Study On Quality Prediction Of Kiwifruit Based On Storage And Transportation Microenvironment Parameters And Neural Network Algorithm

Kiwifruit has a sweet taste and rich nutritional value.It is loved by consumers and is known as the "Pearl in the Fruit".However,due to the thin and juicy skin of kiwifruit,it is easy to lose water and shrink at room temperature.It is prone to rot due to mechanical damage and pathogen infection during the picking,storage and marketing process,and it is not resistant to storage.In the process of long-distance transportation and circulation,the microenvironmental parameters such as temperature and humidity where kiwifruit is located are in dynamic changes,and the quality changes lack effective monitoring and prediction methods.This topic takes "Xuxiang" kiwifruit as the research object,discusses the correlation between microenvironmental parameters（temperature,humidity and gas composition）and quality index changes during storage,and establishes a back propagation algorithm based on microenvironmental parameters and errors（Back Propagation Neural Network,BP neural network）kiwifruit comprehensive quality dynamic prediction model,in order to provide a basis for real-time prediction and monitoring of kiwifruit quality in the circulation process,and reduce economic losses.The results of the study are as follows:1.Taking "Xuxiang" kiwifruit as the research object,the microenvironmental parameters and storage quality changes of kiwifruit under storage temperature of 4℃,10℃ and 20℃ were studied.The results show that temperature is an important factor affecting the microenvironmental parameters and quality of kiwifruit storage.The higher the temperature,the more obvious changes in the microenvironmental parameters of the kiwifruit.The weight loss rate and cell membrane permeability of kiwifruit increase with the extension of storage time,the soluble solid content gradually increases,the hardness,titratable acid content,vitamin C content,color（C*value and L* value）and other quality indicators gradually increase reduce.Low temperature can effectively delay the aging of kiwi tissues,delay the appearance of respiratory peaks,reduce the release of ethylene content,and maintain fresh quality.Kiwifruit can best maintain a relatively stable gas environment under the storage condition of 4℃,and all the quality indexes are significantly better than the storage conditions of 10℃ and 20℃.2.Taking the "Xuxiang" kiwifruit as the research object,combining the advantages of the analytic hierarchy process and the entropy weight method,a quantitative evaluation method of "layer-entropy weight" quality is established.The comprehensive quantitative evaluation of kiwi fruit includes the construction of a fuzzy evaluation matrix of kiwi fruit quality indicators,each quality indicator is dimensionlessly processed,and the weight of each quality indicator is calculated through the analytic hierarchy process and entropy weight to obtain a comprehensive quantitative evaluation value of quality.The results show that the weights of hardness,soluble solids,and vitamin C are relatively high（0.145,0.177,and 0.212,respectively）,while other quality indicators include weight loss rate,color TA content,respiration intensity,ethylene release,and cell membrane permeability.The weight is relatively low（about 0.055-0.079）.The linear weighted summation is used to calculate the comprehensive evaluation value reflecting the overall change of kiwi fruit quality.The quality change of kiwi fruit described by the comprehensive evaluation value is consistent with the sensory observation results.The comprehensive evaluation value calculated by the analytic hierarchy process and entropy weight method can be used as a representative of the overall quality change of kiwifruit,reflecting the quality change process of kiwifruit.3.The parameters of the microenvironment（temperature,relative humidity,carbon dioxide,oxygen,ethylene concentration）of the kiwifruit are the input values,and the comprehensive evaluation value of the kiwifruit quality is the output value.The kiwifruit quality prediction model is constructed based on the BP neural network algorithm,and different working conditions are analyzed.The prediction accuracy of the next model.The results show that the average relative error of the BP neural network model is less than 10%,and the accuracy is high,and it can accurately predict the comprehensive quality changes of kiwifruit during storage.