweigher to meet the requirements in terms of speed and accuracy.In order to improve the accuracy of the dynamic weighing system,it is necessary to study various factors that affect the accuracy of the checkweigher,and explore key methods to reduce errors.First of all,this thesis analyzes the current domestic and foreign research standards,product development and corresponding technical status of the checkweigher,and elaborates the development trend of the checkweigher.The influencing factors of the measurement accuracy are summarized,and the influence of the belt vibration and the nonlinear characteristics of the sensor on the measurement accuracy is introduced.The measurement process of the strain gauge resistance load cell is deduced,and the influence of the nonlinear characteristic of the checkweigher sensor on the checkweighing result is analyzed.Secondly,breaking through the limitations of traditional dynamic weighing signal processing methods based on digital filtering and system identification,a dynamic weighing signal processing algorithm based on ADAM improved BP neural network is proposed.The implementation of the algorithm will be introduced in detail in Chapter2 of this thesis.The simulation research and actual test of this method show that this method has accuracy and superiority in anti-vibration application of checkweigher.Then,according to the algorithm and actual project requirements,a set of Qtbased dynamic weighing management system was designed.Introduce in detail the software design of modules such as data display,communication protocol,neural network training,data storage,checkweighing signal and photoelectric signal acquisition,and provide the overall embedded software development ideas and implementation process based on this device.Finally,by inspecting the checkweighing performance of the checkweigher system at different speeds at different speeds,the minimum scale,the maximum scale,and the resonant speed,the effectiveness of the method in this thesis for the nonlinear compensation of the sensor is verified,and the verification of the local algorithm The effectiveness and practicability of the improvement of the weight accuracy,the realization of the 0.5g verification graduation value,meets the national standard ⅩⅢ(1)level of the checkweigher. |
