表示第i个判别指标在输人第k个样本下，属于第j类（A、B、C）的隶属度

Indicates that the i-th discriminant index belongs to the degree of membership of the j-th category (A, B, C) under the input of the k-th sample. The degree of membership can be divided by using the Delphi method to consult and investigate multiple experts. In each factor that affects the ABC classification of spare parts, hit the critical point where the score of each influencing factor belongs to A, B, and C, that is, A and The critical point of the scores of B, B and C. Finally, through the weighted average of the division criteria of the experts, the discriminant boundary of the ABC classification discriminant index of the spare parts and the membership degree of the input value of each factor to the A, B, and C categories are obtained. <br>Output layer: 3 nodes, corresponding to the A, B, and C types of spare parts. <br>Hidden layer: The determination of the number of hidden layer nodes has always been a bottleneck problem in the structural design of BP fuzzy neural network. The selection of the number of hidden layer nodes is related to the number of input nodes and output nodes, and the change of hidden layer nodes will directly affect the internal expression of the network. The number of hidden layer nodes in this paper is selected according to the method proposed by American scholar Hebb

Indicates that the i discrimination indicator belongs to the membership of category j (A, B, C) under the 0th sample of the loser. The division of degree of membership can be investigated by using Delphi to consult multiple experts, and in each factor set affecting the ABC classification of spare parts, each influence factor score belongs to the critical point of A, B, C, i.e. the score threshold of Class A and B, B and C. Finally, by weighting the criteria of each expert, the criteria for the classification of the ABC classification of spare parts are obtained, and the input values of the factors are subordinate to Class A, B and C.<br>Output layer: 3 nodes, corresponding to spare parts class A, B, Class C.<br>Hidden layer: The determination of the number of hidden layer nodes has always been the bottleneck of the structural design of BP's fuzzy neural network. The selection of the number of hidden nodes is related to the number of input nodes and the number of output nodes, and the change of the hidden layer nodes will directly affect the internal expression of the network. The number of hidden nodes in this paper is selected according to the method proposed by Hebb, an American scholar.

It indicates the membership degree of the i-th discriminant index belonging to class J (a, B, c) under the kth sample input. The membership degree can be divided by consulting and investigating many experts with Delphi method. In each factor set that affects the ABC classification of spare parts, the critical points of each influencing factor's score belonging to a, B and C, i.e. the critical points of a and B, B and C are determined. Finally, through the weighted average of the classification criteria of each expert, the discrimination limits of ABC classification criteria of spare parts and the membership degree of input values of each factor to class A, B and C are obtained.<br>Output layer: 3 nodes, corresponding to class A, B and C of spare parts.<br>Hidden layer: the determination of the number of hidden layer nodes has always been a bottleneck problem in the structural design of BP fuzzy neural network. The selection of the number of hidden nodes is related to the number of input nodes and output nodes, and the change of hidden nodes will directly affect the internal expression of the network. In this paper, the number of hidden layer nodes is selected according to the method proposed by American scholar Hebb