CN115390946A - A Mine Accident Rescue Task Offloading Method Based on Mobile Edge Computing - Google Patents

Abstract

The invention discloses a mine accident rescue task unloading method based on mobile edge computing. The method completes the rescue task based on an evolutionary game theory dynamic unloading strategy in mobile edge computing, and belongs to the field of mine accidents. This method mainly includes the following four steps: First, construct a "cloud + edge" mixed mine accident scenario composed of "central mine accident + multiple chain mine accidents", and calculate the optimization target formula; secondly, use dynamic evolutionary game rules, Design the form of the evolutionary game, including participants, groups, strategies and customize the evolutionary stable strategy; then, design the replication dynamic equation and introduce the incentive coefficient; finally, drive the evolutionary game model through the probability distribution, and finally generate a multi-evolutionary stability of service offloading Strategy; the present invention considers that the chain mine accidents in the real environment are unstable and fluctuate with time, and the dynamic equation is designed to replicate, and the incentive coefficient is introduced to improve the efficiency strategy of the strategy stability.

Description

A Mine Accident Rescue Task Offloading Method Based on Mobile Edge Computing

technical field

The invention relates to a mine accident rescue task unloading method based on mobile edge computing, which belongs to the field of mine accidents and edge computing, and in particular relates to a solution to the mine accident rescue task unloading problem.

Background technique

In recent years, mine accidents have occurred frequently in our country, and the economy and society have been seriously affected. How to improve the level of emergency management and effectively deal with various mine accidents is a problem of widespread concern to the society. As the core throughout the entire life cycle of emergency management, the emergency rescue team is very necessary to deploy emergency rescue tasks.

The traditional game theory is Nash equilibrium, which requires a single rescue team to maximize its own income, and then obtain an equilibrium state that every rescue team can accept. At the same time, due to the late start of research on rescue systems in my country and the complexity of accidents in mining areas, there are problems in research on mine rescue: most of the existing research only considers ideal conditions, making it impossible to solve complex problems. The problem of mine accidents. Moreover, the scene of a mine accident is not fixed and changes with time. If the rescue plan is designed according to ideal conditions, it will cause a great waste of resources. And in the environment described in the present invention, rescue team can't reach complete rationality. In this incompletely rational state, the present invention considers the pursuit of the maximum fitness of the group, and then obtains an evolutionary stable strategy.

Aiming at the scene of task unloading of multiple rescue teams in the above-mentioned hybrid "cloud + edge" environment, the present invention designs a mine accident rescue task unloading method based on mobile edge computing; the reason why the evolutionary game is used to solve this problem is due to the traditional Game theory pursues Nash equilibrium, which requires a single rescue team to maximize its own benefits under completely rational conditions, so as to achieve an equilibrium state acceptable to each participant. In the hybrid "cloud + edge" environment, multiple rescue teams geographically distributed do not have completely transparent information and overall logical deduction capabilities, so they cannot achieve complete rationality. In this incompletely rational state, the present invention considers the pursuit of the maximum fitness of the population, and then obtains an evolutionary stable strategy; this strategy will not cause the overall recalculation due to the change of a certain participant's decision, thus saving a lot of decision-making calculations overhead.

Contents of the invention

The present invention provides a mine accident rescue task unloading method based on mobile edge computing, which is used to solve the problem of how to efficiently unload work rescue tasks when mine accidents occur.

The technical solution of the present invention is: a mine accident rescue task unloading method based on the game strategy of replicating dynamic evolution, and the specific steps of the method are as follows:

Step1. Construct a "cloud + edge" mixed mine accident scenario composed of "central mine accident + multiple chain mine accidents", and calculate the optimization target formula.

Step2. Using dynamic evolutionary game rules, design the form of evolutionary game, including participants, groups, strategies and customize evolutionary stable strategies.

Step3. Design and copy the dynamic equation, introduce the excitation coefficient, and solve the model.

