CN117236923A

CN117236923A - Operation and maintenance strategy optimization method under double operation strategies of polymorphic system

Info

Publication number: CN117236923A
Application number: CN202311160997.2A
Authority: CN
Inventors: 赵先; 吴丛珊; 章继文; 柴筱菲; 郑晓亚
Original assignee: Beijing Institute of Technology BIT
Current assignee: Beijing Institute of Technology BIT
Priority date: 2023-09-08
Filing date: 2023-09-08
Publication date: 2023-12-15

Abstract

The invention provides an operation and maintenance strategy optimization method under a double operation strategy of a multi-state system, which comprises the steps of constructing a Markov chain of the multi-state system, determining the task processing amount of each multi-state component when the multi-state system operates at the same load level, and obtaining the state transition rate of each multi-state component according to the task processing amount and the degradation rate; obtaining a state transition rate matrix of the polymorphic system; obtaining a reliability function of the polymorphic system; determining the running cost of the polymorphic system and the total income brought by each completed subtask; obtaining a cost function and a income function of the polymorphic system; the net profit function is obtained, the net profit of the polymorphic system executing different task allocation strategies and part of task abandoning strategies is calculated through the net profit function under different fixed operation time, the optimal net profit under each fixed operation time is obtained, and the task allocation strategy and part of task abandoning strategies corresponding to the optimal net profit under each fixed operation time are output as the optimal operation and maintenance strategy under each fixed operation time.

Description

Operation and maintenance strategy optimization method under double operation strategies of polymorphic system

Technical Field

The invention belongs to the field of operation and maintenance optimization of polymorphic systems, and particularly discloses an operation and maintenance strategy optimization method under a double operation strategy of a polymorphic system.

Description of the background

With the expansion of digital economics, the need to improve the reliability and net profit of related engineering systems has become more and more urgent, redundancy is a common method of improving the performance of any complex engineering system, which can improve the reliability and availability in the system architecture, and in recent years, the k-out-of-n system has been widely used in various fields such as climate, biology, physiological architecture, power generation and transmission systems, computer networks and communication systems. In addition to the architecture of the system, the operational strategy of the system is also important for reliability analysis and to improve the efficiency of the task system. In previous studies, many methods have also been proposed to increase the efficiency of a task system. Task allocation strategies are also a common strategy for improving reliability, and in particular, in recent years, they are becoming more and more widely used in the field of reliability, and in addition, some task relinquishing strategies have also been considered for use in many life-critical or safety-critical systems to improve reliability. However, there is little research on how to improve the net profit of a multi-state system having a plurality of tasks and a plurality of multi-state components, and the existing multi-state system operation and maintenance optimization method cannot more precisely conform to the actual operation and maintenance, so how to improve the net profit of a multi-state system through a task allocation strategy and a partial task relinquishing strategy is an important problem at present. Aiming at the problems, the operation and maintenance strategy optimization method under the double operation strategies of the novel polymorphic system is designed, and the problems existing in the existing polymorphic system research are very necessary to be overcome.

Disclosure of Invention

The invention provides an operation and maintenance strategy optimization method under a double operation strategy of a polymorphic system, which aims to solve the problems that no effective method is available in the prior art to improve the net profits of the polymorphic system with a plurality of tasks and a plurality of polymorphic components and the operation and maintenance optimization method of the existing polymorphic system is not accurate to conform to the actual operation and maintenance.

The invention provides an operation and maintenance strategy optimization method under a double operation strategy of a polymorphic system, which comprises the following steps:

s1, constructing a Markov chain of a polymorphic system, wherein the polymorphic system consists of a plurality of polymorphic components, each polymorphic component has a plurality of states, determining task processing amounts of each polymorphic component in unit time under different stages of the polymorphic system when the polymorphic system operates at the same load level according to historical information of the polymorphic system, determining basic degradation rates of the polymorphic components and the component degradation rates of the polymorphic components caused by the task amounts when the polymorphic system operates at the same load level, and obtaining state transition rates of each polymorphic component under different stages of the polymorphic system according to a task allocation strategy and the task processing amounts, the basic degradation rates and the component degradation rates;

S2, obtaining a state transition rate matrix of the polymorphic system according to the state transition rate of each polymorphic component obtained in the step S1 under different stages of the polymorphic system;

s3, obtaining a reliability function of the polymorphic system according to the state transition rate matrix obtained in the step S2;

s4, determining the running cost of the polymorphic system and the total income brought by each completed subtask;

s5, obtaining a cost function of the polymorphic system according to the operation cost of the polymorphic system obtained in the step S4 and the reliability function obtained in the step S3; according to the total income brought by each completed subtask obtained in the step S4, introducing an indication function to obtain the income function of the polymorphic system;

s6, according to the cost function and the income function of the polymorphic system obtained in the step S5, obtaining a net profit function, calculating the net profit of the polymorphic system for executing different task allocation strategies and part of task abandonment strategies through the net profit function under different fixed operation time, obtaining the optimal net profit of the polymorphic system under each fixed operation time, and outputting the task allocation strategies and part of task abandonment strategies corresponding to the optimal net profit under each fixed operation time as the optimal operation and maintenance strategies of the polymorphic system under each fixed operation time.

In the step S1, the state of each polymorphic component in the polymorphic system is denoted as g= {1,2, …, G }, where 1 indicates that the polymorphic component is in a brand new state, G indicates that the polymorphic component is in a failure state, and numbers between 1-G indicate that the polymorphic component is in an intermediate state from the brand new state to the failure state;

the total task amount L processed in the unit time of the multi-state system is fixed, a plurality of tasks in the multi-state system are divided into m grades according to the importance degree, and the tasks belonging to different grades are defined as m subtasks Z _i (i=1, 2,., m), wherein Z ₁ Z is the most important subtask _m For the least important subtasks, each subtask corresponds to a task quantity L _i (i=1, 2,) m), and

the whole operation of the multi-state system is divided into three stages, including a first stage, a second stage and a third stage;

in the first stage, the multi-state system is in a good state, all multi-state components normally operate, at the moment, all multi-state components process tasks simultaneously, the task quantity is distributed evenly, and the task processing quantity in unit time of each multi-state component is as follows Wherein n is the number of polymorphic components in the polymorphic system;

