Research On Data Processing Algorithm Based On Intelligent Agricultural Planting

Intelligent agriculture is the advanced stage of agricultural production,it integrates the emerging Internet,mobile Internet,cloud computing and Internet of things technology,and it relies on a variety of sensor nodes deployed in the field of agricultural production(ambient temperature and humidity,soil moisture,carbon dioxide,images,etc.)and wireless communication networks to realize the intelligent perception,intelligent warning,intelligent decision-making,intelligent analysis,expert online guidance on agricultural production environment,and provide accurate planting,visual management and intelligent decision making of agricultural production.This paper takes a mushroom factory as the research background,and aims to study the data processing algorithms in the field of intelligent agricultural cultivation,to provide intelligent decision-making for agricultural production,the following work has been completed:1.Pre-process the original data collected by the sensor,and do key research on the problem of how to remove the noise data.According to demand of low data distortion,this paper choose k-means clustering method to remove noise data.Through the experiment we found that the running time of the algorithm is long and the clustering result is not stable.Therefore,this paper proposes a k-means algorithm based on the most distant priority strategy(FPKM algorithm).It has been proved by experiments that the algorithm has a significant improvement on the running speed and the clustering results compared with the common k-means algorithm.2.Forecasting the production of mushroom whose growth cycle is not complete,and do key research on how to establish the model of the relationship between growth environment and yield in the cultivation process.After analyzing the requirement of statistical measurement model and neural network forecasting model,this pager choose to use BP neural network to established the model between growing environment in mushroom cultivation process and the final yield.And proposed the plan that use genetic algorithm to optimize the initial weights of BP neural network to solve the problem of too many iterations and easily fall in local optimal solution when using BP neural network to train the prediction model.The experimental results show that the convergence speed of improved BP neural network in the training process becomes faster,and the accuracy of the prediction model is improved.3.Do optimization on the growth environment of mushroom cultivation process,and do key research on how to find the best solution between the model of growing environment in mushroom cultivation process and the final yield.Genetic algorithm is used to optimize the production environment continuously,after many experiments we found that the convergence speed of the traditional genetic algorithm is slow and it easily fall into local optimum.Aiming at this problem,this paper improves the selection operator and crossover operator of the traditional genetic algorithm.It is proved by experiment that the ability to find the best of the improved genetic algorithm is improved.In practical applications,the environmental parameters are difficult to be stable in a certain value,so this paper use the average output value of the surrounding environment to optimize the fitness function,which improves the robustness of the optimal scheme and the realization of the actual operation.4.Using Python language to achieve the above research and integrate them to data processing module according to some certain logic,and then add them to the wisdom of the agricultural production platform.It makes the collected data to be used,and finally through the practical application of the platform to verify the output prediction and environmental optimization method proposed in this paper.The results show that the proposed k-means algorithm based on the farthest priority strategy makes up the defect of k-means algorithm on unstable clustering results and many times of iteration,improves the clustering effect and provides accurate and effective data for subsequent data mining.The BP neural network training algorithm after optimization of weights makes up the lack of the traditional BP neural network,and the yield prediction model which is trained by the improved algorithm is more accurate.The method for finding the optimal solution in mushroom cultivation process based on improved Genetic Algorithm makes up for the traditional genetic algorithm on the problem that the convergence speed is slow and it easily fall into local optimum,improves the ability to find optimal,and enhances the robustness and feasibility of the optimal program.