An Improved InVEST Ecological Service Evaluation Model Based on BP Neural Network Optimization

The land is the material basis for human survival, and the contradiction between people and land has become increasingly prominent. The land ecological problem has gradually become a hot spot of concern. It is imperative to make a scientific evaluation of the land ecological quality and propose reasonable and feasible improvement measures and recommendations. At present, domestic research on environmental cost and environmental cost degradation mostly focuses on theoretical discussion, and there are few applications and practical research on enterprise environmental cost management. Based on the principle of protecting the ecological environment, this paper creates an ecological service assessment model to assess the real economic cost of land use development projects. From small community projects to large-scale national projects, because environmental costs are difficult to estimate, this paper uses the recovery cost method and the preventive expenditure method to quantify environmental costs. The cost of environmental degradation mainly comes from water pollution and air pollution. This paper uses the pollution function method to quantify the cost of environmental degradation. The InVEST model is used to evaluate the value of ecosystem services, and the BP neural network method is used to optimize the ecosystem service model, and the sensitivity analysis of the data is used to feedback the impact of the project on ecosystem services. The ecosystem service model based on neural network optimization makes the accuracy of data measurement results reaching 99.7%, which makes the model having a good generalization. Taking a paper mill as an example, this paper evaluates environmental costs by resource consumption cost, environmental degradation value and environmental governance cost, and estimates environmental degradation costs by major environmental governance costs. Finally, the environmental cost and environmental degradation cost are integrated, and the ecosystem service model is established. The neural network model was established in the Matlab environment based on the InVEST model, and the model simulation results of the ecosystem service system were obtained. Compared with the InVEST results, the results of this paper have better authenticity and market utilization value. Although a paper mill was used as an example, the system was evaluated and the evaluation results were analysed. Compared with the actual situation, there is a certain reliability. However, due to the limited data, the number of verifications is insufficient for the system. It is hoped that more data can be verified later to ensure its reliability.