in the second stage, the state of each polymorphic component in the polymorphic system is reduced to different degrees along with the operation of the polymorphic system, the polymorphic system classifies the polymorphic component by the state of each polymorphic component, the state class of the polymorphic component comprises a low-level state and a high-level state, the polymorphic component is classified into the low-level state component and the high-level state component according to the state class, the dividing line is s, and s epsilon G= {1, 2. Each polymorphic component in the polymorphic system adopts a task allocation strategy, and the task allocation strategy comprises that if the state of any polymorphic component of the polymorphic system is reduced to a low-level state, all tasks are sent to a task balancer of the polymorphic system, the task balancer transfers loads through the state grades of the polymorphic components, task amounts with different proportions are allocated based on the state grades of the polymorphic components, the task amounts with the proportion of 1:alpha and alpha & gt1 are allocated to the low-level state components by the task balancer, and the task amounts with the proportion of alpha are allocated to the high-level state components by the task balancer;

Setting the total number of high-level state components in the multi-state system asSetting the total number of low-level state components in the polymorphic system to +.>Wherein n is _i For the number of components in state i, i=1, 2,..n, after re-allocation of task amounts according to the task allocation policy, the task throughput per unit time per low-level state component is +.>The task processing amount per unit time of each high-level state component is +>

In the third stage, the polymorphic components in the polymorphic system begin to appear as failed components and the number of failed components continues to increase, the polymorphic system employing a partial task relinquishing strategy while employing a task allocation strategy, the partial task relinquishing strategy comprising when a failure occurs in the polymorphic systemWhen the number of components reaches a predetermined threshold, the multi-state system ceases to execute the corresponding subtask, defining a vector k= (K) _m ,k _m-1 ,…,k ₂ ,k ₁ ) An abort threshold value for a partial task abort policy, where k _m ≤k _m-1 ≤…≤k ₂ ≤k ₁ When the number of failed parts reaches k =k _m When the polymorphic system gives up the lowest level subtask Z _m The rest of the polymorphic components continue to process all other subtasks according to the task allocation strategy when the number of failed components reaches a threshold k _j When the polymorphic system gives up subtask Z _j Wherein j e i (i=1, 2,.. M) the task amount of all the abandoned sub-tasks per unit time is added to the remaining task amount of the polymorphic system, which is the total task amount remaining in the polymorphic systemWherein L is _Z For subtask Z _i Is a task amount of (1);

the number of high-level state components in the multi-state system isThe number of low-level state components in the multi-state system is +.>The multi-state system continues to distribute tasks by a task distribution strategy, and the task processing amount per unit time of each low-level state component is +.>The task processing capacity per unit time of each high-level state component is

According to some embodiments of the present application, in the step S1, the operation phase of the polymorphic system includes a first phase, a second phase and a third phase, so as to obtain a state transition rate of each polymorphic component when the polymorphic system is operated in different phases:

the state transition rate when the multi-state component in the first stage operates is shown in the formula (1):

wherein,representing the state transition rate lambda of the polymorphic member when in the first stage of operation _i Represents the basic degradation rate of the polymorphic component from state i to state i+1, λ represents the component degradation rate of the polymorphic component due to the mission quantity;

The state transition rate at the time of the operation of the multi-state component in the second stage is shown in the formula (2):

wherein,representing a state transition rate of the polymorphic component when in the second phase of operation;

the state transition rate at the time of the operation of the multi-state component in the third stage is shown in the formula (3):

wherein,indicating the state transition rate when the multi-state component is in the third stage operation.

Operation and maintenance strategy optimization method under double operation strategies of polymorphic system according to some embodiments of the present applicationIn the step S2, a state transition rate matrix Q of the polymorphic system is obtained _WW (K) As shown in formula (4):

wherein K represents the discard threshold of the partial task discard strategy, X (t) is the state space of the polymorphic system, X (t) = (n) ₁ ,n ₂ ,...,n _g ），q _bc ＝yλ ^* B=1, 2,3 … beta, c=1, 2,3 … beta, y is the number of components in state i in the polymorphic system, y=1, 2, n,beta=n-1, Ω is the set of all states of the polymorphic system, Ω= (X (t)) ₁ ,X(t) ₂ ,...,X(t) _N ) N represents the number of all states of the markov process of the polymorphic system.

In the operation and maintenance policy optimization method under the dual operation policies of the polymorphic system according to some embodiments of the present application, in the step S3, the calculation of the reliability function of the polymorphic system is as described in formula (5):

R(t,K)＝π ₁ exp(Q _WW (K)t)e ^T (5)

Wherein R (t, K) is a reliability function of the polymorphic system, t represents time, K represents a discard threshold of a partial task discard strategy, pi ₁ = (1, 0,) 0 (representing the initial state distribution of the system, e= (1,) 1,0 _1×(N+1) N represents the number of all states of the markov process of the polymorphic system and T represents the transpose of the formula.

In the operation and maintenance strategy optimization method under the double operation strategies of the multi-state system according to some embodiments of the present application, in the step S4, the operation cost of the multi-state system includes the fault cost of the multi-state system and the subsequent cost of each multi-state component generated in different states;

the fault cost c _f Including cost of data loss and information leakageThe cost of the present, restart and maintenance;

the state class of the subsequent cost root polymorphic component is classified into two categories: if the polymorphic component is at a high level, the subsequent cost of a single polymorphic component is c _r If the state of the polymorphic component is at a low level or fails, the subsequent cost of a single polymorphic component is c _p ；

The revenue generated by each subtask is related to the corresponding amount of subtasks, and the total revenue generated by each completed subtask is shown in equation (6):

R _i ＝r _i L _i (6)

Wherein R is _i Total revenue for each completed subtask, r _i To complete income brought by each task of subtasks, L _i Is the amount of subtasks.

