Research On Solutions Of Physical Contradictions Based On Neural Network

The key factor for enterprises to win in the fierce market competition is the innovation of technological products,and only those enterprises that are good at innovation can survive in the market.Innovation can not only provide excess profits for enterprises,but also provide sustained and strong support for national economic construction.The principle of solving physical contradictions is one of the important tools used in TRIZ to solve the problem of innovative invention.However,when applying the principle of solving physical contradictions to solve more complex problems,there are many principles of recommendation and invention.Researchers still need a lot of energy to analyze the recommendation results one by one.Therefore,this paper integrates and improves the traditional methods of solving physical contradictions by building a neural network model,and passes through a case.Finally,a neural network model for solving physical contradictions is obtained by training examples.Based on the Separation Principle,the Contradiction matrix and the Neural Network Principle,and the method of constructing and training the neural network,this paper optimizes and improves the traditional method of solving physical contradictions.At the same time,the BP neural network model with the optimal number of hidden layer neurons is determined by debugging the neural network.This paper breaks through the traditional thinking of solving physical contradictions,creatively combines four separation principles with 48 general engineering parameters as input of neural network model,and takes 40 invention principles as output of model.Through construction,training and debugging,the BP neural network model of solving physical contradictions is obtained.This paper combs the application process of BP neural network physical contradiction solving model,making the application of this method more perfect and convenient.The research results of this paper are BP neural network physical contradiction solving model and its application process,which combines intelligent algorithm with TRIZ's physical contradiction tool,which makes the solution of physical contradiction more convenient,and provides a powerful model for innovation activities,which can effectively accelerate the development of innovative activities to achieve the purpose of solving innovative problems.