As a further solution of the present invention, said Step1 includes:

其中J是矿山事故集合{EM₀,EM₁,...,EM_j}；救援队S_i将从A_i中选择一个矿山事故进行救援，救援时间

由四部分组成，分别为任务派遣时间、前往路程时间、回归路程时间及事故处理时间；其中任务派遣时间T_i ^as非常短可忽略不计，

T_i ^re和T_i ^ac则需要根据矿山事故距离和难度计算；因此救援队S_i的救援时间可以表示为：For the rescue team S _i , it is first necessary to determine the distance between it and each mine accident; if r(i,k)<R _k is satisfied, it means that the rescue team S _i can provide rescue for the emergency EM _j ; A _i represents the set of candidate mine accidents of S _i , satisfying

Where J is the set of mine accidents {EM ₀ , EM ₁ ,...,EM _j }; the rescue team S _i will select a mine accident from A _i for rescue, and the rescue time

It consists of four parts, which are task dispatch time, forward travel time, return travel time and accident processing time; among them, the task dispatch time T _i ^as is very short and negligible,

T _i ^re and T _i ^ac need to be calculated according to the mine accident distance and difficulty; therefore, the rescue time of rescue team S _i can be expressed as:

和

分别表示EM_j单位运输成本的价格和单位物资成本的费用；如果矿山事故的等级G_j更高，并且有更多的救援队救援，则每个救援队分摊的救援人工成本就越低；因此救援队S_i的救援货币成本为：The monetary cost of the rescue team includes three parts, namely, transportation cost, material cost and labor cost;

and

Respectively represent the price of the unit transportation cost of EM _j and the cost of the unit material cost; if the grade G _j of the mine accident is higher, and there are more rescue teams to rescue, the rescue labor cost shared by each rescue team will be lower; therefore The rescue monetary cost of the rescue team S _i is:

—表示j号矿山事故单位运输成本的价格；r(i,j)—i号救援队距离j号矿山事故的距离；

—j号矿山事故单位物资成本的费用；G_j—j号矿山事故的等级；M_j—j号矿山事故总的人工耗用成本；num_j—j号矿山事故救援队总数量。In the formula: M _i —rescue monetary cost of rescue team i; M _j ^tr —transportation cost of mine accident j selected by rescue team i; M _j ^ma —material cost of mine accident j selected by rescue team _i ; ^ar —the labor cost of No. j mine accident selected by No. i rescue team;

- the price representing the unit transportation cost of No. j mine accident; r(i, j) - the distance between No. i rescue team and No. j mine accident;

—the unit material cost of No. j mine accident; G _j —the grade of No. j mine accident; M _j —the total labor cost of No. j mine accident; num _j —the total number of rescue teams for No. j mine accident.

In the present invention, FS _i is used to represent the expected attainment degree of the rescue team of S _i , which can be expressed as the probability that both unloading time and monetary cost constraints are satisfied:

FS _i =Pr(T _i ≤t _i )·Pr(M _i ≤m _i ) (13)

In the formula: FS _i —the expected achievement degree of the rescue team i; T _i —the actual rescue time of the rescue team i; t _i —the rescue time constraint of the rescue team i; M _i —the rescue currency cost of the rescue team i; m _i — Rescue monetary cost constraints of rescue team i.

代表区域内k号救援队群体Q_k的期望达成度，其中num_k表示Q_k中的救援队数量：

Represents the expected attainment degree of rescue team Q _k in the area k, where num _k represents the number of rescue teams in Q _k :

—k号救援队群体的期望达成度；FS_i—i号救援队期望达成度；num_k—k号救援队群体中救援队数量。In the formula: Q _k —k number rescue team group;

—the expected achievement degree of the rescue team group k; FS _i —the expected achievement degree of the rescue team number i; num _k —the number of rescue teams in the rescue team group k.