In step S5, the probability of each state of the polymorphic system is determined, and the probability P of each state of the polymorphic system after time t is determined according to the markov process _W (t, K) is as shown in formula (7):

P _W (t,K)＝π ₁ exp(Q _WW (K)t) (7)

wherein P is _W (t,K)∈[P ₁ (t,K),P ₂ (t,K),…,P _N-1 (t,K)]，

If the multi-state system is in a fault state, the running cost of the multi-state system is the fault cost c _f All subtasks cannot be fully executed, and the net profit is 0;

if the multi-state system is still in a non-fault state after a fixed time τ, the cost function of the multi-state system is as shown in formula (8):

wherein W (t, K) represents a cost function of the polymorphic system,

for each subtask Z _i I=1, 2, …, m introduces a sexual functionAs shown in formula (9):

wherein A is _i Represents an event, n _i ≥k _i Representing subtask Z _i Is abandoned event, n _i ＜k _i Representing subtask Z _i The event is not discarded and the event is not discarded,

the successful execution rate of each subtask is expressed as the expected probability that each subtask can successfully complete within a fixed time τ, and as such, the expected probability that each subtask successfully completes depends on the abort threshold K of the partial task abort strategy, so the expected probability of success of the subtask May be calculated from the indirection function of each subtask in each state as shown in equation (10):

the revenue function is shown in equation (11):

wherein D (t, K) represents the revenue function of the polymorphic system.

In step S6, the operation time is changed to a fixed operation time τ to obtain a net profit function Max f (τ, K), as shown in formula (12):

Max f(τ,K)＝D(τ,K)-W(τ,K) (12)

the decision variable of the net profit function Max f (tau, K) is the discard threshold K of the partial task discard strategy, when the number of subtasks is small, all possible values of the discard threshold K of the partial task discard strategy are found through a enumeration method, and when the number of subtasks is large, all possible values of the discard threshold K of the partial task discard strategy are found through a genetic algorithm; the method comprises the steps of listing all possible values of a discarding threshold K of all partial task discarding strategies through an enumeration method, respectively calculating net profits of all task allocation strategies and partial task discarding strategies when the polymorphic system is in different fixed operation time, respectively comparing the net profits of the polymorphic system in different fixed operation time to obtain the optimal net profits under each fixed operation time, and outputting a task allocation strategy and a partial task discarding strategy corresponding to the optimal net profits under each fixed operation time as the optimal operation and maintenance strategy of the polymorphic system under each fixed operation time.

The invention also provides an electronic device comprising a memory and a processor, the memory storing a computer program; the processor is configured to execute the computer program in the memory to implement the method described above.

The invention also provides a computer readable storage medium storing a computer program which when executed by a processor implements the above method.

The operation and maintenance strategy optimization method under the double operation strategies of the polymorphic system analyzes the reliability of the polymorphic system under the double operation strategies of the task allocation strategy and the partial task abandonment strategy, models and analyzes degradation rules of the system by using a Markov process through constructing a Markov chain of the polymorphic system, obtains a reliability function, solves the operation random state distribution of the polymorphic system according to the operation cost of the polymorphic system and the total income when each subtask is completed by using the Markov process, calculates the net profit, further obtains the net profit function of the polymorphic system, and obtains the optimal operation and maintenance strategy for balancing the reliability and the net profit of the polymorphic system under different fixed operation times by using the net profit function and an enumeration method, namely the optimal task allocation strategy and the partial task abandonment strategy.

Drawings

FIG. 1 is a flow chart of an operation and maintenance strategy optimization method under a double operation strategy of a polymorphic system according to the present invention;

FIG. 2 is a diagram of the task allocation of a polymorphic component in different states of the polymorphic system of the present invention;

FIG. 3 is a graphical representation of the number of failed components versus total task volume in a multi-state system;

FIG. 4 is a state transition rate diagram of the distributed consulting expert system in embodiment 3 of the present invention;

fig. 5 is a reliability function curve of the distributed consulting expert system under the conditions of α=1, k= [3,2,1] in embodiment 3 of the present invention;

fig. 6 is a diagram of three cases of net profit of the distributed consulting expert system at different fixed run times in example 3 of the present invention.

Detailed Description

Embodiments of the present invention are described in further detail below with reference to the accompanying drawings and examples. The following examples are illustrative of the invention but are not intended to limit the scope of the invention.

An operation and maintenance strategy optimization method under double operation strategies of a polymorphic system, as shown in fig. 1, comprises the following steps:

s1, constructing a Markov chain of a polymorphic system, wherein the polymorphic system consists of a plurality of polymorphic components, each polymorphic component has a plurality of states, determining task processing amounts of each polymorphic component in unit time of the polymorphic system under different stages when the polymorphic system operates at the same load level according to historical information of the polymorphic system, determining basic degradation rates of the polymorphic components and component degradation rates of the polymorphic components caused by the task amounts when the polymorphic system operates at the same load level, and obtaining state transition rates of each polymorphic component under different stages according to a task allocation strategy and the task processing amounts, the basic degradation rates and the component degradation rates;

The state of each polymorphic component in the polymorphic system is represented as g= {1, 2..once, G }, where 1 represents that the polymorphic component is in a brand new state, G represents that the polymorphic component is in a failure state, and numbers between 1-G represent that the polymorphic component is in an intermediate state from the brand new state to the failure state;

the total task quantity L processed in unit time of the multi-state system is fixed, the tasks in the multi-state system are divided into m grades according to the importance degree, and the tasks belonging to different grades are defined as m subtasks Z _i (i=1, 2,., m), wherein Z ₁ Z is the most important subtask _m For the least important subtasks, each subtask corresponds to a task quantity L _i (i=1, 2,) m), and

in the first stage, the multi-state system is in a good state, all multi-state components normally operate, at the moment, all multi-state components process tasks simultaneously, the task quantity is distributed evenly, and the task processing quantity in unit time of each multi-state component is as followsWherein n is the number of polymorphic components in the polymorphic system;

in the second phase, as the state of each polymorphic component in the polymorphic system is lowered to different degrees, as shown in fig. 2, the polymorphic system classifies the polymorphic component by the state of each polymorphic component, the state class of the polymorphic component includes a low-level state and a high-level state, the polymorphic component is classified into the low-level state component and the high-level state component according to the state class, the boundary line is s, and s e g= {1, 2..g }, if the state of the polymorphic component is better than s, the polymorphic component is the high-level state component, and if the state of the polymorphic component is lower than s, but no fault occurs, the polymorphic component is the low-level state component; each polymorphic component in the polymorphic system adopts a task allocation strategy, wherein the task allocation strategy comprises that if the state of any polymorphic component of the polymorphic system is reduced to a low-level state, all tasks are sent to a task balancer of the polymorphic system, the task balancer transfers loads through the state level of the polymorphic component, task amounts with different proportions are allocated based on the state level of the polymorphic component, the task amount proportion is 1:alpha and alpha is more than 1, the task balancer allocates task amounts with the proportion of 1 to the low-level state component, and the task amounts with the proportion of alpha to the high-level state component;

