The System Design And Optimization For Moisture Meter Based On Thermogravimetric Analysis

A moisture content is one of the vital characteristics of physical and chemical properties in substance.The level of moisture content directly determines the growth of crops and storage time,it determine the degree of production and quality of products in industrial production at the same time.It is of great significance to production and life.Real-time detection and control of material moisture can effectively increase the production,increase storage time and stimulate citizen consumption.At present,the thermogravimetric analysis is the commonly used moisture measurement method,this method is simple and the operation is strong.However,there exist some defects such as poor measurement repeatability,low inductance,and unstable temperature control in the moisture meter system which used the strain gauge sensor.Aiming at the problems of low temperature control accuracy and unstable weighing under different environment during the process of thermogravimetric analysis,a digital moisture meter based on strain gauge load cells is established,the principle of water measurement,measurement accuracy and compensation methods are discussed and investigated in depth.In the thesis,the importance of water detection and research significance are described at the beginning,the commonly used moisture detection methods and the research status of moisture detection in domestic and overseas are introduced.In the meantime,the advantages and disadvantages of domestic moisture testing technology and the influencing factors are summarized.Then,the 220g/1mg moisture analyzer architecture is present based on the low power and high performance single-chip MSP430F5438A,the working principle of weighing,temperature control and management module are depicted in the second chapter.The hardware design and software flow are described to further explain the working principle of the moisture.After analyzing the mechanism of the strain gauge sensor,the thesis presents a new weighing sensor's nonlinear error compensation method based on least squares support vector machine(LSSVM)and mutated modified particle swarm optimization(PSO):This method first uses noise covariance variable Kalman filter algorithm to filter data,reducing the noise influence;and then establishes a model based on LSSVM taking the filtered data and working temperature into consideration;a mutated PSO algorithm is used to optimize the parameters of the model at last.To achieve high precision moisture content detection of the oven,the segmented PID algorithm is used to improve the accuracy of temperature at the same time.Finally,the performance indicators of the moisture analyzer is tested in the thesis,the results show that the moisture content repeatability error is less than 0.5%,and sense of detection accuracy meet 2mg,indicating repeatability errors and other indicators of the moisture meter is in line with national standards,it also has property of stability and high precision.The innovation and deficiencies are summarized,and some suggestion on future research and improvement also proposed at the end of the thesis.