A Weighted Combination Method for Conflicting Evidence in Multi-Sensor Data Fusion

Xiao, Fuyuan; Qin, Bowen

doi:10.3390/s18051487

Cited by 68 publications

(44 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As shown in Table 13, the method proposed in [30] had a prediction accuracy that was lower than that of the basic DS theory when the sliding window sizes were 120 and 960. When the sliding window sizes were 240 and 360, the prediction accuracies of the methods proposed in [31,32] were also lower than those of the basic DS evidence theory. Compared with the basic DS evidence theory results, the method proposed and employed in this study could be used to effectively guarantee the original prediction accuracy and achieve different degrees of improvement.…”

Section: Decision-level Fusion Fault Prediction Experiments Based On Pmentioning

confidence: 88%

“…It can be seen that the final fault prediction result is outer race centered@6:00 fault, which is the same as the real value. Based on the results of the PSO-ANN multifeature fusion fault prediction experiment in Section 6.3, Table 13 shows that multiple PSO-ANN models trained with different single features are combined with basic DS evidence theory, DS evidence theory and Deng entropy [30], DS evidence theory combined with evidence distance and Deng entropy [31], DS evidence theory combined with cosine similarity and Deng entropy [32], and the proposed method for fault prediction accuracy of decision-level fusion. As shown in Table 13, the method proposed in [30] had a prediction accuracy that was lower than that of the basic DS theory when the sliding window sizes were 120 and 960.…”

Section: Decision-level Fusion Fault Prediction Experiments Based On Pmentioning

confidence: 99%

“…Wang et al [31] proposed the preprocessing of original evidence using evidence distance and belief entropy. Xiao et al [32] proposed the preprocessing of original evidence with improved cosine similarity and belief entropy; although this improved strategy combined with the DS evidence theory can be used in some scenarios, it is still unable to effectively improve the decision-making ability of the original DS evidence theory for uncertain data.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Research on Mechanical Fault Prediction Method Based on Multifeature Fusion of Vibration Sensing Data

Huang

Liu

2019

Sensors

View full text Add to dashboard Cite

Vibration sensing data is an important resource for mechanical fault prediction, which is widely used in the industrial sector. Artificial neural networks (ANNs) are important tools for classifying vibration sensing data. However, their basic structures and hyperparameters must be manually adjusted, which results in the prediction accuracy easily falling into the local optimum. For data with high levels of uncertainty, it is difficult for an ANN to obtain correct prediction results. Therefore, we propose a multifeature fusion model based on Dempster-Shafer evidence theory combined with a particle swarm optimization algorithm and artificial neural network (PSO-ANN). The model first used the particle swarm optimization algorithm to optimize the structure and hyperparameters of the ANN, thereby improving its prediction accuracy. Then, the prediction error data of the multifeature fusion using a PSO-ANN is repredicted using multiple PSO-ANNs with different single feature training to obtain new prediction results. Finally, the Dempster-Shafer evidence theory was applied to the decision-level fusion of the new prediction results preprocessed with prediction accuracy and belief entropy, thus improving the model's ability to process uncertain data. The experimental results indicated that compared to the K-nearest neighbor method, support vector machine, and long short-term memory neural networks, the proposed model can effectively improve the accuracy of fault prediction.

show abstract

Section: Decision-level Fusion Fault Prediction Experiments Based On Pmentioning

confidence: 88%

Section: Decision-level Fusion Fault Prediction Experiments Based On Pmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Research on Mechanical Fault Prediction Method Based on Multifeature Fusion of Vibration Sensing Data

Huang

Liu

2019

Sensors

View full text Add to dashboard Cite

show abstract

“…In this step, the weights calculated in the previous step are adopted to modify its corresponding BPAs before using the Dempster's combination rule. For this, the following Equation (29) is used as it is proposed by Xiao and Qin 75

WAE false(c false) = \sum_{c = 1}^{k} ω^{c} * m_{c} .