Let the total number of high-level state components in the multi-state system beLet the total number of low-level state components in the multi-state system be +.>Wherein n is _i For the number of components in state i, i=1, 2,..g, after the task amount is redistributed according to the task distribution policy, the task processing amount per unit time of each low-level state component is +>The task processing amount per unit time of each high-level state component is +>

In the third stage, the polymorphic system starts to generate failed components and the number of failed components is continuously increased, and the polymorphic system uses a partial task discard strategy simultaneously with the task allocation strategy, wherein the partial task discard strategy comprises stopping executing corresponding subtasks when the number of failed components in the polymorphic system reaches a predetermined threshold, as shown in fig. 3, defining a vector k= (K = _m ,k _m-1 ,...,k ₂ ,k ₁ ) An abort threshold value for a partial task abort policy, wherek _m ≤k _m-1 ≤…≤k ₂ ≤k ₁ When the number of failed parts reaches k =k _m When the polymorphic system gives up the lowest level subtask Z _m The rest of the polymorphic components continue to process all other subtasks according to the task allocation strategy when the number of failed components reaches a threshold k _j When the polymorphic system gives up subtask Z _j Wherein j e i (i=1, 2,.. M), the task amounts of all of the abandoned sub-tasks are added to the remaining task amounts of the polymorphic system per unit time, and the remaining total task amounts of the polymorphic system are Wherein L is _Z For subtask Z _i Is a task amount of (1);

the number of high-level state components in the multi-state system isThe number of low-level state components in the multi-state system isThe multi-state system continues to distribute tasks by a task distribution strategy, and the task processing capacity of each low-level state component in unit time is +>The task processing amount per unit time of each high-level state component is +>

The multi-state system operation stage comprises a first stage, a second stage and a third stage, and the state transition rate of each multi-state component when the multi-state system is operated in different stages is obtained:

S2, obtaining a state transition rate matrix of the polymorphic system according to the state transition rate of each polymorphic component obtained in the step S1 in different stages of the polymorphic system;

obtaining a state transition rate matrix Q of the polymorphic system _WW (K) As shown in formula (4):

wherein,k represents the discard threshold of the partial task discard strategy, X (t) is the state space of the multi-state system, and X (t) = (n) ₁ ,n ₂ ,…,n _g )，q _bc ＝yλ ^* B=1, 2,3 … beta, c=1, 2,3 … beta, y is the number of components in state i in the polymorphic system, y=1, 2, …, n,beta=n-1, Ω is all state sets of the polymorphic system, Ω= (X (t)) ₁ ,X(t) ₂ ,...,X(t) _N ) N represents the number of all states of the markov process for the polymorphic system.

the reliability function of the polymorphic system is calculated as in equation (5):

R(t,K)＝π ₁ exp(Q _WW (K)t)e ^T (5)

wherein R (t, K) is a reliability function of the polymorphic system, t represents time, K represents a discard threshold of the partial task discard strategy, pi ₁ = (1, 0,..0) represents an initial state distribution of the system, e= (1,..1, 0) _1＝(N+1) N represents the number of all states of the markov process of the polymorphic system and T represents the transpose of the formula.

The running cost of the multi-state system comprises the fault cost of the multi-state system and the subsequent cost of each multi-state component under different states;

failure cost c _f Including data loss costs, information leakage costs, restart costs, and maintenance costs;

the subsequent cost root polymorphic part states are classified into two categories: if the polymorphic component is at a high level, the subsequent cost of a single polymorphic component is c _r If the state of the polymorphic component is at a low level or fails, the subsequent cost of a single polymorphic component is c _p ；

R _i ＝r _i L _i (6)

determining the probability of each state of the multi-state system, and determining the probability P of each state of the multi-state system after time t according to the Markov process _W (t, K) is as shown in formula (7):

P _W (t,K)＝π ₁ exp(Q _WW (K)t) (7)

wherein P is _W (t,K)∈[P ₁ (t,K),P ₂ (t,K),...,P _N-1 (t,K)]，

if the multi-state system is still in a non-fault state after a fixed time τ, the cost function of the multi-state system is shown in formula (8):

wherein W (t, K) represents a cost function of the polymorphic system,

for each subtask Z _i I=1, 2, where, m introducing a sexual functionAs shown in formula (9):

the successful execution rate of each subtask is expressed as the expected probability that each subtask can successfully complete within a fixed time τ, and as such, the expected probability that each subtask successfully completes depends on the abort threshold K of the partial task abort strategy, so the expected probability of success of the subtaskMay be calculated from the indirection function of each subtask in each state as shown in equation (10):

the revenue function is shown in equation (11):

where D (t, K) represents the revenue function of the polymorphic system.

S6, according to the cost function and the income function of the polymorphic system obtained in the step S5, obtaining a net profit function, calculating net profits of the polymorphic system for executing different task allocation strategies and part task abandoning strategies through the net profit function under different fixed operation time, obtaining optimal net profits of the polymorphic system under each fixed operation time, and outputting the task allocation strategies and part task abandoning strategies corresponding to the optimal net profits under each fixed operation time as optimal operation and maintenance strategies of the polymorphic system under each fixed operation time;

Changing the run time to a fixed run time τ, resulting in a net profit function Max f9 τ, K), as shown in equation (12):

Max f(τ,K)＝D(τ,K)-W(τ,K) (12)

the decision variable of the net profit function Max f (tau, K) is the discard threshold K of the partial task discard strategy, when the number of subtasks is small, all possible values of the discard threshold K of the partial task discard strategy are found through a enumeration method, and when the number of subtasks is large, all possible values of the discard threshold K of the partial task discard strategy are found through a genetic algorithm; the method comprises the steps of listing all possible values of a giving-up threshold K of all partial task giving-up strategies through an enumeration method, respectively calculating net profits of all task allocation strategies and the partial task giving-up strategies when the polymorphic system is in different fixed operation times, respectively comparing the net profits of the polymorphic system in different fixed operation times to obtain the optimal net profits under each fixed operation time, and outputting the task allocation strategies and the partial task giving-up strategies corresponding to the optimal net profits under each fixed operation time as the optimal operation and maintenance strategies of the polymorphic system under each fixed operation time.

