CN109408228B - A Crowdsensing Task Allocation Method Based on Budget Allocation - Google Patents

A Crowdsensing Task Allocation Method Based on Budget Allocation Download PDF

Info

Publication number
CN109408228B
CN109408228B CN201811159885.4A CN201811159885A CN109408228B CN 109408228 B CN109408228 B CN 109408228B CN 201811159885 A CN201811159885 A CN 201811159885A CN 109408228 B CN109408228 B CN 109408228B
Authority
CN
China
Prior art keywords
task
user
budget
formula
turn
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201811159885.4A
Other languages
Chinese (zh)
Other versions
CN109408228A (en
Inventor
张立臣
闫斌
李丽霞
王阿娜
李黎
王小明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangzhou Aladdin Intelligent Technology Co ltd
Original Assignee
Shaanxi Normal University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shaanxi Normal University filed Critical Shaanxi Normal University
Priority to CN201811159885.4A priority Critical patent/CN109408228B/en
Publication of CN109408228A publication Critical patent/CN109408228A/en
Application granted granted Critical
Publication of CN109408228B publication Critical patent/CN109408228B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00Arrangements for program control, e.g. control units
    • G06F9/06Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46Multiprogramming arrangements
    • G06F9/50Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5005Allocation of resources, e.g. of the central processing unit [CPU] to service a request

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种基于预算调配的群智感知任务分配方法,收集任务请求者提交的感知任务集合Γ={τ1,...,τi,...,τn},n为一个有限的自然数,任务τi=(Bi,Qi),Bi是完成任务τi的总预算,为一个有限的正实数,Qi是任务τi的质量要求值,为一个有限的正整数;收集用户集合U={u1,...,uj,...,um}和该用户集合中用户uj的质量等级qj,j∈{1,...,m};计算用户uj完成任务τi的代价cij,并确定用户uj完成任务τi的收益wij;根据任务τi的质量需求Qi和用户uj的质量等级qj对感知任务集合Γ中的任务进行分配,根据任务分配结果,计算并支付用户报酬rj

A crowdsensing task assignment method based on budget allocation, collects a set of sensing tasks submitted by task requesters Γ={τ 1 ,...,τ i ,...,τ n }, where n is a finite natural number, Task τ i =(B i ,Q i ), Bi is the total budget for completing the task τ i , which is a finite positive real number, and Qi is the quality requirement value of the task τ i , which is a finite positive integer; Set U={u 1 ,...,u j ,..., um } and the quality level q j of user u j in the user set, j∈{1,...,m}; calculate user u The cost c ij of j completing the task τ i , and the benefit w ij of the user u j completing the task τ i is determined; according to the quality requirement Q i of the task τ i and the quality level q j of the user u j , the tasks in the perception task set Γ are determined. Assignment, according to the task assignment result, calculate and pay the user reward r j .

Description

基于预算调配的群智感知任务分配方法A Crowdsensing Task Allocation Method Based on Budget Allocation

技术领域technical field

本发明属于点数字数据处理技术领域,具体涉及到一种基于预算调配的群智感知任务分配方法。The invention belongs to the technical field of point digital data processing, and in particular relates to a method for assigning a group intelligence perception task based on budget allocation.

背景技术Background technique

近年来,移动智能设备(如智能手机、平板电脑、智能手表等)伴随着移动计算和无线通信技术的迅速发展得到广泛普及,其存储、计算、通信能力得到不断加强,并配备了众多传感器(如摄像头、麦克风、重力仪、GPS、陀螺仪、温度传感器、血压传感器等等)。在此基础上,人们利用大量普通用户使用的移动设备作为基本感知单元,通过物联网/移动互联网进行协作,实现感知任务分发与感知数据收集利用,最终完成大规模、复杂的城市与社会感知任务。为了充分、低成本地利用这些移动智能设备的感知、计算和存储资源,群智感知计算模式应运而生并得到广泛研究和应用。与基于传感网和物联网的感知方式不同,群智感知以大量普通用户作为感知源,强调利用大众的广泛分布性、灵活移动性和机会连接性进行感知,并为城市及社会管理提供智能辅助支持。它可应用在很多重要领域,如智能交通、公共安全、社会化推荐、环境监测、城市公共管理等诸多领域,因而得到了世界各国政府、产业界和学术界的高度关注。In recent years, mobile smart devices (such as smart phones, tablet computers, smart watches, etc.) have been widely popularized with the rapid development of mobile computing and wireless communication technologies. Such as camera, microphone, gravimeter, GPS, gyroscope, temperature sensor, blood pressure sensor, etc.). On this basis, people use the mobile devices used by a large number of ordinary users as the basic perception unit to cooperate through the Internet of Things/mobile Internet to realize the distribution of perception tasks and the collection and utilization of perception data, and finally complete large-scale and complex urban and social perception tasks. . In order to make full and low-cost use of the sensing, computing and storage resources of these mobile smart devices, the crowd-sensing computing model emerges and has been widely studied and applied. Different from the sensing methods based on sensor networks and the Internet of Things, crowd sensing uses a large number of ordinary users as the sensing source, emphasizes the use of the widespread distribution, flexible mobility and opportunity connectivity of the public for sensing, and provides intelligence for urban and social management. Auxiliary support. It can be applied in many important fields, such as intelligent transportation, public safety, social recommendation, environmental monitoring, urban public management and many other fields, so it has received great attention from governments, industries and academia around the world.

一个群智感知应用系统一般包含3个组成部分:云平台(即系统),任务请求者和任务完成者。任务请求者向云平台发布感知任务(下文简称任务)需求,并从云平台接收任务结果;云平台主要负责任务发布、任务分配、感知数据收集和任务质量评估;任务完成者,即携带移动智能设备的普通移动用户(下文简称用户),从云平台接收所分配的任务并负责数据感知和收集,并将感知数据发送给云平台。在群智感知系统中,任务分配是实施群智感知的基础。云平台需要根据任务请求者提交的任务需求,并结合当前在线用户的状态,将各个任务分配(即指派)给参与群智感知的众多用户,指派的依据是实现特定的优化目标,如最大化任务接收率、最大化整体收益等。A crowd-sensing application system generally includes three components: cloud platform (ie system), task requester and task completer. The task requester publishes the perception task (hereinafter referred to as the task) requirements to the cloud platform, and receives the task results from the cloud platform; the cloud platform is mainly responsible for task release, task allocation, perception data collection and task quality assessment; task completers, that is, carry mobile intelligence Ordinary mobile users of the device (hereinafter referred to as users) receive assigned tasks from the cloud platform and are responsible for data perception and collection, and send the perception data to the cloud platform. In the crowd-sensing system, task assignment is the basis for implementing crowd-sensing. The cloud platform needs to allocate (ie, assign) each task to many users participating in crowdsensing according to the task requirements submitted by the task requester and combined with the status of the current online users. The basis for the assignment is to achieve specific optimization goals, such as maximizing Task acceptance rate, maximizing overall revenue, etc.

从现有的研究成果上看,当前的面向群智感知的任务分配方法和系统尚未考虑以下几个方面:1.平台对任务预算的合理调度分配以提高任务接受率及系统收益;2.选取合适的质量等级的用户并通过用户的合作来满足任务的质量需求,以节省任务预算。在现有任务分配方法中,一般由平台收集当前需要执行的所有任务,并向所有用户发送需要执行的任务集合,然后根据各个用户自身的具体情况分配任务,一般假设一个任务的预算只用于本任务的分配。事实上,用户的质量等级有限,多个用户合作完成任务则可以获得更高质量的感知数据,因此,多个不同质量等级的用户可以通过合作来完成一个高质量需求的任务。此外,平台可以提高任务预算分配的合理程度,比如平台可以充分利用任务的结余预算,将其分配给部分预算不足的任务,在节省任务预算的同时有望大幅提高任务分配率。但是,当前的在线任务分配方法尚未考虑这些因素,在很大程度上限制了群智感知的应用范围。Judging from the existing research results, the current task allocation methods and systems for crowd awareness have not considered the following aspects: 1. The platform allocates the task budget reasonably to improve the task acceptance rate and system benefits; 2. Select Users of the appropriate quality level and through the cooperation of users to meet the quality requirements of the task, to save the task budget. In the existing task allocation method, the platform generally collects all tasks that need to be performed currently, sends the task set to be performed to all users, and then allocates tasks according to the specific conditions of each user. Generally, it is assumed that the budget of a task is only used for assignment of this task. In fact, the quality level of users is limited, and multiple users can obtain higher-quality perceptual data by cooperating to complete a task. Therefore, multiple users with different quality levels can cooperate to complete a task requiring high quality. In addition, the platform can improve the rationality of task budget allocation. For example, the platform can make full use of the remaining budget of tasks and allocate it to some tasks with insufficient budgets, which is expected to greatly improve the task allocation rate while saving task budgets. However, the current online task assignment methods have not considered these factors, which largely limits the application scope of crowd sensing.