…”

Section: The Proposed Methodsmentioning

confidence: 99%

Sensor fusion based on Dempster‐Shafer theory of evidence using a large scale group decision making approach

Köksalmiş

Kabak

2020

Int J Intell Syst

View full text Add to dashboard Cite

In group decision making (GDM), the quality of the solution relies primarily on the quality and the expertize of decision makers. At that point, deriving the weights, which reflects their importance or perceived reliability of decision makers, presents as a new challenge. In addition to that, the uncertainty is also a common problem for GDM. These problems are also faced in the sensor fusion problem where information from multiple sources must be aggregated.Therefore, in this study, a large scale GDM approach for sensor fusion is proposed. Since the proposed method is a clustering-based method, it provides acceptable results in the sensor networks consisting of multiple sensors. It can work under uncertainty as a result of converting the raw data obtained from sensors to the basic probability assignments. It also considers the reliability of the sensors clusters by assigning three objective weights. In addition to these objective weights, the proposed method enables to assign subjective weights to integrate supervisors/intelligence analyst experiences and knowledge in the problem field. The applicability and the validity of the proposed method are checked through two real classification data sets: ionosphere and forest type mapping data set. Experiments show that the classification rate is increased significantly when the proposed method is applied to two data sets. Finally, effect of extension parameter, objective weights, reliability threshold, number of clusters and clustering method on the classification rate and the detection probability are examined, and future studies are provided in conclusion.classification, Dempster-Shafer theory of evidence, large scale group decision making, sensor fusion, sensor weighting | INTRODUCTIONAs the complexity of the problem grows with the complexity of the socioeconomic environment, it gets more difficult for a single decision maker to cover all the relevant characteristic of a decision problem. Hence, in the real world, it is required to fuse the group of decision makers' opinion to obtain a feasible and satisfactory decision making result by using group decision making (GDM) approaches. 1,2 Here, the quality of the solution relies primarily on the quality and the expertize of decision makers in the problem area and how to aggregate decision makers' evaluations. 3 Therefore, it is critical to find a good way of aggregating decision makers' opinions whose expertize, background, and knowledge differ. In the literature, it is mostly assumed that the decision makers have the same importance level. 4 However, disregarding the weights of decision makers can cause improper and inaccurate outcomes that cannot be compensated for in the final solution. Therefore, when decision makers do not share similar levels of expertize, background, or knowledge, it is better to assign weights to the decision makers reflecting their importance or perceived reliability. At that point, deriving the weights of decision makers presents as a new challenge. Koksalmis and Kabak 5 presents the objectiv...

show abstract

“…Since Zadeh's paradox 15 was proposed, the means of overcoming the shortcomings of Dempster's rule and integrating conflicting sources of evidence effectively has become a crucial research field in DST. Quite a few novel or alternative combination rules have been developed to improve the performance of Dempster's rule, such as the conjunctive rule, [16][17][18] disjunctive rule, [19][20][21] cautious conjunctive rule, 22,23 and the bold disjunctive rule. [24][25][26][27][28][29][30] To correct the deviation of the fusion rules, many studies have been carried out, and there are two main popular ideas.…”

Section: Introductionmentioning

confidence: 99%

Evidence combination using OWA‐based soft likelihood functions

2019

View full text Add to dashboard Cite

Dempster's combination rule has been widely regarded and applied since it is an effective and rigorous method of synthesizing multisource information with its special information representation (ie, mass function or basic probability assignment). However, it has also been criticized and debated upon regarding some of its unreasonable behaviors and restrictive requirements, such as the counterintuitive results in some cases. To address these issues from different perspectives, in this study, an alternative fusion rule is developed under the framework of Dempster‐Shafer evidence theory. A novel evidence combination rule called CR‐SLF is proposed based on soft likelihood functions (SLF) considering the ordered weighted average aggregation operator. Some illustrative examples are shown, and the corresponding analyses demonstrate the good performance of CR‐SLF to fuse multisource evidence. To extend CR‐SLF further, the reliability of multisource evidence is considered from two aspects, subsequently two reliability‐based combination rules are presented, including the discount‐based rule and the SLF improvement‐based rule. The simulation results show that the reliability‐based CR‐SLF has a better fusion effect than the rule without considering the reliability.

show abstract

A Weighted Combination Method for Conflicting Evidence in Multi-Sensor Data Fusion

Cited by 68 publications

References 58 publications

Research on Mechanical Fault Prediction Method Based on Multifeature Fusion of Vibration Sensing Data

Research on Mechanical Fault Prediction Method Based on Multifeature Fusion of Vibration Sensing Data

Sensor fusion based on Dempster‐Shafer theory of evidence using a large scale group decision making approach

Evidence combination using OWA‐based soft likelihood functions

Contact Info

Product

Resources

About