The embodiment also provides an electronic device, which comprises a memory and a processor, wherein the memory stores a computer program; the processor is configured to execute the computer program in the memory to implement the above method.

The present embodiment also provides a computer readable storage medium storing a computer program which when executed by a processor implements the above method.

Example 2

In this embodiment, taking a computer digital service system as an example, the polymorphic system is a digital service system formed by n computers, specifically, a distributed consultation expert system formed by n computers, and the polymorphic component is a computer, and the operation and maintenance policy optimization method under the dual operation policy includes the following steps:

s1, constructing a Markov chain of a distributed consultation expert system, wherein the distributed consultation expert system consists of a plurality of computers, each computer has a plurality of states, the task processing amount of each computer in unit time of the distributed consultation expert system under different stages when the distributed consultation expert system operates at the same load level is determined according to the historical information of the distributed consultation expert system, the basic degradation rate of the computer and the component degradation rate of the computer caused by the task amount when the distributed consultation expert system operates at the same load level are determined, and the state transition rate of each computer under different stages of the distributed consultation expert system is obtained according to a task allocation strategy and the task processing amount, the basic degradation rate and the component degradation rate;

The state of each computer in the distributed consultation expert system is expressed as G= {1,2, …, G }, wherein 1 indicates that the computer is in a brand new state, G indicates that the computer is in a failure state, and numbers between 1 and G indicate that the computer is in an intermediate state from the brand new state to the failure state;

the total task amount L processed in the unit time of the distributed consultation expert system is fixed, a plurality of tasks in the distributed consultation expert system are divided into m grades according to the importance degree, and the tasks belonging to different grades are defined as m subtasks Z _i (i=1, 2, …, m), wherein Z ₁ Z is the most important subtask _m For the least important subtasks, each subtask corresponds to a task quantity L _i (i=1, 2, …, m), and

the whole operation of the distributed consultation expert system is divided into three stages, including a first stage, a second stage and a third stage;

in the first stage, the distributed consultation expert system is in a good state, all computers normally operate, at the moment, all computers process tasks simultaneously, the task quantity is distributed evenly, and the task processing quantity in unit time of each computer is as followsWherein n is the number of computers in the distributed consultation expert system;

in the second stage, along with the operation of the distributed consultation expert system, the state of each computer in the distributed consultation expert system is reduced to different degrees, as shown in fig. 2, the distributed consultation expert system classifies the computer by the state of each computer, the state level of the computer comprises a low-level state and a high-level state, the computer is divided into a low-level state component and a high-level state component according to the state level, the dividing line is s, s epsilon G= {1,2, …, G }, if the state of the computer is better than s, the computer is a high-level state component, and if the state of the computer is lower than s, but no fault occurs, the computer is a low-level state component; each computer in the distributed consultation expert system adopts a task allocation strategy, wherein the task allocation strategy comprises that if the state of any computer in the distributed consultation expert system is reduced to a low-level state, all tasks are sent to a task balancer of the distributed consultation expert system, the task balancer transfers loads through the state grades of the computers, task quantities with different proportions are allocated based on the state grades of the computers, the task quantity proportion is 1:alpha and alpha is more than 1, the task balancer allocates the task quantity with the proportion of 1 to a low-level state component, and allocates the task quantity with the proportion of alpha to a high-level state component;

Setting the total number of high-level state components in the distributed consultation expert system asSetting the total number of low-level state components in the distributed consultation expert system as +.>Wherein n is _i For the number of components in state i, i=1, 2,..g, after the task amount is redistributed according to the task distribution policy, the task processing amount per unit time of each low-level state component is +>The task processing amount per unit time of each high-level state component is +>

In the third stage, the computers in the distributed consultation expert system begin to fail, and the number of failed computers is continuously increasing, the distributed consultation expert system uses a partial task discarding strategy while using a task allocation strategy, and the partial task discarding strategy includes stopping the distributed consultation expert system from executing the corresponding subtasks when the number of failed computers in the distributed consultation expert system reaches a predetermined threshold, as shown in fig. 3, defining a vector k= (K) _m ,k _m-1 ,...,k ₂ ,k ₁ ) An abort threshold value for a partial task abort policy, where k _m ≤k _m-1 ≤…≤k ₂ ≤k ₁ When the number of failed parts reaches k =k _m When the distributed consulting expert system gives up the lowest level subtask Z _m The other computers continue to process all other subtasks according to the task allocation strategy when the number of failed components reaches the threshold k _j At this time, the distributed consulting expert system gives up subtask Z _j Wherein j e i (i=1, 2,.., m), the task amounts of all of the discarded subtasks are added to the remaining task amounts of the distributed consultation expert system per unit time, and the remaining total task amounts of the distributed consultation expert system areWherein L is _Z For subtask Z _i Is a task amount of (1);

the number of high-level status components in the distributed consultation expert system isThe number of low-level status parts in the distributed consultation expert system is +.>The distributed consultation expert system continues to distribute tasks by a task distribution strategy, and the task processing amount per unit time of each low-level state component is +.>The task processing amount per unit time of each high-level state component is +>

The operation stage of the distributed consultation expert system comprises a first stage, a second stage and a third stage, and the state transition rate of each computer when the distributed consultation expert system operates in different stages is obtained:

the state transition rate of the computer in the first stage when running is as shown in formula (13):

wherein,representing the state transition rate, lambda, of a computer when the computer is in a first phase of operation _i Representing the basic degradation rate of the computer from state i to state i+1, λ representing the component degradation rate of the computer due to the task amount;

The state transition rate of the computer in the second stage when running is as shown in formula (14):

wherein,representing the state transition rate of the computer in the second stage of operation;

the state transition rate of the computer in the third stage when running is as shown in formula (15):

wherein,representing the state transition rate of the computer when it is in the third phase of operation.