因此,为适应大规模群智感知应用场景,同时提高平台的任务分配率,在考虑任务的质量需求和用户的质量等级之间的关系以及平台对任务预算的合理调配的情况下,设计一种基于预算调配的群智感知任务分配方法是十分必要的。Therefore, in order to adapt to large-scale crowd-sensing application scenarios and improve the task allocation rate of the platform, considering the relationship between the quality requirements of tasks and the quality level of users and the reasonable allocation of task budgets by the platform, a A crowd-sensing task assignment method based on budget allocation is very necessary.

发明内容SUMMARY OF THE INVENTION

本发明所要解决的技术问题在于克服现有群智感知任务分配方法的缺点,提供一种设计合理、任务分配率高、节省任务预算的基于预算调配的群智感知任务分配方法。The technical problem to be solved by the present invention is to overcome the shortcomings of the existing crowd-sensing sensing task assignment methods, and provide a crowd-sensing sensing task assignment method based on budget allocation with reasonable design, high task assignment rate and task budget saving.

解决上述技术问题所采用的技术方案是一种基于预算调配的群智感知任务分配方法,其特征在于,由以下步骤组成:The technical solution adopted to solve the above-mentioned technical problems is a method for allocating tasks for crowdsensing based on budget allocation, which is characterized in that it consists of the following steps:

(1)收集任务请求者提交的感知任务集合Γ={τ1,...,τi,...,τn},i∈{1,...,n},n为一个有限的自然数,任务τi=(Bi,Qi),Bi是完成任务τi的总预算,为一个有限的正实数,Qi是任务τi的质量要求值,为一个有限的正整数;(1) Collect the set of perception tasks submitted by the task requester Γ={τ 1 ,...,τ i ,...,τ n }, i∈{1,...,n}, where n is a finite A natural number, task τ i =(B i ,Q i ), Bi is the total budget for completing the task τ i , which is a finite positive real number, and Qi is the quality requirement value of the task τ i , which is a finite positive integer;

(2)收集用户集合U={u1,...,uj,...,um}和该用户集合中用户uj的质量等级qj,j∈{1,...,m},m为一个有限的自然数;(2) Collect the user set U={u 1 ,...,u j ,..., um } and the quality level q j of the user u j in the user set, j∈{1,...,m }, m is a finite natural number;

(3)计算用户uj完成任务τi的代价cij,并确定用户uj完成任务τi的收益wij(3) Calculate the cost c ij for the user u j to complete the task τ i , and determine the profit w ij for the user u j to complete the task τ i ;

(4)根据任务τi的质量需求Qi和预算Bi以及用户uj的质量等级qj对感知任务集合Γ中的任务进行分配,具体分配公式如下:(4) Allocate the tasks in the sensing task set Γ according to the quality requirement Qi and budget B i of the task τ i and the quality level q j of the user u j . The specific allocation formula is as follows:

式中,yi表示任务τi是否成功分配,yi=1表示任务τi成功分配,τi∈Γ,yi=0表示任务τi分配失败;xij表示是否将任务τi分配给用户uj,uj∈U,xij=1表示将任务τi分配给用户uj,xij=0表示不将任务τi分配给用户uj,δ表示参与任务τi的所有用户的报酬之和与任务总预算的差值,

Figure BDA0001819778780000041
Figure BDA0001819778780000042
表示平台以往任务分配后的结余预算,为一个正实数,η∈(0,1)表示结余预算的利用率;In the formula, y i indicates whether the task τ i is successfully assigned, y i =1 indicates that the task τ i is successfully assigned, τ i ∈Γ, y i =0 indicates that the task τ i assignment fails; x ij indicates whether the task τ i is assigned to the User u j , u j ∈ U, x ij =1 indicates that task τ i is assigned to user u j , x ij =0 indicates that task τ i is not assigned to user u j , δ indicates that all users participating in task τ i The difference between the sum of the rewards and the total budget of the task,
Figure BDA0001819778780000041
Figure BDA0001819778780000042
Represents the balance budget after the platform's previous task allocation, which is a positive real number, and η∈(0,1) represents the utilization rate of the balance budget;

(5)根据任务分配结果,计算并支付用户报酬rj(5) Calculate and pay the user reward r j according to the task assignment result;

作为一种优选的技术方案,所述的步骤(3)用户uj完成任务τi的代价cijAs a preferred technical solution, in the step (3) the cost c ij for the user u j to complete the task τ i ,

cij=f(Qi,qj) (2)c ij =f(Q i ,q j ) (2)

式中,qj为用户uj的质量等级,函数f(Qi,qj)是一个非负的随qj单调递增函数;所述的用户uj完成任务τi的收益wijIn the formula, q j is the quality level of the user u j , and the function f(Q i , q j ) is a non-negative function that increases monotonically with q j ; the user u j 's income w ij for completing the task τ i ,

wij=g(Qi,qj) (3)w ij =g(Q i ,q j ) (3)

函数g(Qi,qj)是一个非负的随qj单调递增函数;The function g(Q i , q j ) is a non-negative function that increases monotonically with q j ;

作为一种优选的技术方案,所述的步骤(4)中的公式(1)是一个整数规划问题,为了高效解决该问题按照下述方法进行分配:As a preferred technical solution, the formula (1) in the step (4) is an integer programming problem, and in order to efficiently solve the problem, it is allocated according to the following method:

S1:令集合Γ'为待分配的任务集合,初始时Γ'=Γ,对于任意任务τi,τi∈Γ',令Q′i表示任务τi剩余的质量需求值,B′i表示任务τi的剩余预算,

Figure BDA0001819778780000043
表示平台以往任务分配后的结余预算,初始时Q′i=Qi,B′i=Bi
Figure BDA0001819778780000044
令集合
Figure BDA0001819778780000045
为剩余的用户集合,集合U'为可选用的用户集合,初始时集合
Figure BDA0001819778780000047
表示能执行任务τi的用户集合,初始
Figure BDA0001819778780000048
S1: Let the set Γ' be the set of tasks to be assigned, initially Γ'=Γ, for any task τ i , τ i ∈Γ', let Q' i represent the remaining quality requirement value of the task τ i , and B' i represent the remaining budget of task τ i ,
Figure BDA0001819778780000043
Represents the balance budget of the platform after the assignment of tasks in the past. Initially, Q′ i =Q i , B′ i =B i ,
Figure BDA0001819778780000044
set of orders
Figure BDA0001819778780000045
is the remaining user set, and the set U' is the optional user set. gather
Figure BDA0001819778780000047
represents the set of users who can perform the task τ i , the initial
Figure BDA0001819778780000048

S2:判断集合是否成立,若成立,按照下式S2: Judgment set Whether it is established, if so, according to the following formula

Figure BDA00018197787800000410
Figure BDA00018197787800000410

找出一个特定任务τi*,特定任务τi*为在单位预算内质量需求最低的任务,初始特定任务τi*的质量需求值Q′i*=Qi*,特定任务τi*的剩余预算B′i*=Bi*,特定任务τi*可选用的用户集合

Figure BDA00018197787800000411
否则,停止;Find a specific task τ i* , the specific task τ i* is the task with the lowest quality requirement within the unit budget, the initial quality requirement value of the specific task τ i* is Q′ i* =Q i* , the specific task τ i* Remaining budget B′ i* =B i* , the set of users that can be selected for a specific task τ i*
Figure BDA00018197787800000411
Otherwise, stop;