FS is the expected attainment degree of the entire area, and it is used as the final optimization goal. The optimization goal formula is:

—k号救援队群体的期望达成度；K—区域内救援队群体的总数。In the formula: FS—the expected achievement degree of the whole area;

—The expected attainment degree of rescue team group k; K—the total number of rescue team groups in the area.

As a further solution of the present invention, said Step2 designs the evolutionary game form as follows:

The form of the evolutionary game is designed as follows:

Participants: For the "cloud + edge" mixed multi-rescue team environment shown in Figure 1, each terminal rescue team in the area is a participant in the evolutionary game; under the assumption of bounded rationality, the purpose of participating in the game is to pass Participate in the game to maximize their own benefits.

Group: As shown in Figure 1, the rescue teams can be divided into K different groups according to their positions, and the number of rescue teams in each group is represented by num _k ; the present invention uses {Q ₀ ,Q ₁ ,..., Q _K-1 } to represent the set of rescue teams of K groups, the participants in each group are located in the same geographical area and the sum of the number of rescue teams satisfying all groups is the total number of rescue teams.

Strategy: The strategy of each rescue team refers to the mine accidents it can choose. In this game environment, there are 1+J kinds of mine accidents for participants to choose from. 1 is a central mine accident, and J is a chain mine accident; use x _j to indicate whether the rescue team chooses the mine accident EM _j to unload the task. The actual situation is as follows:

Among them, the optional mine accident set S _k of Q _k is a subset of J.

表示在Q_k群体中选择EM_j矿山事故进行救援的救援队总数，

表示整个群体选择EM_j进行救援的群体份额；

Group share:

Indicates the total number of rescue teams who choose EM _j mine accident rescue in the Q _k population,

Indicates the group share of the whole group choosing EM _j for rescue;

来表示群体P_k的策略选择状态，且满足

用一个矩阵P来表示包含J个群体状态的总群体状态空间如下：Crowd state: use a vector

To represent the strategy selection state of the group P _k , and satisfy

A matrix P is used to represent the total population state space containing J population states as follows:

Benefits: The benefits of the participants depend on their net utility function, which is determined by their maximum bearable rescue time and cost as well as the actual rescue time and cost; then the rescue team S _i of the Q _k group chooses the value obtained from the EM _j mine accident The benefit can be expressed as follows:

Step3. Design and copy the dynamic equation, introduce the excitation coefficient, and solve the model.

In traditional games, all participants can reach a stable state, that is, no participant can obtain additional benefits by unilaterally changing the strategy. This state is called NE; the selection of servers and resource allocation under the hybrid architecture can be seen As a game Γ, the set of participants is N, and the set of strategies is U=[u1,u2,u3....uN], which means that the strategy of each participant has been counted; use u _-i =[u ₁ , ..., u _i , u _i+1 ..., u _N ] represent the strategy combination chosen by other people except participant S _i ; in the income set W=[γ1,γ2,γ3...,γN] Among them, the profit of the participant S _i is related to the strategy chosen by himself and the strategy chosen by others, expressed as γ(u _i ,u _-i ).

In a group of players with bounded rationality, strategies with results better than the average level will gradually be adopted by more players, and the proportion of players' strategies will also change accordingly; the details are as follows:

—对时间进行一阶求导得到的函数；

—群体中的参与者选择服务器EM_j时所获得的当前收益；

_t 时刻群体的平均收益；In the formula: λ—controls the convergence speed of strategy adaptation of participants in the same group;

— the function obtained by first-order derivation with respect to time;

— the current benefits obtained by the participants in the group when they choose the server EM _j ;

_ The average return of the group at time t;

表示对时间进行一阶求导得到的函数，

是群体中的参与者选择服务器EM_j时所获得的当前收益，

为t时刻群体的平均收益，可以通过下式计算得到：where δ is used to control the convergence speed of the strategy adaptation of the participants in the same group; the growth rate

Represents the function obtained by taking the first-order derivative with respect to time,

is the current payoff obtained by the participants in the group when they choose the server EM _j ,

is the average income of the group at time t, which can be calculated by the following formula:

本发明可以得到复制者动态的不动点，在该不动点上，由于同一群体中的所有参与者都有相同的收益，因此群体状态不会改变，也没有参与人愿意改变策略。Based on the replicator dynamic equation of strategy selection in group Q _k , when the income of choosing strategy u is higher than the average income of the same group, the number of terminal rescue teams who choose this strategy will generally show a positive growth trend. by setting

The present invention can obtain the dynamic fixed point of the replicator, at which point, since all participants in the same group have the same benefits, the state of the group will not change, and no participant is willing to change the strategy.

Beneficial effects of the present invention:

The present invention studies the environment of multiple mine accidents concurrently, that is, the rescue problem in the mixed environment formed by a central mine accident and multiple mine accidents; it is different from many traditional methods and strategies, considering that the real-time situation of mine accidents changes with time Fluctuations and changes; by proposing an evolutionary game strategy in a targeted manner, the rescue team's rescue strategy is dynamically obtained, which improves the rescue team's rescue efficiency.

Description of drawings

Figure 1 is a flowchart of the invention.

FIG. 2 is a schematic diagram of a specific scenario and a specific implementation manner.

Detailed ways

The present invention will be described in further detail below in conjunction with specific examples, but the protection scope of the present invention is not limited to the content described.

In order to solve the problem of more efficient completion of rescue tasks, the embodiment of the present invention provides a mine accident rescue task unloading method based on mobile edge computing, and constructs a mine accident rescue task unloading model based on mobile edge computing for use in Solve the problem of mine accidents and how to efficiently unload work rescue tasks.

Example 1

Specific applicable environments include but are not limited to scenarios where multiple mine accidents occur concurrently. First of all, in terms of accidents, when a central mine accident occurs, multiple mine accidents will occur, which has a chain effect. Secondly, in the area where multiple mine accidents occur, there are multiple rescue teams distributed in different locations, and the rescue teams can be divided into multiple groups. Then, there are many options for rescue teams to choose mine accidents for rescue. In this environment, this embodiment aims at rescue cost and time limit, uses dynamic evolutionary game analysis, obtains the best rescue strategy of the rescue team, and improves the rescue efficiency, specifically including the following steps:

Step1. Construct a "cloud + edge" mixed mine accident scenario composed of "central mine accident + multiple chain mine accidents", and calculate the optimization target formula.

The present invention assumes that both the central mine accident and the chain mine accident device have their own rescue ranges, and rescue teams within the range can rescue them; chain mine accidents can be rescued nearby by the rescue team. Also consider a rescue environment consisting of a single central mine accident and J chained mine accidents. The central mine accident is rescued in a relatively large area; while the J chain mine accidents are closer to the rescue team. Chain mine accident rescue costs. The above-mentioned mixed environment area can usually be divided into K small service areas, and the rescue team in each area is regarded as a group. The principle of division is to ensure that the number of chain mine accidents in each area is almost equal.

其中J是矿山事故集合{EM₀,EM₁,...,EM_j}；救援队S_i将从A_i中选择一个矿山事故进行救援，救援时间

由四部分组成，分别为任务派遣时间、前往路程时间、回归路程时间及事故处理时间；其中任务派遣时间T_i ^as非常短可忽略不计，

T_i ^re和T_i ^ac则需要根据矿山事故距离和难度计算；因此救援队S_i的救援时间可以表示为：For the rescue team S _i , it is first necessary to determine the distance between it and each mine accident; if r(i,k)<R _k is satisfied, it means that the rescue team S _i can provide rescue for the emergency EM _j ; A _i represents the set of candidate mine accidents of S _i , satisfying

Where J is the set of mine accidents {EM ₀ , EM ₁ ,...,EM _j }; the rescue team S _i will select a mine accident from A _i for rescue, and the rescue time

It consists of four parts, which are task dispatch time, forward travel time, return travel time and accident processing time; among them, the task dispatch time T _i ^as is very short and negligible,

T _i ^re and T _i ^ac need to be calculated according to the mine accident distance and difficulty; therefore, the rescue time of rescue team S _i can be expressed as:

和

分别表示EM_j单位运输成本的价格和单位物资成本的费用；如果矿山事故的等级G_j更高，并且有更多的救援队救援，则每个救援队分摊的救援成本就越低；因此救援队S_i的救援货币成本为：The monetary cost of the rescue team includes three parts, namely, transportation cost, material cost and labor cost;

and

Respectively represent the price of the unit transportation cost of EM _j and the cost of the unit material cost; if the grade G _j of the mine accident is higher, and there are more rescue teams to rescue, the rescue cost shared by each rescue team will be lower; therefore the rescue The rescue monetary cost of team S _i is:

In the present invention, FS _i is used to represent the expected attainment degree of the rescue team of S _i , which can be expressed as the probability that both unloading time and monetary cost constraints are satisfied:

FS _i =Pr(T _i ≤t _i )·Pr(M _i ≤m _i ) (23).

代表区域内k号救援队群体Q_k的期望达成度，其中num_k表示Q_k中的救援队数量：

Represents the expected attainment degree of rescue team Q _k in the area k, where num _k represents the number of rescue teams in Q _k :

FS is the expected attainment degree of the entire region, and it is used as the final optimization goal:

Step2. Using dynamic evolutionary game rules, design the form of evolutionary game, including participants, groups, strategies and customize evolutionary stable strategies.

The standard formula of the classic game includes three factors: the participants of the game, the set of strategies that each participant can choose and the combination of strategies that may be selected by all participants, and the profit function of each participant when choosing a strategy; in the evolution In the context of games, a population can be used to represent a group of players with the same properties.

The form of the evolutionary game is designed as follows:

Participants: For the "cloud + edge" mixed multi-rescue team environment shown in Figure 1, each terminal rescue team in the area is a participant in the evolutionary game; under the assumption of bounded rationality, the purpose of participating in the game is to pass Participate in the game to maximize their own benefits.

Group: As shown in Figure 1, the rescue teams can be divided into K different groups according to their positions, and the number of rescue teams in each group is represented by num _k ; the present invention uses {Q ₀ ,Q ₁ ,..., Q _K-1 } to represent the set of rescue teams of K groups, the participants in each group are located in the same geographical area and the sum of the number of rescue teams satisfying all groups is the total number of rescue teams.

Strategy: The strategy of each rescue team refers to the mine accidents it can choose. In this game environment, there are 1+J kinds of mine accidents for participants to choose from. 1 is a central mine accident, and J is a chain mine accident; use x _j to indicate whether the rescue team chooses the mine accident EM _j to unload the task. The actual situation is as follows:

Among them, the optional mine accident set S _k of Q _k is a subset of J.

表示在Q_k群体中选择EM_j矿山事故进行救援的救援队总数，

表示整个群体选择EM_j进行救援的群体份额；

Group share:

Indicates the total number of rescue teams who choose EM _j mine accident rescue in the Q _k population,

Indicates the group share of the whole group choosing EM _j for rescue;

来表示群体P_k的策略选择状态，且满足

用一个矩阵P来表示包含J个群体状态的总群体状态空间如下：Crowd state: use a vector

To represent the strategy selection state of the group P _k , and satisfy

A matrix P is used to represent the total population state space containing J population states as follows:

Benefits: The benefits of the participants depend on their net utility function, which is determined by their maximum bearable rescue time and cost as well as the actual rescue time and cost; then the rescue team S _i of the Q _k group chooses the value obtained from the EM _j mine accident The benefit can be expressed as follows:

Step3. Design and copy the dynamic equation, introduce the excitation coefficient, and solve the model.

In the present invention, the terminal rescue team under the cloud-edge hybrid architecture adjusts their strategies in a limited set of strategies to obtain better returns. At each moment, the terminal rescue team can obtain its own strategy set, and at the same time obtain the average income information of its group; thus, it can continuously evolve its own server selection strategy as time goes by; after enough repetitions and deduction, use The number of mutant participants of other strategies will continue to decrease due to low income, and finally the mutant players will become extinct and reach the evolutionary stable state of the group. This dynamic process can be solved by replicating the dynamic equation in the present invention.

The evolutionary game model is mainly based on the mutation mechanism and the selection mechanism. The mutation refers to that in a group, a small number of participants randomly choose a strategy different from that of the majority of participants, and the selected strategy may obtain higher results. It may also obtain lower returns, but only excellent strategies will not be eliminated by history, but will be retained; selection means that the strategy can obtain higher returns and can be continuously used by other participants in the future. Adoption; this selection mechanism is the key to the dynamic model of replicators, it provides a way to obtain information about other groups, and converges to the equilibrium point until the strategy adapts to reach evolutionary equilibrium (EE), that is, the group will not change its choose. The basic idea is that in a group of players with bounded rationality, strategies with results better than the average level will gradually be adopted by more players, and the proportion of players' strategies will also change accordingly; the details are as follows:

表示对时间进行一阶求导得到的函数，

是群体中的参与者选择服务器EM_j时所获得的当前收益，

为t时刻群体的平均收益，可以通过下式计算得到：where δ is used to control the convergence speed of the strategy adaptation of the participants in the same group; the growth rate

Represents the function obtained by taking the first-order derivative with respect to time,

is the current payoff obtained by the participants in the group when they choose the server EM _j ,

is the average income of the group at time t, which can be calculated by the following formula:

本发明可以得到复制者动态的不动点，在该不动点上，由于同一群体中的所有参与者都有相同的收益，因此群体状态不会改变，也没有参与人愿意改变策略。Based on the replicator dynamic equation of strategy selection in group Q _k , when the income of choosing strategy u is higher than the average income of the same group, the number of terminal rescue teams who choose this strategy will generally show a positive growth trend. by setting

The present invention can obtain the dynamic fixed point of the replicator, at which point, since all participants in the same group have the same benefits, the state of the group will not change, and no participant is willing to change the strategy.

The specific embodiments of the present invention have been described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned embodiments. Variations.

Claims (4)

1. A mine accident rescue task unloading method based on mobile edge computing, characterized in that, the specific steps of the construction method are as follows: Step1. Construct a "cloud + edge" mixed mine accident scenario composed of "central mine accident + multiple chain mine accidents", and calculate the optimization target formula; Step2. Using dynamic evolutionary game rules, design the form of evolutionary game, including participants, groups, strategies and customize evolutionary stable strategies; Step3. Design and copy the dynamic equation, introduce the excitation coefficient, and solve the model. 2. The Mine Accident Rescue Task Unloading Method Based on Mobile Edge Computing according to claim 1, characterized in that: said Step1 comprises: The monetary cost of the rescue team includes three parts, namely transportation cost, material cost and labor cost;

and

Respectively represent the price of unit transportation cost and unit material cost of No. j mine accident; if the level of mine accident G _j is higher and there are more rescue teams to rescue, the additional rescue cost shared by each rescue team will be lower ; Therefore, the rescue currency cost of rescue team S _i of number i is:

In the formula: M _i —rescue monetary cost of rescue team i;

—The transportation cost of No. j mine accident selected by No. i rescue team;

—The material cost of No. j mine accident selected by No. i rescue team;

—The labor cost of No. j mine accident selected by No. i rescue team;

— represents the price of unit transportation cost of No. j mine accident; r(i,j)—the distance between No. i rescue team and No. j mine accident;