S2, obtaining a state transition rate matrix of the distributed consultation expert system according to the state transition rate of each computer obtained in the step S1 in different stages of the distributed consultation expert system;

obtaining a state transition rate matrix Q of the distributed consultation expert system _WW (K) As shown in equation (16):

wherein K represents the discard threshold of the partial task discard strategy, X (t) is the state space of the distributed consulting expert system, X (t) = (n) ₁ ,n ₂ ,...,n _g ），q _bc ＝yλ ^* B=1, 2,3 … β, c=1, 2,3 … β, y is the number of components in state i in the distributed consultation expert system, y=1, 2, n,beta=n-1, Ω is all state sets of the distributed consulting expert system, Ω= (X (t)) ₁ ,X(t) ₂ ,...,X(t) _N ) N represents the number of all states of the markov process of the distributed consulting expert system.

S3, obtaining a reliability function of the distributed consultation expert system according to the state transition rate matrix obtained in the step S2;

The reliability function of the distributed consulting expert system is calculated as in equation (17):

R(t,K)＝π ₁ exp(Q _WW (K)t)e ^T (17)

wherein R (t, K) is the reliability function of the distributed consultation expert system, t represents time, K represents the discard threshold of the partial task discard strategy, pi ₁ = (1, 0, …, 0) represents the initial state of the systemDistribution, e= (1,., 1, 0) _1×(N+1) N represents the number of all states of the markov process of the distributed consulting expert system and T represents the transpose of the formula.

S4, determining the running cost of the distributed consultation expert system and the total income brought by each completed subtask;

the running cost of the distributed consultation expert system comprises the fault cost of the distributed consultation expert system and the subsequent cost generated by each computer in different states;

the state class of the root computer is classified into two classes: if the computers are in a high-level state, the subsequent cost of a single computer is c _r If the state of the computer is in a low level state or fails, the subsequent cost of the single computer is c _p ；

The revenue generated by each subtask is related to the corresponding amount of subtasks, and the total revenue generated by each completed subtask is shown in equation (18):

R _i ＝r _i L _i (18)

S5, obtaining a cost function of the distributed consultation expert system according to the operation cost of the distributed consultation expert system obtained in the step S4 and the reliability function obtained in the step S3; according to the total income brought by each completed subtask obtained in the step S4, introducing an indication function to obtain the income function of the distributed consultation expert system;

determining the probability of each state of the distributed consultation expert system, and determining the probability P of each state of the distributed consultation expert system after time t according to the Markov process _W (t, K) is as shown in formula (19):

P _W (t,K)＝π ₁ exp(Q _WW (K)t) (19)

wherein P is _W (t,K)∈[P ₁ (t,K),P ₂ (t,K),...,P _N-1 (t,K)]，

If the distributed consultation expert system is in a fault state, the running cost of the distributed consultation expert system is the fault cost c _f All subtasks cannot be fully executed, and the net profit is 0;

if the distributed consulting expert system is still in a non-fault state after a fixed time τ, the cost function of the distributed consulting expert system is shown in formula (20):

wherein W (t, K) represents a cost function of the distributed consulting expert system,

For each subtask Z _i I=1, 2, …, m introduces a sexual functionAs shown in formula (21):

the successful execution rate of each subtask is expressed as the expected probability that each subtask can successfully complete within a fixed time τ, and as such, the expected probability that each subtask successfully completes depends on the abort threshold K of the partial task abort strategy, so the expected probability of success of the subtaskMay be calculated from the indirection function of each subtask in each state as shown in equation (22):

the revenue function is shown in equation (23):

where D (t, K) represents the revenue function of the distributed consulting expert system.

S6, according to the cost function and the income function of the distributed consultation expert system obtained in the step S5, obtaining a net profit function, calculating net profits of different task allocation strategies and partial task abandon strategies by the distributed consultation expert system under different fixed operation time, obtaining optimal net profits of the distributed consultation expert system under each fixed operation time, and outputting the task allocation strategies and the partial task abandon strategies corresponding to the optimal net profits under each fixed operation time as optimal operation and maintenance strategies of the distributed consultation expert system under each fixed operation time;

Changing the run time to a fixed run time τ yields a net profit function Max f (τ, K) as shown in equation (24):

Max f(τ,K)＝D(τ,K)-W(τ,K) (24)

the decision variable of the net profit function Max f (tau, K) is the discard threshold K of the partial task discard strategy, when the number of subtasks is small, all possible values of the discard threshold K of the partial task discard strategy are found through a enumeration method, and when the number of subtasks is large, all possible values of the discard threshold K of the partial task discard strategy are found through a genetic algorithm; the method comprises the steps of listing all possible values of a discarding threshold K of all partial task discarding strategies through an enumeration method, respectively calculating net profits of all task allocation strategies and partial task discarding strategies when a distributed consultation expert system is in different fixed operation times, respectively comparing the net profits of the distributed consultation expert system in different fixed operation times to obtain optimal net profits under each fixed operation time, and outputting a task allocation strategy and a partial task discarding strategy corresponding to the optimal net profits under each fixed operation time as an optimal operation and maintenance strategy of the distributed consultation expert system under each fixed operation time.

Example 3

In this embodiment, the computer digital service system is taken as an example, the polymorphic system is a digital service system formed by three computers, and is supported by a plurality of computers with the same performance, when the computers fail, the polymorphic system distributes tasks to the remaining non-failed computers, because the load of the computers is too large, the polymorphic system can split the tasks, the task balancer distributes the tasks to the remaining different computers according to the task distribution policy, so as to realize unequal task distribution, and as the number of the computers increases, the polymorphic system will discard part of the tasks to reduce the failure rate of each computer, discard the tasks such as using static pages, limiting domain name repeated access, shortening automatic logout time, rejecting excessive flow requests, and the like. When the number of failures of computers reaches a certain number, each non-failed computer is assigned a task that exceeds the maximum allowable load, which will result in a crash of the digital service system, i.e., a system failure.

The method mainly uses a Markov process method, models and analyzes the reliability of the polymorphic system according to the degradation rate and the double operation strategies of the polymorphic system, and searches for the optimal double operation strategies, namely a task allocation strategy and a partial task abandon strategy, and the specific method of the embodiment comprises the following steps:

determining a composition structure, a task system architecture and a task allocation strategy and a partial task abandon strategy of the polymorphic system, namely the digital service system;

the digital service system of the embodiment is specifically a 2/3:F distributed consultation expert system which executes tasks within a fixed time, namely the distributed consultation expert system is composed of three computers, namely the computers are polymorphic components, the three computers run simultaneously, and when two of the computers fail, the distributed consultation expert system fails. Each computer in the distributed consultation expert system has the same performance, three computers are in three different states, and the state of each computer in the distributed consultation expert system is G= {1,2,3}, wherein '1, 2' indicates that the computer is still in a working state, and '3' indicates that the computer is in a failure state and is a failure computer. The division threshold value of the high-level state and the low-level state of the computer is 2, that is, the computer in the state 1 is the high-level state component, and the computer in the state 2 is the low-level state component.

The total task amount L of the distributed consultation expert system is 6, and can be divided into three subtasks according to importance degree, namely: expert online communication service, online data submission and intelligent query service. The task amount corresponding to the expert online communication service subtask is 3, the task amount corresponding to the online data submitting subtask is 2, and the task amount corresponding to the intelligent query service subtask is 1. Each unit time a task is dynamically arrived at and is considered to be successfully executed if and only if the subtask is still in execution for a fixed period of time.

Basic degradation rate lambda of polymorphic components corresponding to distributed advisory expert system states from state 1 to state 2 ₁ A basic degradation rate lambda of 0.05 for the multi-state component from state 2 to state 3 ₂ The component degradation rate λ due to the task amount was 0.10. And in the running process of the distributed consultation expert system, the task quantity of each computer in different states is calculated according to the task allocation strategy and the partial task abandonment strategy.

Analyzing the random state distribution of the whole distributed consultation expert system in the operation environment, and describing the state transition rule of the random state distribution;

determining a state space X (t) = (n) of a distributed consulting expert system ₁ ,n ₂ ,...,n _g )。n _i Representing the number of components in state i. For the distributed consulting expert system of this embodiment, the state space of the distributed consulting expert system can be expressed as:

{(3,0,0),(2,1,0),(1,2,0),(0,3,0),(2,0,1),(1,1,1),(0,2,1)}∪{E}。

the state transition rate of each computer is composed of two parts, one is the basic degradation rate of the polymorphic component being operated and one is the component degradation rate due to the amount of tasks. For example, in the first stage, the task amounts are distributed equally, so the state transition rate due to the tasks is also uniform for each computer, so the state transition rate when the computers in the first stage are running is as shown in the formula (25):

the state transition rate of the computer in the second stage when running is shown in the formula (26):

the state transition rate of the computer in the third stage when running is as shown in formula (27):

and drawing a state transition rate diagram and writing a state transition rate matrix of the distributed consultation expert system. In this embodiment, fig. 4 is a state transition diagram of the distributed consulting expert system, and the state transition rate matrix is shown in formula (28):

wherein,q ₂₂ ＝-2λ ₁ -λ ₂ -λ，/> q ₃₃ ＝-λ ₁ -2λ ₂ -λ，/>

based on the above, the reliability distribution of the distributed consulting expert system is calculated, and the reliability function expression of the distributed consulting expert system is written, as shown in formula (29):

R(t,K)＝π ₁ exp(Q _WW (K)t)e ^T (29)

Fig. 5 is a reliability function curve of the distributed consultation expert system under the conditions of α=1 and k= [3,2,1], in fig. 5, when the running time is 0, the system reliability is 1, but the reliability is reduced to different degrees with the increase of time, and the basic trend of the reduction is that the reliability is reduced slowly before quickly.

The cost and income of the distributed consultation expert system are calculated, firstly, three types of cost including fault cost, high-level state follow-up cost and low-level state follow-up cost are generated in the operation process of the system, wherein the fault cost is 3.5, the high-level state follow-up cost is 0.2 and the low-level state follow-up cost is 0.6. In addition, the expert communicates the unit profit r corresponding to the service subtask online ₁ 1.2, the unit profit r corresponding to the on-line data submitting subtask ₂ 1, the unit profit r corresponding to the intelligent inquiry service subtask ₃ 0.8.

And comparing the optimal net profits of the distributed consultation expert system under different part task abandoning strategies to obtain the maximum optimal net profits. For comparison of policies, the present embodiment enumerates all partial task relinquishing policies according to a distributed consulting expert system model, as shown in Table 1, where k ₁ 、k ₂ And k ₃ Representing a discard threshold. The net profit value for all schemes at different fixed run times τ is also calculated as shown in fig. 6.

TABLE 1 multiple parameter valuation under partial task relinquishing policy

The optimal net profit function is shown in equation (30):

Max f(τ,K)＝D(τ,K)-W(τ,K) (30)

wherein,

the comparison of the optimal net profit for the distributed consultation expert system can be seen to be [0, 15 ] by plotting the optimal net profit function curves for different strategies as shown in FIG. 6]The time interval of the distributed consultation expert system is divided into three stages, when the running time of the distributed consultation expert system is short, the fault rate is very low, and the states of all computers of the distributed consultation expert system are very good, so that the net profit of the distributed consultation expert system is highest when tasks are not abandoned; however, when the fixed run time of the distributed consulting expert system increases to some extent, the corresponding net profit for a solution that does not implement part of the task relinquishing strategy is minimal, the longer the time, the greater the gap, and the other two solutions k ₁ ＝k ₂ ＝2,k ₃ =1 and k ₁ ＝2,k ₂ ＝k ₃ The net profit of =1 reaches a maximum at different fixed run times, respectively. Assuming τ=8, it can be seen by visual numerical comparison that when scheme k is taken ₁ ＝k ₂ ＝2,k ₃ When=1, the net profit of the distributed consultation expert system is maximum 2.5722, and scheme k is selected ₁ ＝k ₂ ＝k ₃ When=2, the net profit is 2.4132, which is three schemesThis illustrates that the net profit is less when the distributed consulting expert system is not doing task abandonment, which also demonstrates the effectiveness of a partial task abandonment strategy.