S3:按照下式找出一个最接近特定任务τi*质量需求的用户uj*S3: Find a user u j* that is closest to the quality requirement of a specific task τ i* according to the following formula,

式中,ci*j表示用户uj执行所分配的特定任务τi*时付出的代价,wi*j表示用户uj执行所分配的特定任务τi*时所带来的纯收益,若能找到,则转向步骤S4;否则,在任务集合Γ中删除此任务,并令yi*=0,即特定任务τi*分配失败,再转向步骤S2;In the formula, c i*j represents the cost paid by user u j to perform the assigned specific task τ i* , w i*j represents the pure benefit brought by user u j when performing the assigned specific task τ i* , If it can be found, then turn to step S4; otherwise, delete this task in the task set Γ, and make y i* =0, that is, the assignment of the specific task τ i* fails, and then turn to step S2;

S4:判断用户uj*执行特定任务τi*的报酬与特定任务τi*预算的差值即下式是否成立,S4: Determine whether the difference between the reward of the user u j* performing the specific task τ i* and the budget of the specific task τ i* , that is, whether the following formula holds,

(ci*j*+wi*j*)-B′i*≤δ (6)(c i*j* +w i*j* )-B′ i* ≤δ (6)

式中ci*j*表示用户uj*执行所分配的特定任务τi*时付出的总代价,wi*j*表示用户uj*执行所分配特定任务τi*时带来的纯收益,若成立则转向步骤S5;否则,在可选用的用户集合U′中删除用户uj*,再转向步骤S3寻找新的用户;In the formula, c i*j* represents the total cost paid by user u j * to perform the assigned specific task τ i* , and w i*j* represents the pure cost incurred by user u j* when performing the assigned specific task τ i* . If the income is established, turn to step S5; otherwise, delete user u j* in the optional user set U', and then turn to step S3 to find a new user;

S5:判断平台是否能利用结余预算

Figure BDA0001819778780000052
来帮助特定任务τi*完成分配,即下式是否成立,S5: Determine whether the platform can use the balance budget
Figure BDA0001819778780000052
to help a specific task τ i* complete the assignment, that is, whether the following formula holds,

Figure BDA0001819778780000053
Figure BDA0001819778780000053

若成立,则把用户uj*加入满足任务τi*质量需求的用户集合

Figure BDA0001819778780000054
中,转向步骤S6;否则,应把用户uj*从可选用的用户集合U'中删除,再转向步骤S3寻找新的用户;If so, add user u j* to the set of users that meet the quality requirements of task τ i*
Figure BDA0001819778780000054
In, turn to step S6; Otherwise, user u j* should be deleted from the optional user set U', then turn to step S3 to find new users;

S6:将用户uj*从可选用的用户集合U'中删除,根据公式Q′i*=Qi*′-qj*和公式B′i*=B′i*-(ci*j*+wi*j*)更新Q′i*和B′i*的值,再判断任务τi*质量需求是否被满足,即公式Q′i*≤0是否成立,若成立,转向步骤S7;若不成立,转向步骤S3继续寻找用户;S6: delete the user u j* from the selectable user set U', according to the formula Q' i* =Q i* '-q j* and the formula B' i* =B' i* -(c i*j * +w i*j* ) update the values of Q′ i* and B′ i* , and then judge whether the quality requirements of the task τ i* are satisfied, that is, whether the formula Q′ i* ≤ 0 is established, if so, turn to step S7 ; If not established, turn to step S3 and continue to look for the user;

S7:用户集合

Figure BDA0001819778780000055
即为执行任务τi*的一个可行的用户集合,令yi*=1,即特定任务τi*成功分配,令xi*j*=1,即将任务τi分配给用户uj*计算结余预算令Γ'=Γ'-{τi*},
Figure BDA0001819778780000058
再转向步骤S2下一任务进行分配。S7: User Collection
Figure BDA0001819778780000055
is a feasible set of users to perform task τ i* , let y i* =1, that is, a specific task τ i* is successfully assigned, let x i*j* =1, that is, assign task τ i to user u j* , Calculate balance budget Let Γ'=Γ'-{τ i* },
Figure BDA0001819778780000058
Then turn to step S2 to assign the next task.

作为一种优选的技术方案,所述的步骤(5)中对于执行特定任务τi*的用户uj*获得的报酬rj*As a preferred technical solution, the remuneration r j * obtained by the user u j* performing the specific task τ i* in the step (5),

Figure BDA0001819778780000061
Figure BDA0001819778780000061

当用户uj*完成特定任务τi*并获得报酬后,将所有xi*j*的值设为0,τi*∈Γ,并将用户uj*加入用户集合U中。When user u j* completes a specific task τ i* and gets paid, set all the values of x i*j* to 0, τ i* ∈ Γ, and add user u j* to the user set U.

作为一种优选的技术方案,所述的步骤(1)中当收集到任务请求者所提交的新的感知任务时,将新任务增加到集合Γ中;当一个已分配的任务被完成,则将其从集合Γ中去除。As a preferred technical solution, in the step (1), when a new perception task submitted by the task requester is collected, the new task is added to the set Γ; when an assigned task is completed, then remove it from the set Γ.

作为一种优选的技术方案,所述的步骤(2)中当检测到用户离开系统时,将该用户从集合U中删除;当收集到新的用户时,将该用户增加到集合U中。As a preferred technical solution, in the step (2), when it is detected that the user leaves the system, the user is deleted from the set U; when a new user is collected, the user is added to the set U.

本发明的有益效果如下:The beneficial effects of the present invention are as follows:

1、本发明考虑了任务的质量需求和用户质量等级之间的关系,由于多个用户合作完成任务可以获得更高质量的感知数据,因此在任务预算的约束条件下,根据任务的质量需求值来选取不同质量等级的用户协作完成感知任务,在满足感知数据质量的同时有望大幅节省任务预算。1. The present invention considers the relationship between the quality requirements of the task and the user's quality level. Since multiple users cooperate to complete the task, higher quality perception data can be obtained. Therefore, under the constraint of the task budget, according to the quality requirement value of the task To select users of different quality levels to cooperate to complete the perception task, it is expected to greatly save the task budget while satisfying the quality of the perception data.

2、本发明考虑了任务的预算对任务分配的影响,提出了一种能够提高任务分配率的方法,任务的预算将被平台拆分为两部分,一部分预算用来保证本任务的正常分配,剩余的部分预算将由平台进行合理分配给那些因为预算不足而无法分配的任务。分配过程中所有预算不足的任务都平等获得平台预算补充,体现了公平分配的思想。这种任务分配方法既促使任务的预算得到更充分地应用,也能使更多的任务被成功分配。2. The present invention considers the impact of the task budget on task allocation, and proposes a method that can improve the task allocation rate. The task budget will be split into two parts by the platform, and a part of the budget is used to ensure the normal distribution of the task, The remaining part of the budget will be reasonably allocated by the platform to those tasks that cannot be allocated due to insufficient budget. During the allocation process, all tasks with insufficient budget are equally supplemented by the platform budget, which reflects the idea of fair distribution. This task allocation method not only promotes the full application of the task budget, but also enables more tasks to be successfully allocated.

附图说明Description of drawings

图1是本发明一种基于预算调配的群智感知任务分配方法的流程图。FIG. 1 is a flow chart of a method for assigning tasks for crowdsensing based on budget allocation according to the present invention.

具体实施方式Detailed ways

下面结合附图和实施例对本发明进一步详细说明,但本发明不限于下述的实施方式。The present invention will be further described in detail below with reference to the accompanying drawings and examples, but the present invention is not limited to the following embodiments.

在图1中,本实施例的基于预算调配的群智感知任务分配方法由以下步骤组成:In FIG. 1 , the budget allocation-based crowdsensing task allocation method of this embodiment consists of the following steps:

(1)任务请求者提交的感知任务集合设为Γ={τ123},任务τi=(Bi,Qi),其中i∈{1,2,3,4},τ1=(4,75),τ2=(8,240),τ3=(5,110),τ4=(6,130);(1) The set of perception tasks submitted by the task requester is set to Γ={τ 123 }, task τ i =(B i ,Q i ), where i∈{1,2,3,4} , τ 1 =(4,75), τ 2 =(8,240), τ 3 =(5,110), τ 4 =(6,130);

(2)收集用户集合U={u1,u2,u3,u4,u5,u6},qj是用户uj的质量等级,其中j∈{1,2,3,4},q1=3,q2=5,q3=3,q4=2,q5=3,q6=2;(2) Collect user set U={u 1 , u 2 , u 3 , u 4 , u 5 , u 6 }, q j is the quality level of user u j , where j∈{1,2,3,4} , q 1 =3, q 2 =5, q 3 =3, q 4 =2, q 5 =3, q 6 =2;

(3)计算用户uj完成任务τi的代价cij,并确定用户uj完成任务τi的收益wij(3) Calculate the cost c ij for the user u j to complete the task τ i , and determine the profit w ij for the user u j to complete the task τ i ;

用户uj完成任务τi的代价cijThe cost c ij for the user u j to complete the task τ i ,

cij=f(Qi,qj)=5Qi+10qj (1)c ij =f(Q i ,q j )=5Q i +10q j (1)

用户uj完成任务τi的收益wijThe income w ij of user u j for completing task τ i ,

wij=g(Qi,qj)=Qi+2qj (2)w ij =g(Q i ,q j )=Q i +2q j (2)

根据上式(1)和(2)计算出用户u1,u2,u3,u4,u5,u6执行任务τ1234的代价和收益分别用矩阵表示:According to the above formulas (1) and (2), the costs and benefits of users u 1 , u 2 , u 3 , u 4 , u 5 , and u 6 to perform tasks τ 1 , τ 2 , τ 3 , τ 4 are calculated by matrices, respectively. express:

Figure BDA0001819778780000071
Figure BDA0001819778780000071

(4)根据任务τi的质量需求Qi和用户uj的质量等级qj对感知任务集合Γ中的任务进行分配,具体分配公式如下:(4) Allocate the tasks in the sensing task set Γ according to the quality requirement Qi of the task τ i and the quality level q j of the user u j . The specific allocation formula is as follows:

Figure BDA0001819778780000081
Figure BDA0001819778780000081

式中,yi表示任务τi是否成功分配,yi=1表示任务τi成功分配,τi∈Γ,yi=0表示任务τi分配失败;xij表示是否将任务τi分配给用户uj,uj∈U,xij=1表示将任务τi分配给用户uj,xij=0表示不将任务τi分配给用户uj,δ表示参与任务τi的所有用户的报酬之和与任务总预算的差值,

Figure BDA0001819778780000082
Figure BDA0001819778780000083
表示平台以往任务分配后的结余预算,为一个正实数,η∈(0,1)表示结余预算的利用率;In the formula, y i indicates whether the task τ i is successfully assigned, y i =1 indicates that the task τ i is successfully assigned, τ i ∈Γ, y i =0 indicates that the task τ i assignment fails; x ij indicates whether the task τ i is assigned to the User u j , u j ∈ U, x ij =1 indicates that task τ i is assigned to user u j , x ij =0 indicates that task τ i is not assigned to user u j , δ indicates that all users participating in task τ i The difference between the sum of the rewards and the total budget of the task,
Figure BDA0001819778780000082
Figure BDA0001819778780000083
Represents the balance budget after the platform's previous task allocation, which is a positive real number, and η∈(0,1) represents the utilization rate of the balance budget;

上述公式(3)是一个整数规划问题,为了高效解决该问题按照下述方法进行分配:The above formula (3) is an integer programming problem, and in order to solve the problem efficiently, it is allocated according to the following method:

S1:令集合Γ'为待分配的任务集合,初始时Γ'=Γ,对于任意任务τiS1: Let the set Γ' be the set of tasks to be assigned, initially Γ'=Γ, for any task τ i ,

τi∈Γ',令Q′i表示任务τi剩余的质量需求值,B′i表示任务τi的剩余预算,

Figure BDA0001819778780000084
表示平台以往任务分配后的结余预算,初始时Q′i=Qi,B′i=Bi令集合
Figure BDA0001819778780000086
为剩余的用户集合,集合U′为可选用的用户集合,初始时
Figure BDA0001819778780000087
集合
Figure BDA0001819778780000088
表示能执行任务τi的用户集合,初始
Figure BDA0001819778780000089
τ i ∈Γ', let Q′ i denote the remaining quality requirement value of task τ i , and B′ i denote the remaining budget of task τ i ,
Figure BDA0001819778780000084
Represents the balance budget of the platform after the assignment of tasks in the past. Initially, Q′ i =Q i , B′ i =B i , set of orders
Figure BDA0001819778780000086
is the remaining user set, and the set U' is the optional user set.
Figure BDA0001819778780000087
gather
Figure BDA0001819778780000088
represents the set of users who can perform the task τ i , the initial
Figure BDA0001819778780000089

S2:判断集合

Figure BDA00018197787800000810
是否成立,若成立,按照下式S2: Judgment set
Figure BDA00018197787800000810
Whether it is established, if so, according to the following formula

Figure BDA0001819778780000091
Figure BDA0001819778780000091

找出一个特定任务τi*,特定任务τi*为在单位预算内质量需求最低的任务,初始特定任务τi*的质量需求值Q′i*=Qi*,特定任务τi*的剩余预算B′i*=Bi*,特定任务τi*可选用的用户集合

Figure BDA0001819778780000092
否则,停止;Find a specific task τ i* , the specific task τ i* is the task with the lowest quality requirement within the unit budget, the initial quality requirement value of the specific task τ i* is Q′ i* =Q i* , the specific task τ i* Remaining budget B′ i* =B i* , the set of users that can be selected for a specific task τ i*
Figure BDA0001819778780000092
Otherwise, stop;

集合按照公式(4)找出一个在单位预算内质量需求最低的任务τi*,求得τi*=τ2,初始特定任务τ2的质量需求值Q2′=Q2=8,特定任务τ2的剩余预算B2′=B2=240,特定任务τ2可选用的用户集合

Figure BDA0001819778780000094
gather According to formula (4), find out a task τ i* with the lowest quality requirement within the unit budget, obtain τ i*2 , the initial quality requirement value of the specific task τ 2 is Q 2 ′=Q 2 =8, the specific task The remaining budget of τ 2 B 2 ′=B2=240, the set of users that can be selected for a specific task τ 2
Figure BDA0001819778780000094

S3:按照下式找出一个最接近特定任务τi*质量需求的用户uj*S3: Find a user u j* that is closest to the quality requirement of a specific task τ i* according to the following formula,

式中,ci*j表示用户uj执行所分配的特定任务τi*时付出的代价,wi*j表示用户uj执行所分配的特定任务τi*时所带来的纯收益,若能找到,则转向步骤S4;否则,在任务集合Γ中删除此任务,并令yi*=0,即特定任务τi*分配失败,再转向步骤S2;In the formula, c i*j represents the cost paid by user u j to perform the assigned specific task τ i* , w i*j represents the pure benefit brought by user u j when performing the assigned specific task τ i* , If it can be found, then turn to step S4; otherwise, delete this task in the task set Γ, and make y i* =0, that is, the assignment of the specific task τ i* fails, and then turn to step S2;

由步骤S2得出τi*=τ2,按照公式(5)找出uj*=u2,转向步骤S4;Obtain τ i*2 from step S2, find out u j* =u 2 according to formula (5), and turn to step S4;

S4:判断用户uj*执行特定任务τi*的报酬与特定任务τi*预算的差值即下式是否成立,S4: Determine whether the difference between the reward of the user u j* performing the specific task τ i* and the budget of the specific task τ i* , that is, whether the following formula holds,

(ci*j*+wi*j*)-B′i*≤δ (6)(c i*j* +w i*j* )-B′ i* ≤δ (6)

式中ci*j*表示用户uj*执行所分配的特定任务τi*时付出的总代价,wi*j*表示用户uj*执行所分配特定任务τi*时带来的纯收益,若成立则转向步骤S5;否则,在可选用的用户集合U'中删除用户uj*,再转向步骤S3寻找新的用户;In the formula, c i*j* represents the total cost paid by user u j * to perform the assigned specific task τ i* , and w i*j* represents the pure cost incurred by user u j* when performing the assigned specific task τ i* . If it is established, turn to step S5; otherwise, delete user u j* in the optional user set U', and then turn to step S3 to find a new user;

用户u2执行所分配的特定任务τ2时付出的总代价c22,用户u2执行所分配特定任务τ2时带来的纯收益w22,由于c22=90,w22=18,B′2=240,δ=25,则(c22+w22)-B′2≤δ成立,转向步骤S5;The total cost c 22 paid by the user u 2 when performing the assigned specific task τ 2 , the net benefit w 22 brought by the user u 2 when performing the assigned specific task τ 2 , since c 22 =90, w 22 =18, B ' 2 =240, δ = 25, then (c 22 +w 22 )-B' 2 ≤δ is established, and turn to step S5;

S5:判断平台是否能利用结余预算来帮助特定任务τi*完成分配,即下式是否成立,S5: Determine whether the platform can use the balance budget to help a specific task τ i* complete the assignment, that is, whether the following formula holds,

若成立,则把用户uj*加入满足任务τi*质量需求的用户集合

Figure BDA0001819778780000103
中,转向步骤S6;否则,应把用户uj*从可选用的用户集合U'中删除,再转向步骤S3寻找新的用户;If so, add user u j* to the set of users that meet the quality requirements of task τ i*
Figure BDA0001819778780000103
In, turn to step S6; Otherwise, user u j* should be deleted from the optional user set U', then turn to step S3 to find new users;

对于用户u2,令η=0.8,由于

Figure BDA0001819778780000104
c22+w22=90+18=108,则公式(7)成立,将用户u2加入集合
Figure BDA0001819778780000105
中,转向步骤S6;For user u 2 , let n=0.8, since
Figure BDA0001819778780000104
c 22 +w 22 =90+18=108, then formula (7) is established, and user u 2 is added to the set
Figure BDA0001819778780000105
, turn to step S6;

S6:将用户uj*从可选用的用户集合U'中删除,根据公式Q′i*=Qi*′-qj*和公式B′i*=B′i*-(ci*j*+wi*j*)更新Q′i*和B′i*的值,再判断任务τi*质量需求是否被满足,即公式Q′i*≤0是否成立,若成立,转向步骤S7;若不成立,转向步骤S3继续寻找用户;S6: delete the user u j* from the selectable user set U', according to the formula Q' i* =Q i* '-q j* and the formula B' i* =B' i* -(c i*j * +w i*j* ) update the values of Q′ i* and B′ i* , and then judge whether the quality requirements of the task τ i* are satisfied, that is, whether the formula Q′ i* ≤ 0 is established, if so, turn to step S7 ; If not established, turn to step S3 and continue to look for the user;

在用户集合U'中删除用户u2,更新Qi'和Bi'的值,Q2'=Q2'-q2=3,B2'=B2'-(c22+w22)=240-(90+18)=132,由于Q2'=3,则公式Q2'≤0不成立,转向步骤S3;Delete user u 2 in user set U', update the values of Qi ' and B i ' , Q 2 '=Q 2 '-q 2 =3, B 2 '=B 2 '-(c 22 +w 22 ) =240-(90+18)=132, since Q 2 '=3, the formula Q 2 '≤0 does not hold, and turn to step S3;

在步骤S3中,

Figure BDA0001819778780000106
转向步骤S4;In step S3,
Figure BDA0001819778780000106
Go to step S4;

在步骤S4中,对于用户u1,由于(c21+w21)-B2'=(70+14)-132=-48,δ=25,则公式(6)成立,转向步骤S5;In step S4, for user u 1 , since (c 21 +w 21 )-B 2 '=(70+14)-132=-48, δ=25, then formula (6) is established, and turn to step S5;

在步骤S5中,对于用户u1,判断公式(7)是否成立,由于

Figure BDA0001819778780000107
c21+w21=70+14=84,则公式(7)成立,将用户u1加入集合
Figure BDA0001819778780000108
中,转向步骤S6;In step S5, for user u 1 , it is judged whether formula (7) holds, because
Figure BDA0001819778780000107
c 21 +w 21 =70+14=84, then formula (7) is established, user u 1 is added to the set
Figure BDA0001819778780000108
, turn to step S6;

在步骤S6中,在用户集合U'中删除用户u1,更新Qi'和Bi'的值,Q2'=Q2'-q1=0,B2'=B2'-(c21+w21)=132-(70+14)=48,由于Q2'=0,则公式Q2'≤0成立,转向步骤S7;In step S6, delete the user u 1 in the user set U', update the values of Q i ' and B i ', Q 2 '=Q 2 '-q 1 =0, B 2 '=B 2 '-(c 21 +w 21 )=132-(70+14)=48, since Q 2 '=0, the formula Q 2 '≤0 is established, and turn to step S7;

S7:用户集合

Figure BDA00018197787800001111
即为执行任务τi*的一个可行的用户集合,令yi*=1,即特定任务τi*成功分配,令xi*j*=1,即将任务τi分配给用户uj*
Figure BDA0001819778780000111
计算结余预算
Figure BDA0001819778780000112
令Γ'=Γ'-{τi*},
Figure BDA0001819778780000113
再转向步骤S2下一任务进行分配;S7: User Collection
Figure BDA00018197787800001111
is a feasible set of users to perform task τ i* , let y i* =1, that is, a specific task τ i* is successfully assigned, let x i*j* =1, that is, assign task τ i to user u j* ,
Figure BDA0001819778780000111
Calculate balance budget
Figure BDA0001819778780000112
Let Γ'=Γ'-{τ i* },
Figure BDA0001819778780000113
Then turn to step S2 to assign the next task;

用户集合U2 +即为执行任务τ2的一个可行的用户集合,令y2=1,对任一用户

Figure BDA0001819778780000114
都令x2j=1,计算结余预算
Figure BDA0001819778780000115
令Γ'=Γ'-{τ2},
Figure BDA0001819778780000116
转向步骤S402为下一任务进行分配;User set U 2 + is a feasible set of users to perform task τ 2 , let y 2 =1, for any user
Figure BDA0001819778780000114
Let x 2j = 1, calculate the balance budget
Figure BDA0001819778780000115
Let Γ'=Γ'-{τ 2 },
Figure BDA0001819778780000116
Turn to step S402 to assign the next task;

需要指出的是,任务τ2是在预算充足的条件下找到了满足质量要求的用户集合U2 +顺利完成分配的;It should be pointed out that the task τ 2 is successfully assigned when the user set U 2 + that meets the quality requirements is found under the condition of sufficient budget;

在步骤S2中,集合

Figure BDA0001819778780000117
按照公式(4)找出一个在单位预算内质量需求最低的任务τi*,求得τi*=τ3,初始特定任务τ3的质量需求值Q3'=Q3=5,特定任务τ3的剩余预算B3'=B3=110,特定任务τ3可选用的用户集合
Figure BDA0001819778780000118
In step S2, the collection
Figure BDA0001819778780000117
According to formula (4), find out a task τ i* with the lowest quality requirement within the unit budget, obtain τ i*3 , the quality requirement value of the initial specific task τ 3 Q 3 '=Q 3 =5, the specific task The remaining budget of τ 3 B 3 ′=B 3 =110, the set of users that can be selected for a specific task τ 3
Figure BDA0001819778780000118

在步骤S3中,按照公式(5)找出uj*=u3,转向步骤S4;In step S3, find out u j* =u 3 according to formula (5), and turn to step S4;

在步骤S4中,用户u3执行所分配的特定任务τ3时付出的总代价c33,用户u3执行所分配特定任务τ3时带来的纯收益w33,由于c33=55,w33=11,B′3=110,δ=25,则(c33+w33)-B′3≤δ成立,转向步骤S5;In step S4, the total cost c 33 paid by the user u 3 when performing the assigned specific task τ 3 , and the net benefit w 33 brought by the user u 3 when performing the assigned specific task τ 3 , since c 33 =55, w 33 =11, B' 3 =110, δ=25, then (c 33 +w 33 )-B' 3 ≤δ is established, and turn to step S5;

在步骤S5中,对于用户u3,由于

Figure BDA0001819778780000119
c33+w33=55+11=66,则公式(7)成立,将用户u3加入集合中,转向步骤S6;In step S5, for user u 3 , since
Figure BDA0001819778780000119
c 33 +w 33 =55+11=66, then formula (7) is established, and user u 3 is added to the set , turn to step S6;

在步骤S6中,在用户集合U'中删除用户u3,再更新Qi'和Bi'的值,Q3'=Q3'-q3=5-3=2,B3'=B3'-(c33+w33)=110-66=44,由于Q3'=2,则公式Q3'≤0不成立,转向步骤S3;In step S6, delete the user u 3 in the user set U', and then update the values of Q i ' and B i ', Q 3 '=Q 3 '-q 3 =5-3=2, B 3 '=B 3 '-(c 33 +w 33 )=110-66=44, since Q 3 '=2, the formula Q 3 '≤0 does not hold, and turn to step S3;

在步骤S3中,uj*=u4,转向步骤S4;In step S3, u j* =u 4 , turn to step S4;

在步骤S4中,对于用户u4,由于c34=45,w34=9,B′3=44,δ=25,则(c34+w34)-B′3≤δ成立,转向步骤S5;In step S4, for user u 4 , since c 34 =45, w 34 =9, B′ 3 =44, and δ=25, then (c 34 +w 34 )-B′ 3 ≤δ is established, and turn to step S5 ;

在步骤S5中,对于用户u4,判断公式(7)是否成立,由于c34+w34=45+9=54,则公式(7)成立,将用户u4加入集合

Figure BDA0001819778780000122
中,转向步骤S6;In step S5, for user u 4 , it is judged whether formula (7) holds, because c 34 +w 34 =45+9=54, then formula (7) is established, user u 4 is added to the set
Figure BDA0001819778780000122
, turn to step S6;

在步骤S6中,在用户集合U'中删除用户u4,再更新Qi'和Bi'的值,Q3'=Q3'-q4=2-2=0,B3'=B3'-(c34+w34)=44-54=-10,由于Q3'=0,则公式Q3'≤0成立,转向步骤S7;In step S6, delete the user u 4 in the user set U', and then update the values of Q i ' and B i ', Q 3 '=Q 3 '-q 4 =2-2=0, B 3 '=B 3 '-(c 34 +w 34 )=44-54=-10, since Q 3 '=0, the formula Q 3 '≤0 is established, and turn to step S7;

在步骤S7中,用户集合U3 +即为执行任务τ3的一个可行的用户集合,令y3=1,对任一用户

Figure BDA0001819778780000123
都令x3j=1,计算结余预算
Figure BDA0001819778780000124
令Γ'=Γ'-{τ3},
Figure BDA0001819778780000125
转向步骤S2为下一任务进行分配;In step S7, the user set U 3 + is a feasible set of users to perform task τ 3 , let y 3 =1, for any user
Figure BDA0001819778780000123
Let x 3j = 1, calculate the balance budget
Figure BDA0001819778780000124
Let Γ'=Γ'-{τ 3 },
Figure BDA0001819778780000125
Turn to step S2 to allocate the next task;

任务τ3虽然预算不足,但由于平台利用以往任务结余的预算用于弥补任务预算,任务τ3得以找到满足质量要求的用户集合U3 +顺利完成分配;Although the budget of task τ 3 is insufficient, because the platform uses the budget from previous tasks to make up for the task budget, task τ 3 can find the user set U 3 + that meets the quality requirements and successfully complete the assignment;

在步骤S2中,集合

Figure BDA0001819778780000126
按照公式(4)找出一个在单位预算内质量需求最低的任务τi*,求得τi*=τ4,初始特定任务τ4的质量需求值Q4'=Q4=6,特定任务τ4的剩余预算B4'=B4=130,特定任务τ4可选用的用户集合
Figure BDA0001819778780000127
In step S2, the collection
Figure BDA0001819778780000126
According to formula (4), find out a task τ i* with the lowest quality requirement within the unit budget, obtain τ i*4 , the quality requirement value of the initial specific task τ 4 Q 4 '=Q 4 =6, the specific task The remaining budget of τ 4 B 4 ′=B 4 =130, the set of users that can be selected for a specific task τ 4
Figure BDA0001819778780000127

在步骤S3中,按照公式(5)找出uj*=u5,转向步骤S4;In step S3, find out u j* =u 5 according to formula (5), and turn to step S4;

在步骤S4中,用户u5执行所分配的特定任务τ5时付出的总代价c45,用户u5执行所分配特定任务τ5时带来的纯收益w45,c45=60,w45=12,B′4=130,δ=25,则(c45+w45)-B′4≤δ成立,转向步骤S5;In step S4, the total cost c 45 paid by the user u 5 to perform the assigned specific task τ 5 , and the net benefit w 45 brought by the user u 5 performing the assigned specific task τ 5 , c 45 =60, w 45 =12, B' 4 =130, δ=25, then (c 45 +w 45 )-B' 4 ≤δ is established, and turn to step S5;

在步骤S5中,对于用户u5,由于c45+w45=60+12=72,则公式(7)成立,将用户u5加入集合中,转向步骤S6;In step S5, for user u 5 , since c 45 +w 45 =60+12=72, then formula (7) is established, and user u 5 is added to the set , turn to step S6;

在步骤S6中,用户集合U'中删除用户u5,再更新Qi'和Bi'的值,Q4'=Q4'-q5=6-3=3,B4'=B4'-(c45+w45)=130-72=58,由于Q4'=3,则公式Q4'≤0不成立,转向步骤S3;In step S6, delete the user u 5 from the user set U', and then update the values of Qi ' and B i ' , Q 4 '=Q 4 '-q 5 =6-3=3, B 4 '=B 4 '-(c 45 +w 45 )=130-72=58, since Q 4 '=3, the formula Q 4 '≤0 does not hold, and turn to step S3;

在步骤S3中,uj*=u6,转向步骤S4;In step S3, u j* =u 6 , turn to step S4;

在步骤S4中,对于用户u6,由于c46=50,w46=10,B′4=58,δ=25,则(c46+w46)-B′4≤δ成立,转向步骤S5;In step S4, for the user u 6 , since c 46 =50, w 46 =10, B′ 4 =58, δ=25, then (c 46 +w 46 )-B′ 4 ≤δ is established, and turn to step S5 ;

在步骤S5中,对于用户u4,判断公式(7)是否成立,由于

Figure BDA0001819778780000131
c46+w46=50+10=60,则公式(7)成立,将用户u6加入集合
Figure BDA0001819778780000135
中,转向步骤S6;In step S5, for user u 4 , it is judged whether formula (7) holds, because
Figure BDA0001819778780000131
c 46 +w 46 =50+10=60, then formula (7) is established, and user u 6 is added to the set
Figure BDA0001819778780000135
, turn to step S6;

在步骤S6中,从用户集合U'中删除用户u6,再更新Qi'和Bi'的值,Q4'=Q4'-q6=3-2=1,B4'=B4'-(c46+w46)=68-60=8,由于Q4'=1,则公式Q4'≤0不成立,转向步骤S3;In step S6, delete the user u 6 from the user set U', and then update the values of Q i ' and B i ', Q 4 '=Q 4 '-q 6 =3-2=1, B 4 '=B 4 '-(c 46 +w 46 )=68-60=8, since Q 4 '=1, the formula Q 4 '≤0 does not hold, and turn to step S3;

在步骤S3中,由于找不到新的用户可以执行任务τ4,因此,任务τ4无法分配,因此令y4=0,并在任务集合Γ中删除此任务,再转向步骤S2;In step S3, since no new user can perform the task τ 4 , the task τ 4 cannot be assigned, so let y 4 =0, delete this task in the task set Γ, and turn to step S2;

任务τ4由于找不到满足任务质量需求的用户集合,因此无法完成分配;Task τ 4 cannot complete the assignment because it cannot find a user set that meets the task quality requirements;

在步骤S2中,集合

Figure BDA0001819778780000132
按照公式(4)找出一个在单位预算内质量需求最低的任务τi*,求得τi*=τ1,初始特定任务τ1的质量需求值Q1'=Q1=4,特定任务τ1的剩余预算B1'=B1=75,特定任务τ1可选用的用户集合
Figure BDA0001819778780000133
In step S2, the collection
Figure BDA0001819778780000132
According to formula (4), find out a task τ i* with the lowest quality requirement within the unit budget, obtain τ i*1 , the quality requirement value of the initial specific task τ 1 is Q 1 '=Q 1 =4, the specific task The remaining budget of τ 1 B 1 '=B 1 =75, the set of users that can be selected for a specific task τ 1
Figure BDA0001819778780000133

