RAPID SELECTING UAVs FOR COMBAT BASED ON THREE-WAY MULTIPLE ATTRIBUTE DECISION

Abstract

Unmanned aerial vehicles (UAVs) can carry out more and more dangerous missions and strike deep in the skies over hostile military sites. Thus, selecting appropriate UAVs to attend combat through rapid assessment is a hot topic in current research. In consideration of formulating practical evaluation as a three-way multiple attribute decision making (MADM) problem, a comprehensive assessment method based on interval-valued intuitionistic fuzzy set (IVIFs) is introduced under the context of determining the precision combat mission. First, the critical attributes of the UAV combat effectiveness are determined according to battlefield intelligence. Second, the attribute weights are computed by exploring the feature information of attribute orders given by experts. Third, the conditional probto collect more information to reach decision conclusions. It is a new research to combine a three-way decision algorithm and multiple attribute decision making (MADM) [13]…[17] in recent years. This hybrid method can consider both the MADM matrix and di erent loss functions for individual UAVs. In the application of decision, it is di cult for experts to give an accurate assessment with exact numbers due to the complexity of the battle. The de“nition of the intuitionistic fuzzy set (IFs) as an extension of the fuzzy number [18]…[20] was proposed by Atanassov, in which both membership and non-membership degrees were introduced. As a further extension of IFs, Atanassov and Gargov proposed the concept of interval-valued intuitionistic fuzzy set (IVIFs) [21]…[29]. It is clear that IVIFs enables experts to give preference judgments on UAV performance through interval-valued membership degrees to reduce errors. In this paper, the UAV evaluation method is given as a three-way MADM problem with IVIFs. First, a general attribute framework is constructed by discussing the in”uence factors on UAV performance and the battle information, and a method to determine the weight of attribute is given based on the superiority index of attribute. Second, IVIFs is used for the subjective judgment of the proposed method, and the conditional probability that the UAV can be selected is calculated based on MADM. Third, the classi“cation of UAVs is obtained combined with the given loss functions of individual UAVs. Finally, a numerical example further illustrates the e ectiveness and advantage of the proposed method. This paper is the “rst attempt to study the selection of UAVs based on the three-way MADM under the IVIFs environment. The other sections are set out as follows. Section 2 proposes the evaluation system of UAVs combat e ectiveness. In Section 3, we brie”y describe the proposed UAVs evaluation model based on a new three-way MADM method with IVIFs. A case study about the UAV selection in a battle shows the applicability and power of the introduced methodology in Section 4. Finally, concluding remarks and future directions are presented in Section 5. 2. Evaluation System of Unmanned Aerial Vehicles’ Combat Effectiveness Usually, speci“c battle“eld scenarios determine the survivability and environmental adaptability of UAVs. With the continuous innovation of science and technology, war also appears with di erent characteristics of combat, for example, the battle“eld space is more extensive, the operational command is more accurate, and the weapon killing speed is increased. All of these accurately provide intelligence for the battle“eld and a basis for the formulation of accurate strategic, tactical strategies and special missions in each battle. In air combat, it is key to the current precision operations and national military strategy research. The scienti“c evaluation index system is an essential prerequisite for combat e ectiveness evaluation. Whether the selection of evaluation system index is proper or not is directly related to the evaluation result. Therefore, we should perceive the battle“eld environment to determine UAV combat missions and select attributes of UAV combat e ectiveness to construct the decision framework. 2.1 Evaluation Attributes of Unmanned Aerial Vehicles’ Combat Effectiveness The evaluation system of UAVs combat e ectiveness should be established based on the basic elements of evaluation and the features of the object, which are the main factors in”uencing the information warfare capability. It strives to respond to UAVs combat capability to its best and fully follow the principles of purpose, uniqueness, comprehensiveness and personality. Then, the main factors a ecting the optimal selection of procedural UAVs are determined. Suppose that the assessed attribute set is {c1, c2, . . . , c6}. The various attributes are as follows: c1 is the reconnaissance target capability. It is the ability to explore the operation target, which is mainly decided by the performance of airborne detection equipment and airborne radar. The main parameters include detection range, search angle, resolution and the ability to discover and identify the targets and operate UAVs. c2 is the battle“eld ”exibility. This performance includes UAVs pitching agility, axial agility, high performance, conversion performance and other parameters. c3 is the attack capacity. It is the capability and quantity of the UAVs airborne equipment, mainly including the power range of the missile, the payload distance of the seeker, the angle of departure from the shaft and the e ective launch distance. c4 is the air survival ability. The parameters mainly include electronic countermeasures capability, navigation, radar re”ectance area and geometry size. c5 is the coordinated combat capability. It denotes the ability to coordinate operation and maintain uninterrupted communication among UAVs under a uni“ed organization and command. c6 is the logistics support capability. It is the maintenance capability of UAVs. 2.2 The Determination Method of Attribute Weight The role of the attribute weights is vital in MADM problems, which can be obtained by the superiority index of attribute. Let {e1, e2, . . . , et} be the set of experts, and the weight of expert ek be λk with t k=1 λk = 1. Now, we discuss the method to determine the weights of the attributes. First, the de“nition of superiority index is given. Definition 1. Suppose that the priority order of attributes is cki1 c k i2 • • • ckimgiven by ek, denote r k ij as: