Neural network and evidence theory-based water pollution event intelligent decision-making method

Abstract

The invention discloses a neural network and evidence theory-based water pollution event intelligent decision-making method. The method comprises the following steps of collecting a water body surface image of a to-be-detected water region, extracting image feature parameters from the image, and performing normalization on various image feature parameters; performing fuzzy reasoning based on the various image feature parameters to obtain preliminary judgment of a water pollution event type; according to the preliminary judgment of the water pollution event type, calling a corresponding water quality sensor to extract water quality feature parameters, and performing normalization on numerical values of the water quality feature parameters; and finally performing training by utilizing a neural network to obtain a nonlinear mapping relationship between the multiple feature parameters and a specific water pollution event, performing weighted processing operation on the previously established mapping relationship according to a D-S evidence theory, and finally performing prediction and decision-making on the water pollution type. According to the method, a target water region is effectively monitored in real time, so that stable and normal water quality is ensured; and the method has relatively high flexibility and relatively good self-adaptation capability.

Description

technical field [0001] The invention relates to an intelligent decision-making method for water pollution events, in particular to a water pollution event prediction and decision-making method combined with a neural network pattern recognition method and evidence theory, and belongs to the technical field of artificial intelligence. Background technique [0002] With the advancement of urbanization and the gradual development of the economy, the demand for water resources is also increasing, and the accompanying water pollution caused by industrial waste and domestic waste has become a growing phenomenon. issues worthy of attention. Especially in the case of residential water use, agricultural irrigation and breeding, and fine chemical industry water use, the quality of water quality is increasing, and it has become a new industry to realize the detection of water quality in water source areas and the timely prediction and decision-making of water pollution incidents. . ...

AI Technical Summary

Problems solved by technology

[0003] At present, there are mainly two kinds of water pollution detection and prediction schemes. The first one is to manually sample the water quality, and then conduct complicated chemical analysis and inspection on the samples, so as to obtain a detailed water quality report. Although this scheme can be obtained very accurately The detailed situation of water pollution, but it takes a long time to obtain real-time updated water quality data, and the inspection cost is high and the economic benefit is low; the second is to use a single water quality monitoring sensor to monitor the water pollution in the water area to be detected in real time , transmit the information obtained by the sensor to the central computer in real time, and then carry out detailed screening and prediction

This solution is low in cost and can realize real-time monitoring, but the water environment itself is a complex and changeable environment to be detected, and the monitoring of a single sensor often has weaknesses such as ambiguous information, poor fault tolerance, poor detection efficiency, and small monitoring range.

Images