在步骤S3中,按照公式(5)找出uj*=u5,转向步骤S4;In step S3, find out u j* =u 5 according to formula (5), and turn to step S4;

在步骤S4中,用户u5执行所分配的特定任务τ1时付出的总代价c15,用户u5执行所分配特定任务τ5时带来的纯收益w15,c15=50,w15=10,B′1=75,δ=25,则(c15+w15)-B′1≤δ成立,转向步骤S5;In step S4, the total cost c 15 paid by the user u 5 to perform the assigned specific task τ 1 , the net benefit w 15 brought by the user u 5 when the user u 5 performs the assigned specific task τ 5 , c 15 =50, w 15 =10, B' 1 =75, δ=25, then (c 15 +w 15 )-B' 1 ≤δ is established, go to step S5;

在步骤S5中,对于用户u5,令η=0.8,由于

Figure BDA0001819778780000134
c15+w15=60,则公式(7)成立,将用户u5加入集合
Figure BDA0001819778780000136
中,转向步骤S6;In step S5, for user u 5 , let n=0.8, since
Figure BDA0001819778780000134
c 15 +w 15 =60, then formula (7) is established, and user u 5 is added to the set
Figure BDA0001819778780000136
, turn to step S6;

在步骤S6中,从用户集合U'中删除用户u5,再更新Qi'和Bi'的值,Q1'=Q1'-q5=4-3=1,B1'=B1'-(c15+w15)=75-60=15,由于Q1'=1,则公式Q1'≤0不成立,转向步骤S3;In step S6, delete the user u 5 from the user set U', and then update the values of Q i ' and B i ', Q 1 '=Q 1 '-q 5 =4-3=1, B 1 '=B 1 '-(c 15 +w 15 )=75-60=15, since Q 1 '=1, the formula Q 1 '≤0 does not hold, and turn to step S3;

在步骤S3中,uj*=u6,转向步骤S4;In step S3, u j* =u 6 , turn to step S4;

在步骤S4中,对于用户u6,由于c16=40,w16=8,B′1=15,δ=25,则(c16+w16)-B′1≤δ不成立,转向步骤S3;In step S4, for user u 6 , since c 16 =40, w 16 =8, B′ 1 =15, and δ=25, then (c 16 +w 16 )-B′ 1 ≤δ does not hold, and turn to step S3 ;

在步骤S3中,由于找不到新的用户可以执行任务τ1,因此,任务τ1无法分配,因此令y1=0,并在任务集合Γ中删除此任务,再转向步骤S2;In step S3, since no new user can perform task τ 1 , task τ 1 cannot be assigned, so set y 1 =0, delete this task in task set Γ, and turn to step S2;

任务τ1预算不足,且已超过平台可以弥补任务预算的范围,因此无法完成分配;The budget of task τ 1 is insufficient and has exceeded the range that the platform can make up for the task budget, so the allocation cannot be completed;

在步骤S2中,集合

Figure BDA0001819778780000141
则集合Γ中任务已全部分配完成;In step S2, the collection
Figure BDA0001819778780000141
Then the tasks in the set Γ have all been allocated and completed;

(5)根据下式(8)为执行任务τi*i*∈Γ)的用户

Figure BDA0001819778780000142
分配报酬rj*,(5) According to the following formula (8), it is the user who performs the task τ i*i* ∈Γ)
Figure BDA0001819778780000142
Distribution reward r j* ,

Figure BDA0001819778780000143
Figure BDA0001819778780000143

执行任务τ2的用户集合为U2 +,计算得r1=70+14=84,r2=90+18=108,将报酬r1和r2分别付给用户u1和用户u2,执行任务τ2的用户集合为

Figure BDA0001819778780000144
计算得r3=55+11=66,r4=45+9=54,将报酬r3和r4分别付给用户u3和用户u4,其余任务及用户未分配,无报酬;当用户uj*完成任务并获得报酬τi*后,将所有xi*j*的值设为0,其中,τi*∈Γ,并将用户uj*加入用户集合U中,支付报酬也可以在相应用户完成感知任务并上传感知数据后再进行。The set of users who perform task τ 2 is U 2 + , r 1 =70+14=84, r 2 =90+18=108 are calculated, and the rewards r 1 and r 2 are paid to user u 1 and user u 2 respectively, The set of users who perform task τ 2 is
Figure BDA0001819778780000144
It is calculated that r 3 =55+11=66, r 4 =45+9=54, the rewards r 3 and r 4 are paid to user u 3 and user u 4 respectively, and the rest of the tasks and users are not assigned and unpaid; when the user After u j* completes the task and gets paid τ i* , set all x i*j* values to 0, where, τ i* ∈ Γ, and the user u j* is added to the user set U, and the payment can also be done after the corresponding user completes the sensing task and uploads the sensing data.

Claims (1)

1. a kind of intelligent perception method for allocating tasks based on allocation of budget, which is characterized in that comprise the steps of:
(1) perception task set Γ={ τ that task requester is submitted is collected1,...,τi,...,τn, i ∈ { 1 ..., n }, n is One limited natural number, task τi=(Bi,Qi), BiIt is to complete task τiMaster budget, be a limited positive real number, QiIt is Task τiQuality requirement value, be a limited positive integer;
(2) user's set U={ u is collected1,...,uj,...,umAnd the user set in user ujCredit rating qj, j ∈ { 1 ..., m }, m are a limited natural number;
(3) user u is calculatedjComplete task τiCost cij, and determine user ujComplete task τiIncome wij
(4) according to task τiQuality requirement QiWith budget BiAnd user ujCredit rating qjTo in perception task set Γ Task is allocated, and specific distribution formula is as follows:
Figure FDA0002167272790000011
In formula, yiIndicate task τiWhether at the distribution of work, yi=1 indicates task τiAt the distribution of work, τi∈ Γ, yi=0 indicates task τi Distribution failure;xijIt indicates whether task τiDistribute to user uj, uj∈ U, xij=1 indicates task τiDistribute to user uj, xij =0 indicates not by task τiDistribute to user uj, δ expression participation task τiThe sum of the remuneration of all users and task master budget Difference,
Figure FDA0002167272790000021
Figure FDA0002167272790000022
Unspent budget after indicating the previous task distribution of platform, is a positive real number, η ∈ (0,1) is indicated The utilization rate of unspent budget;
Formula (1) is an integer programming problem, is allocated by the following method:
S1: order set Γ ' is set of tasks to be allocated, Γ '=Γ when initial, for any task τi, τi∈ Γ ', enables Q 'i Indicate task τiRemaining quality requirement value, B 'iIndicate task τiResidual,
Figure FDA0002167272790000023
After indicating the previous task distribution of platform Unspent budget, Q ' when initiali=Qi, B 'i=Bi,
Figure FDA0002167272790000024
Enable set
Figure FDA0002167272790000025
For remaining user set, set U' is optional User set, when initial
Figure FDA0002167272790000026
Set
Figure FDA0002167272790000027
Expression can execute task τiUser set, initially
S2: judgement set
Figure FDA0002167272790000029
It is whether true, if so, according to the following formula
Figure FDA00021672727900000210
Find out a particular task τi*, particular task τi*It is specific initial for the task that the quality requirement in unit budget is minimum Be engaged in τi*Quality requirement value Q 'i*=Qi*, particular task τi*Residual B 'i*=Bi*, particular task τi*Available user Set
Figure FDA00021672727900000211
Otherwise, stop;
S3: one is found out according to the following formula closest to particular task τi*The user u of quality requirementj*,
Figure FDA00021672727900000212
In formula, ci*jIndicate user ujExecute distributed particular task τi*When the cost paid, wi*jIndicate user ujExecution divides The particular task τ matchedi*When brought net earnings if can find turn to step S4;Otherwise, it is deleted in set of tasks Γ This task, and enable yi*=0, i.e. particular task τi*Distribution failure, turns again to step S2;
S4: judge user uj*Execute particular task τi*Remuneration and particular task τi*Whether difference, that is, following formula of budget is true,
(ci*j*+wi*j*)-B′i*≤δ
C in formulai*j*Indicate user uj*Execute distributed particular task τi*When the total cost paid, wi*j*Indicate user uj*It executes Distributed particular task τi*When bring net earnings, if set up if turn to step S5;Otherwise, in available user's set U' Delete user uj*, turn again to step S3 and find new user;
S5: judge whether platform can utilize unspent budget
Figure FDA0002167272790000031
To help particular task τi*Complete distribution, i.e., whether following formula is true,
Figure FDA0002167272790000032
If so, then user uj*Addition meets task τi*The user of quality requirement gathers
Figure FDA0002167272790000033
In, turn to step S6;Otherwise, It should be user uj*It is deleted from available user's set U', turns again to step S3 and find new user;
S6: by user uj*It is deleted from available user's set U', according to formula Q 'i*=Qi*'-qj*With formula B 'i*=B ′i*-(ci*j*+wi*j*) update Q 'i*With B 'i*Value, then judge task τi*Quality requirement whether it is satisfied, i.e. formula Q 'i*≤ 0 is No establishment, if so, turn to step S7;If not, it turns to step S3 and continually looks for user;
S7: user's set
Figure FDA0002167272790000034
As execute task τi*A feasible user set, enable yi*=1, i.e. particular task τi*At The distribution of work enables xi*j*=1, i.e., by task τiDistribute to user uj*,
Figure FDA0002167272790000035
Calculate unspent budget
Figure FDA0002167272790000036
It enables Γ '=Γ '-{ τi*,
Figure FDA0002167272790000037
The lower task of step S2 is turned again to be allocated;
(5) it according to task allocation result, calculates and pays user's remuneration rj, for executing particular task τi*User uj*It obtains Remuneration rj*:
Figure FDA0002167272790000038
As user uj*Complete particular task τi*And after obtaining remuneration, by all xi*j*Value be set as 0,
Figure FDA0002167272790000039
τi*∈ Γ, And by user uj*It is added in user's set U.
CN201811159885.4A 2018-09-30 2018-09-30 A Crowdsensing Task Allocation Method Based on Budget Allocation Active CN109408228B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811159885.4A CN109408228B (en) 2018-09-30 2018-09-30 A Crowdsensing Task Allocation Method Based on Budget Allocation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811159885.4A CN109408228B (en) 2018-09-30 2018-09-30 A Crowdsensing Task Allocation Method Based on Budget Allocation

Publications (2)

Publication Number Publication Date
CN109408228A CN109408228A (en) 2019-03-01
CN109408228B true CN109408228B (en) 2019-10-15

Family

ID=65465821

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811159885.4A Active CN109408228B (en) 2018-09-30 2018-09-30 A Crowdsensing Task Allocation Method Based on Budget Allocation

Country Status (1)

Country Link
CN (1) CN109408228B (en)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110782150B (en) * 2019-10-21 2024-02-09 青岛市黄岛区自然资源局 Natural resource information collection system and method based on crowd sensing
CN110825500B (en) * 2019-10-28 2022-10-25 陕西师范大学 A three-layer structure-oriented crowd-sensing task assignment method
CN111475266B (en) * 2020-03-06 2021-11-30 陕西师范大学 Diversity-constrained crowd sensing task allocation method
CN111639856A (en) * 2020-05-29 2020-09-08 中国人民解放军国防科技大学 City environment object calibration method, system and medium based on crowd sensing technology
CN112990728A (en) * 2021-03-26 2021-06-18 山东大学 Large-scale crowd sensing task allocation method based on characteristic information
CN113139792B (en) * 2021-04-27 2023-05-26 陕西师范大学 Specific material collection method based on crowd sensing technology

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106209874A (en) * 2016-07-18 2016-12-07 沈阳师范大学 A kind of intelligent perception distribution system and method for allocating tasks thereof
CN108171593A (en) * 2017-11-17 2018-06-15 南京邮电大学 It is a kind of that the mobile intelligent perception motivational techniques with budget are covered based on place
CN108364190A (en) * 2018-01-08 2018-08-03 东南大学 In conjunction with the online motivational techniques of mobile intelligent perception of reputation updating

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105282246B (en) * 2015-10-23 2018-08-24 上海交通大学 The method of perception task distribution based on auction mechanism
CN106385387B (en) * 2016-09-27 2019-08-02 中国科学院空间应用工程与技术中心 A kind of resource regulating method, system and the application of Information Network link
CN108269129B (en) * 2018-01-23 2022-02-08 重庆邮电大学 User incentive method in mobile crowd sensing network based on reverse auction

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106209874A (en) * 2016-07-18 2016-12-07 沈阳师范大学 A kind of intelligent perception distribution system and method for allocating tasks thereof
CN108171593A (en) * 2017-11-17 2018-06-15 南京邮电大学 It is a kind of that the mobile intelligent perception motivational techniques with budget are covered based on place
CN108364190A (en) * 2018-01-08 2018-08-03 东南大学 In conjunction with the online motivational techniques of mobile intelligent perception of reputation updating

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Budget-Feasible Online Incentive Mechanisms for Crowdsourcing Tasks Truthfully;Dong Zhao,et.al.;《IEEE/ACM Transactions on Networking (TON)》;20160430;第2卷(第24期);全文 *
基于机会协作传输的群智感知网络激励机制;贾超鹏等;《计算机应用与软件》;20180331(第3期);全文 *
移动群智感知中收益最大化的用户招募算法;郭会东等;《小型微型计算机系统》;20180331(第03期);全文 *

Also Published As

Publication number Publication date
CN109408228A (en) 2019-03-01

Similar Documents

Publication Publication Date Title
CN109408228B (en) A Crowdsensing Task Allocation Method Based on Budget Allocation
CN104038540B (en) Method and system for automatically selecting application proxy server
WO2019024445A1 (en) Collaborative optimization method for geographic distribution interactive service cloud resource
CN106209874B (en) A kind of method for allocating tasks for intelligent perception distribution system
CN108200610B (en) Crowd-sensing resource allocation method using distributed game
CN105204924A (en) Managing private use of program execution capacity
CN103064744B (en) The method for optimizing resources that a kind of oriented multilayer Web based on SLA applies
CN108876567A (en) A kind of intelligent perception motivational techniques based on perception maximization of utility
CN109684078A (en) Resource dynamic distributing method and system for spark streaming
CN110503396A (en) Multi-skill-based task assignment method for complex spatial crowdsourcing
CN107292420A (en) Data processing method and device
CN106485415B (en) A kind of mobile intelligent perception motivational techniques with budget based on relation between supply and demand
CN106875211A (en) Optimal service pricing method based on user's perceived value in a kind of cloud computing environment
CN107563616A (en) A kind of user task distribution and the quorum-sensing system system and method for incentives strategy
CN103514048B (en) Participate in sensory perceptual system and method for allocating tasks thereof
CN106803815B (en) Flow control method and device
Jiao et al. Cost optimization for online social networks on geo-distributed clouds
CN112583609A (en) Cloud charging method and device, cloud management platform and system and storage medium
CN105282246B (en) The method of perception task distribution based on auction mechanism
CN103854208B (en) A kind of implementation method of cloud market multi-kind resource distribution pricing mechanism
Shang et al. An Intelligent Capacity Planning Model for Cloud Market.
CN105338064A (en) Task collaborative execution method based on Stackelberg Game
CN109086976B (en) A Crowdsensing-Oriented Task Assignment Method
CN115169634A (en) A kind of task allocation optimization processing method and device
CN109886581A (en) A data quality-based participant selection method for mobile crowdsensing tasks

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20220913

Address after: Room 884, Room 406, No. 1 Yichuang Street, Huangpu District, Guangzhou City, Guangdong Province (Sino-Singapore Guangzhou Knowledge City) 510000

Patentee after: Guangzhou Aladdin Intelligent Technology Co.,Ltd.

Address before: 710062 No. 199 South Changan Road, Shaanxi, Xi'an

Patentee before: Shaanxi Normal University