Application of convolutional neural network under nonlinear excitation function in the construction of employee incentive and constraint model

It is aimed to explore the relationship between the incentive constraint model and corporate performance, and expand the application of neural networks in the incentive mechanism, thereby providing a direction for the innovation development of the enterprise to a certain extent. Based on the convolutional neural network (CNN), the construction and practice of the employee incentive constraint model are discussed. First, fully combining the excellent performance of the nonlinear excitation function in CNN, a CNN-based PReLUs-Sigmoid (P-S) nonlinear excitation function is proposed and compared with several excitation functions. Second, the P-S nonlinear excitation function is integrated. Based on the law of diminishing marginal returns, the construction of the employee incentive constraint model is completed. Finally, companies with and without equity constraint mechanisms are selected as the research sample to analyze the relationship between the implementation of the incentive constraint mechanism and the performance level of the company. The results show that the P-S nonlinear excitation function based on CNN has both sparse expression ability and smooth nonlinear mapping correction ability. Also, it has applicability in the optimal solution. When the employee’s work effort is $$x = 2.5743$$ x = 2.5743 and excitation coefficient is $$\beta^{*} = 0.8285$$ β ∗ = 0.8285 , the optimal returns can be obtained between the enterprise organization and employees under this incentive constraint model. Before and after the implementation of the equity incentive constraint mechanism, there are significant differences in the performance level of enterprises. The implementation of the incentive constraint mechanism is beneficial to the improvement of enterprise performance level. The employee incentive constraint model constructed expands the application of CNN in the incentive mechanism and provides a direction for the development of enterprise performance.