TL;DR
This paper introduces a novel group decision-making method for selecting algal bloom remediation strategies by integrating real-time sensor data with fuzzy expert opinions. By employing a modified Vague set similarity measure and a grey relational degree approach, the method aims to overcome the subjectivity of traditional methods and provide more rational, data-driven decisions for environmental management.
What Problem Does This Paper Address?
Algal blooms, a consequence of water eutrophication, pose a significant threat to water security and public health globally. While various remediation approaches exist, selecting the most effective one for a specific situation is challenging. Existing methods often rely heavily on subjective expert opinions and lack the integration of real-time environmental data. This leads to potentially one-sided, inefficient, and resource-wasting decisions. The paper addresses this gap by proposing a method that fuses objective sensor data with subjective, yet structured, expert evaluations to improve the rationality and effectiveness of decision-making in algal bloom remediation.
Methods
The proposed method integrates two main components:
- Optimal Similarity Aggregation of Group Opinions: This involves aggregating opinions from multiple experts using a modified Vague set similarity measure. The Vague set is chosen for its ability to handle linguistic and grade variables, representing truth-membership and false-membership functions. The core of this aggregation is an optimization model that minimizes the inconsistency between the aggregated opinion and individual expert opinions. A novel similarity measurement for Vague values is introduced, which considers interval ends, kernel distance, and hesitancy degrees, aiming for improved precision and differentiation.
- Selection of Approaches Based on Sensor Data: After aggregating expert opinions into an approach-index matrix, this matrix is transformed into real numbers using a superiority function. This transformed matrix is then compared with real-time water quality sensor data (standardized from historical and current measurements). The comparison is done using the grey relational degree, which quantifies the similarity between the trend of the alternatives' abilities and the real-time environmental status. The approach with the highest relational degree is selected as the most suitable.
Datasets
The paper uses a dataset of water quality monitoring data from Yuyuantan Lake in Beijing. This dataset includes measurements for pH, total phosphorus (TP), total nitrogen (TN), chlorophyll a (chl_a), and dissolved oxygen (DO). Historical data (49 groups) and real-time data (the 50th group) were used for standardization and comparison.
Key Insights and Ideas
- Bridging Subjectivity and Objectivity: The core innovation lies in effectively combining subjective expert knowledge (expressed through fuzzy Vague sets) with objective, real-time sensor data. This fusion aims to create a more robust and reliable decision-making framework.
- Enhanced Vague Set Similarity: The development of a modified similarity measurement for Vague values, which accounts for hesitancy and interval properties, is crucial for improving the aggregation of expert opinions.
- Real-time Relevance: By directly incorporating real-time sensor data through the grey relational degree, the method ensures that the selected remediation approach is relevant to the current environmental conditions, moving beyond static evaluations.
- Group Wisdom Amplification: Aggregating opinions from multiple experts, rather than relying on a single individual, leverages collective knowledge and reduces the impact of individual biases or limited expertise.
Key Results
The study demonstrates the effectiveness of the proposed method through an example of algal bloom remediation in Yuyuantan Lake. The results indicate that the algaecide method was ranked highest, followed by electrochemistry and adsorption, while algophagous and microorganism methods were ranked lowest. This ranking was consistent with the analysis of real-time water quality data, where high levels of TP and TN favored methods like algaecide. The paper concludes that the selection model successfully identifies remediation approaches that best match the current water quality conditions.
Limitations and Future Work
- Limitations: The paper acknowledges that the method relies on the quality and availability of sensor data and expert opinions. The subjectivity inherent in expert evaluations, even when structured within Vague sets, remains a factor. The computational complexity of the optimization model for aggregating opinions might also be a consideration for very large groups of experts or complex Vague set operations.
- Future Work: The authors suggest that this method and other similar approaches can be further developed by integrating granular computing, fuzzy reasoning, artificial neural networks, and expert systems. Future research could also explore the application of this framework to other environmental management problems requiring complex decision-making under uncertainty.
Summary updated