| After nearly 40 years of development,near-infrared spectroscopy technology has been applied to the internal quality detection of many types of fruits.For traditional electronic weighing fruit sorting devices,only external quality sorting can be performed on fruits,which cannot satisfy the current people.There is a demand for the internal quality classification of fruits,and the high cost of professional fruit internal quality sorting devices is not conducive to promotion.In order to reduce the production cost of fruit internal quality sorting devices,a traditional electronic weighing type fruit sorting device transformation platform was developed.The internal structure of the heavy sorting device was modified to add a sugar sorting module to achieve online sorting of fruit weight and quality at the same time.In this paper,navel orange and peach are used as experimental objects to compare and analyze the anti-spurious light ability of two different integrated optical fiber probes in the sugar sorting module,and optimize the structural parameters of the integrated near-infrared diffuse reflection spectroscopy optical fiber probe in the sugar sorting module.At the same time,it also explores the influence of fruit physical factors such as fruit core and size effect on the detection accuracy of the fruit internal quality sorting model under the sugar sorting module,and establishes the optimal navel orange and peach sugar degree sorting model.The main research results and conclusions are as follows:(1)Explore the influence of two different structures of near-infrared spectrum light source emission and reception integrated diffuse reflection detection probes on the accuracy of the sugar quality sorting module fruit internal quality sorting model,and compare and analyze the structural characteristics of the two different integrated detection probes In the detection mode of ring emission and reception,a circular diaphragm can be formed on the surface of the sample through the structure characteristics of its own ring light source,which effectively avoids the interference of specular reflected light and improves the structure of the integrated fiber probe against stray light.The multi-point emission and reception detection methods are relatively weak against stray light.Compared with other diffuse reflection detection methods,the multi-point emission and reception diffuse reflection detection ability against stray light is still strengthened.Contrast and analyze the structural characteristics of the sugar content sorting module detection platform and diffuse transmission fruit internal quality sorting device based on the integrated optical fiber probe,and briefly describe the method of establishing mathematical model,data analysis method and spectral pretreatment method.Two different diffuse reflection integrated optical fiber probes,ring transmission and reception and multi-point reflection and reception,are built on different detection platforms through experiments.They are used for rapid detection of the internal quality of different kinds of fruits,which is the quality of traditional electronic weighing fruits.The modification of the sugar sorting module added to the detection device provides a theoretical basis and reference.(2)Taking Gannan navel orange as the research object,to explore the influence of two different integrated optical fiber probes in the sugar content sorting module on the accuracy of the citrus fruit sorting model.The self-built sugar content sorting module was used to collect the ring launch and The NIR diffuse reflectance spectrum of navel orange under the two integrated fiber probes of receiving and multi-point emission and receiving.On the basis of retaining the effective information of the spectrum of navel orange samples,the pretreatment methods of navel orange samples are compared and analyzed.The pretreatment methods include multiple scattering Correction(MSC),baseline correction(Baseline),first derivative(1st Derivative)and second derivative(2st Derivative),smoothing(Smoothing)preprocessing methods.Finally,the partial least squares regression sugar model under two different detection methods of ring transmission and reception and multi-point transmission and reception was compared and analyzed.The experimental results show that the stability and model prediction of the ring transmission and reception diffuse reflection detection fiber The ability should be better than the multi-point emission and reception diffuse reflection detection fiber,and it is verified that the dynamic online transformation of the sugar content of the traditional external quality sorter is feasible.Among them,the correlation coefficient of the prediction model of navel orange sugar content in the detection mode of ring transmission and reception reaches 0.81,and the root mean square error of the prediction set is 0.46°Brix.(3)Taking peach as the research object,to explore the influence of two different integrated optical fiber probes in the sugar sorting module on the accuracy of the walnut fruit sorting model,using the self-built sugar sorting module to collect two different The near-infrared diffuse reflectance spectrum of peach with integrated fiber probe.The experimental results show that the stability and prediction ability of the multi-point transmission and reception diffuse reflection detection fiber in the peach sugar model is better than that of the ring-shaped transmission and reception diffuser.Compared with the sugar prediction model of navel orange,the reflection detection fiber found that the sugar prediction model of peach is more robust than the sugar prediction model of navel orange,and the prediction model under the diffuse reflection detection method of multi-point transmission and reception is more robust than the ring transmission and The received detection method is higher,the correlation coefficient of the model's prediction set reaches 0.92,and the root mean square error of the prediction set is 0.32°Brix.(4)Taking navel orange as the research object,the influence of sample size difference on the accuracy of online non-destructive detection of diffuse transmission and diffuse reflection of navel orange was mainly explored.The different effects of the size effect of navel orange on diffuse transmission and diffuse reflection devices were analyzed;the near-infrared spectral characteristics of navel orange fruit under three different detection methods were compared and analyzed,and the large fruit,small fruit,Mixed fruit,mixed fructose degree PLS model and size PLS prediction model after MSC pretreatment,of which the small fruit sugar degree PLS model has the best prediction effect,in the diffuse transmission sugar degree prediction model,after MSC pretreatment,mixed fruit prediction The accuracy of the model has been improved,and the robustness of the prediction model is better than that of Xiaoguo.The prediction correlation coefficient of the model is 0.84,and the root mean square error of the prediction is 0.64°Brix.In the size prediction model under different detection methods,the diffuse transmission detection method The prediction accuracy of the size model is lower than that of the two diffuse reflection detection methods.Finally,the sugar model is corrected by the size model.After the correction,the accuracy of the model is not significantly improved.Therefore,in the diffuse reflection detection method,the size model is first used for size measurement.Sorting,and then use the sugar model to sort the sugar content,and finally achieve the simultaneous online sorting of navel oranges of different sizes under different detection methods.The experimental results show that the traditional electronic weighing type fruit sorting device transformation platform equipped with integrated near-infrared diffuse reflectance spectroscopy integrated fiber probe can better achieve fruit weight,internal quality and online sorting at the same time,saving the fruit quality sorting equipment manufacturing industry cost. |
PDF Full Text Request