Detection Device On Internal Quality Parameters Of Apple

Non-destructive detection of physicochemical indicators in apples is an effective method to analyze the internal quality of apples.To achieve non-destructive prediction of SSC/TAC（the ratio of Soluble Solids Content（SSC）to Titratable Acid Content（TAC））and firmness in apples,this study used dielectric spectroscopy measurement method to collect the electrical parameters of 300‘Fuji’apples from six planting modes at nine frequency points ranging from0.1 to 3980 k Hz,as well as SSC/TAC and firmness values.The correlation between the electrical parameters and physicochemical indicator content was analyzed,and multiple feature extraction algorithms were used to integrate various regression methods to develop the SSC/TAC and firmness prediction models.The optimal prediction model was installed in an industrial computer,and the LCR meter was controlled by the industrial computer to complete the collection,analysis,and prediction of electrical parameters and physicochemical indicators.Finally,a portable prototype circuit for measuring electrical parameters based on Electrical Impedance Spectroscopy measurement technology and Automatic Balance Bridge method was built.The main conclusions of this study are summarized as follows:Aiming at the construction of a prediction model for the SSC/TAC and firmness of apples,this paper analyzed the variation of electrical parameters with frequency and the distribution of physicochemical indicators.The K-S test and correlation analysis method were used to obtain the correlation between them.The non-linear feature extraction method based on Multi-layer Autoencoder and the linear feature extraction method based on Principal Component Analysis,as well as the Full Frequency Spectrum electrical parameter values,were combined with Partial Least Squares Regression,Support Vector Regression and Multi-layer Perceptron methods to construct SSC/TAC and firmness prediction model.The results showed that there was no linear correlation between electrical parameters and physical and chemical indicators.The Multi-layer Autoencoder and Multi-layer Perceptron fusion model had the best effect in predicting SSC/TAC(²00.82,RMSEP02.08)and firmness(²00.88,RMSEP00.66).Aiming at the design of SSC/TAC and firmness measurement system,this paper analyzed the communication mode and interface function of LCR meter,connected to industrial computer through general interface bus,and used industrial computer,LCR-8205 meter and shielding box to build physicochemical indicators detection system.Ba sed on the Python development environment,the optimal prediction model was loaded into the industrial computer.The software interface development and corresponding interface encapsulation were completed through the Py VISA library,Py Qt library and Open Py XL library,and the software package was completed using the Py Installer library.To verify the accuracy of the measurement of the electrical parameters of the system,10‘Fuji’apples were selected to test them.The results showed that the variation law and measurement range of the electrical parameters were consistent with the data distribution law of 300‘Fuji’apples,the average errors of SSC/TAC and firmness prediction were 2.37 and 0.78,and the root mean standard error of SSC/TAC and firmness were 1.97 and 0.71,respectively.The model prediction effect was close to the effect of model pre-training.Aiming at the circuit design of portable electrical parameter measurement prototype,this paper analyzed the measurement range and variation law of electrical parameters,as well as the measurement principle and key links of circuit design.Based on the automatic balance bridge method,a measurement circuit with STM32 as the core was built,including three modules:signal generation circuit,automatic balance measurement circuit and vector proportional measurement circuit.Each module circuit was tested at three preset frequency points and four measurement ranges.However,it can be seen from the test results at different frequency points that there was no obvious distortion and loss in the transmission of the overall signal between the modules,that is,the prototype circuit of the electrical parameter measurement principle met the measurement requirements of the electrical parameters in this paper,and the frequency coverage is between 100 Hz and 5 MHz.The measurement errors of impedance and phase angle of the circuit showed a regular change trend,and the maximum measurement errors of impedance and phase angle were 6.85%and 4.66°,respectively.