—The expense of material cost of the No. j mine accident unit; G _j — grade of mine accident j; M _j — the total labor cost of mine accident j; num _j —the total number of mine accident rescue teams of No. j; Using FS _i to represent the rescue team's expected achievement degree, it can be expressed as the probability that both the rescue time and monetary cost constraints are satisfied: FS _i =Pr(T _i ≤t _i )·Pr(M _i ≤m _i ) (2) In the formula: FS _i — the expected achievement degree of No. i rescue team; T _i —actual rescue time of No. i rescue team; t _i —rescue time constraint of No. i rescue team; M _i —rescue monetary cost of rescue team i; m _i —rescue currency cost constraint of No. i rescue team;

Represents the expected attainment degree of rescue team Q _k in the area k, where num _k represents the number of rescue teams in Q _k :

In the formula: Q _k —k number rescue team group;

—The degree of achievement of the group's expectations of the No.k rescue team; FS _i — the expected achievement degree of No. i rescue team; num _k —the number of rescue teams in the number k rescue team group; FS is the expected attainment degree of the entire area, and it is used as the final optimization goal. The optimization goal formula is:

In the formula: FS—the expectation achievement degree of the whole area;

—The degree of achievement of the group's expectations of the No.k rescue team; K—the total number of rescue team groups in the area. 3. The mine accident rescue task unloading method based on mobile edge computing according to claim 1, characterized in that: said Step2 designs the form of evolutionary game as follows: The form of the evolutionary game is designed as follows: Participants: In the "cloud + edge" mixed multi-rescue team environment, each terminal rescue team in the region is a participant in the evolutionary game; under the assumption of bounded rationality, the purpose of participating in the game is to realize its own benefits by participating in the game maximize; Group: Rescue teams can be divided into K different groups according to their positions, and the number of rescue teams in each group is represented by num _k ; represented by {Q ₀ ,Q ₁ ,...,Q _K-1 } A collection of rescue teams of groups, the participants in each group are located in the same geographical area and the sum of the number of rescue teams satisfying all groups is the total number of rescue teams; Strategy: The strategy of each rescue team refers to the mine accidents it can choose. In this game environment, there are 1+J kinds of mine accidents for participants to choose from. 1 is a central mine accident, and J is a chain mine accident; use x _j to indicate whether the rescue team chooses the mine accident EM _j to unload the task. The actual situation is as follows:

Among them, the optional mine accident set S _k of Q _k is a subset of J; Group share:

Indicates the total number of rescue teams who choose EM _j mine accident rescue in the Q _k population,

Indicates the group share of the whole group choosing EM _j for rescue;

Crowd state: use a vector

To represent the strategy selection state of the group P _k , and satisfy

A matrix P is used to represent the total population state space containing J population states as follows:

Benefits: The benefits of the participants depend on their net utility function, which is determined by their maximum bearable rescue time and cost as well as the actual rescue time and cost; then the rescue team S _i of the Q _k group chooses the value obtained from the EM _j mine accident The benefit can be expressed as follows:

4. The Mine Accident Rescue Task Unloading Method Based on Mobile Edge Computing according to claim 1, characterized in that: said Step3 introduces the replicator dynamic equation into which δ is used to control the convergence speed of participant strategy adaptation in the same group, The design is as follows: In a group of players with bounded rationality, strategies with results better than the average level will gradually be adopted by more players, and the proportion of players' strategies will also change accordingly; the details are as follows:

In the formula: λ—controls the convergence rate of strategy adaptation of participants in the same group;

— the function obtained by first-order derivation with respect to time;

— the current benefits obtained by the participants in the group when they choose the server EM _j ;

—the average income of the group at time t;

is the average income of the group at time t, which can be calculated by the following formula:

Based on the replicator dynamic equation of strategy selection in group Q _k , when the income of choosing strategy u is higher than the average income of the same group, the number of terminal rescue teams who choose this strategy will generally show a positive growth trend; by setting

A fixed point of the replicator dynamic can be obtained, at which the state of the group does not change and no player is willing to change strategy since all players in the same group have the same payoff.

Images