Fuzzy Petri Nets Reasoning Method Research For Processing Tomato Quality Prediction

Forecasting knowledge is used to predict tomato quality accurately andquantitatively by acquiring environmental data and planting managementinformation.The object of prediction is to avoid the species degradation and qualitydecline in long-term accumulation caused by extensive planting stage managementmodel. And this is the specific requirements of modern agriculture.Managementmeasures can be adjusted precisely and quantitatively through dynamic qualityprediction to obtain better quality with the most reasonable agricultural inputs and laythe foundation for the theory and methods of tomato precision farming.Due to a variety of factors affecting the quality of tomato such as light, temperature,fertilization, and these factors exist mutually coupling and restricting relationship,building a multi-output tomato quality prediction model by a reasonable predictionmethod is a key issue to achieve quality dynamic prediction. Therefore, this paperstudies the reasoning method of quality predictin as a starting point for the followingsections:(1) FPN has the ability to represent fuzzy knowledge, visually describe thelogical relationship between rules.And the accumulation process of tomato quality is atime-scale complex process, including a large fuzzy expert knowledge. So this articleuses fuzzy Petri Nets to establish tomato quality prediction model. For the qualityformation characteristics of more relevance and continuous time, this paper putforward a multi-output fuzzy Petri net reasoning method of forward operator takingsmall. The algorithm is more suitable for the longer chain rule, the more complexmodel system of reasoning. The feasibility of this method has been verified bycomparison.(2) For the relevance features of tomato quality with factors, this paperobtained their correlation matrix by correlation analysis of several important factorsaffecting quality traits. The correlation matrix was made advantage analysis toprovide a theoretical basis for determining tomato quality FPN model input parameters.(3)The FPN quality prediction model is been established by the researchof fuzzy Petri nets reasoning methods and quality association In this model, thequality affecting factors data are the input and the content of the quality traits are theoutput.Based on the above research, the shape-quality multi-instance is predicted by usingFPN quality prediction model. The results show that the maximum average relativeerror of inference results with the actual data is2.16%.The accuracy of the model is tomeet the requirements of prediction. Tomato multi-quality quantitative prediction isrealized. And the inference algorithm of fuzzy Petri net to take the small path issuitable to model and inerfence for complex system of the long chain rule, morenetwork layers...