Based on the above embodiments, the present application further provides an electronic device, including: one or more processors, memory, and one or more programs; wherein one or more programs are stored in the memory, the one or more programs comprising instructions, which when executed by the electronic device, cause the electronic device to perform the methods provided by the above embodiments.

Based on the above embodiments, the present application also provides a computer storage medium having a computer program stored therein, which when executed by a computer, causes the computer to perform the method provided in the above embodiments.

Wherein a storage medium may be any available medium that can be accessed by a computer. Taking this as an example but not limited to: the computer readable medium may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The embodiments of the invention have been presented for purposes of illustration and description, and are not intended to be exhaustive or limited to the invention in the form disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, and to enable others of ordinary skill in the art to understand the invention for various embodiments with various modifications as are suited to the particular use contemplated.

Claims

1. An operation and maintenance strategy optimization method under a double operation strategy of a polymorphic system is characterized by comprising the following steps:

2. The method of optimizing operation and maintenance strategies under a dual operation strategy of a polymorphic system according to claim 1, wherein in said step S1, the state of each polymorphic component in said polymorphic system is denoted as g= {1, 2..g }, wherein 1 denotes that the polymorphic component is in a completely new state, G denotes that the polymorphic component is in a failure state, and numbers between 1-G denote that the polymorphic component is in an intermediate state from the completely new state to the failure state;

Setting the total number of high-level state components in the multi-state system asSetting the total number of low-level state components in the polymorphic system to +.>Wherein n is _i For the number of components in state i, i=1, 2, …, g, after the task amount is redistributed according to the task distribution policy, the task processing amount per unit time of each low-level state component is +>The task processing amount per unit time of each high-level state component is +>

In the third stage, the polymorphic system starts to generate failed components and the number of failed components is continuously increased, and the polymorphic system uses a partial task discard strategy simultaneously with a task allocation strategy, wherein the partial task discard strategy comprises stopping executing corresponding subtasks when the number of failed components in the polymorphic system reaches a predetermined threshold value, and defining a vector K= (K) _m ,k _m-1 ,…,k ₂ ,k ₁ ) An abort threshold value for a partial task abort policy, where k _m ≤k _m-1 ≤…≤k ₂ ≤k ₁ When the number of failed parts reaches k =k _m When the polymorphic system gives up the lowest level subtask Z _m The rest of the polymorphic components continue to process all other subtasks according to the task allocation strategy when the number of failed components reaches a threshold k _j When the polymorphic system gives up subtask Z _j Wherein j e i (i=1, 2,.. M) the task amount of all the abandoned sub-tasks per unit time is added to the remaining task amount of the polymorphic system, which is the total task amount remaining in the polymorphic systemWherein L is _Z For subtask Z _i Is a task amount of (1);

the number of high-level state components in the multi-state system isThe number of low-level state components in the multi-state system isThe multi-state system continues to distribute tasks by a task distribution strategy, and the task processing amount per unit time of each low-level state component is +.>The task processing capacity per unit time of each high-level state component is

3. The method for optimizing operation and maintenance strategies under a dual operation strategy of a multi-state system according to claim 2, wherein in the step S1, the operation phases of the multi-state system include a first phase, a second phase and a third phase, so as to obtain a state transition rate of each multi-state component when the multi-state system is operated under different phases:

4. The method for optimizing operation and maintenance strategies under a dual operation strategy of a multi-state system according to claim 3, wherein in said step S2, a state transition rate matrix Q of said multi-state system is obtained _WW (K) As shown in formula (4):

wherein K represents the discard threshold of the partial task discard strategy, X (t) is the state space of the polymorphic system, X (t) = (n) ₁ ,n ₂ ,…,n _g )，q _bc ＝yλ ^* B=1, 2,3 … beta, c=1, 2,3 … beta, y is the number of components in state i in the polymorphic system, y=1, 2, …, n,beta=n-1, Ω is the set of all states of the polymorphic system, Ω= (X (t)) ₁ ,X(t) ₂ ,...,X(t) _N ) N represents the number of all states of the markov process of the polymorphic system.

5. The method for optimizing operation and maintenance strategies under a dual operation strategy of a multi-state system according to claim 4, wherein in the step S3, the reliability function of the multi-state system is calculated as shown in the formula (5):

R(t,K)＝π ₁ exp(Q _WW (K)t)e ^T (5)

Wherein R (t, K) is a reliability function of the polymorphic system, t represents time, K represents a discard threshold of a partial task discard strategy, pi ₁ = (1, 0, …, 0) represents an initial state distribution of the system, e= (1,..1, 0) _1×(N+1) N represents the number of all states of the markov process of the polymorphic system and T represents the transpose of the formula.

6. The method for optimizing operation and maintenance strategies under a dual operation strategy of a multi-state system according to claim 5, wherein in said step S4, the operation cost of said multi-state system includes the failure cost of said multi-state system, the subsequent cost of each multi-state component generated in different states;

the fault cost c _f Including data loss costs, information leakage costs, restart costs, and maintenance costs;

R _i ＝r _i L _i (6)

7. The method for optimizing operation and maintenance strategies under a dual operation strategy of a multi-state system according to claim 6, wherein in said step S5, the probability of each state of said multi-state system is determined, and the probability P of each state of said multi-state system after time t is determined according to a markov process _W (t, K) is as shown in formula (7):

P _W (t,K)＝π ₁ exp(Q _WW (K)t) (7)

wherein P is _W (t,K)∈[P ₁ (t,K),P ₂ (t,K),…,P _N-1 (t,K)]，

wherein W (t, K) represents a cost function of the polymorphic system,

the revenue function is shown in equation (11):

wherein D (t, K) represents the revenue function of the polymorphic system.

8. The method for optimizing operation and maintenance strategies under a dual operation strategy of a multi-state system according to claim 7, wherein in the step S6, the operation time is changed to a fixed operation time τ, and a net profit function Max f (τ, K) is obtained, as shown in the formula (12):

Max f(τ,K)＝D(τ,K)-W(τ,K) (12)

9. An electronic device comprising a memory and a processor, the memory storing a computer program; the processor is configured to execute the computer program in the memory to implement the method of any one of claims 1 to 8.

10. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program which, when executed by a processor, implements the method according to any one of claims 1-8.