A task reliability modeling and analysis method based on a dynamic reconfiguration system

Abstract

The application provides a task reliability modeling and analysis method based on a dynamic reconfiguration system, comprising: using resource operation capability support vectors to model elements of the dynamic reconfiguration system, constructing a resource vector group and a task demand matrix, and obtaining an initial demand constraint model of the system; when a relationship between the resource vector group and the task demand matrix changes, performing change analysis through mathematical modeling, judging existence and uniqueness of a capability-demand constraint solution, and generating a new resource configuration scheme; and then calculating resource residual capability and functional residual capability, combining reliability indexes, evaluating actual residual capability of the system, and completing reliability calculation and analysis. The application can effectively depict behaviors of the system in a dynamic reconfiguration process, clearly define selection and optimization targets in the reconfiguration process, provide a quantitative basis for reconfiguration decisions of the system, and can also calculate resource residual capability and functional residual capability, ensuring comprehensiveness and accuracy of reliability analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of reliability modeling and analysis, in particular to a task reliability modeling and analysis method based on a dynamic reconfiguration system. BACKGROUND

[0002] With the rapid development of information technology, traditional system architecture is difficult to cope with complex and dynamic environment requirements. Dynamic reconfiguration system architecture emerges as the times require. For example, a ship electronic information system is a kind of dynamic reconfiguration system, which needs to provide high reliable information support and decision support services in a dynamic environment. The main functions of the ship electronic information system include navigation, communication, radar detection, electronic countermeasure, etc., each function is supported by multiple sub-function modules. For example, the navigation task can be divided into route planning, position correction, chart updating, etc. The communication task can include data transmission, signal relay, interference avoidance, etc. The implementation of the function module depends on the underlying resource operation capability, such as radar equipment, communication terminal, computing node, power supply, etc.

[0003] However, the existing dynamic reconfiguration system still has deficiencies in many key links. On the one hand, in the reconfiguration decision process, the current method often lacks in-depth analysis of the changes of resource capability and task demand, and it is difficult to accurately predict the trend and degree of change in the dynamic environment. This leads to the generation of reconfiguration scheme too dependent on simple rules or fixed patterns, limiting the adaptability and flexibility of the system in complex scenarios. On the other hand, in the resource allocation process, the existing technology lacks global optimization analysis between task demand and resource capability, especially when facing complex systems, it cannot effectively balance the efficiency and cost of resource use, often leading to resource allocation redundancy or deficiency, affecting the overall effect of system dynamic reconfiguration.

[0004] In addition, for the reliability evaluation after system reconfiguration, most of the existing methods fail to fully consider the dynamic relationship between task demand and residual resource capability, lack of systematic indicators to measure resource and functional residual capacity, and thus it is difficult to fully reflect the actual reliability level of the system. Especially in the case of resource shortage or demand mutation, the reconfiguration scheme of the system lacks optimization strategy, which may further reduce the running stability of the system. SUMMARY

[0005] In order to overcome the deficiencies of the prior art, the purpose of the present application is to provide a task reliability modeling and analysis method based on a dynamic reconfiguration system, which can effectively depict the behavior of the system in the dynamic reconfiguration process, clearly define the selection and optimization target in the reconfiguration process, provide a quantitative basis for the reconfiguration decision of the system, and also calculate the resource residual capability and functional residual capability, ensuring the comprehensiveness and accuracy of the reliability analysis.

[0006] To achieve the above object, the application provides the following scheme: a task reliability modeling and analysis method based on a dynamic reconfiguration system, comprising:

[0007] Modeling tasks, sub-tasks, functions and resources in the dynamic reconfiguration system by using resource operation capability support vectors, constructing a resource vector group and a task demand matrix, and obtaining an initial demand constraint model of the system;

[0008] When the relationship between the resource vector group and the task demand matrix changes, change analysis is performed through mathematical modeling, the existence and uniqueness of the capability-demand constraint solution are judged, a new resource configuration scheme is generated, and system reconfiguration is completed; when the capability-demand constraint solution has a unique solution, the new resource configuration scheme is generated, when the capability-demand constraint solution has infinite solutions, the objective function and constraint condition are introduced into the system reconfiguration mathematical model, and the optimal solution is calculated, and when the capability-demand constraint solution has no solution, the system resources or demand are adjusted, the new resource vector group and the task demand matrix are constructed;

[0009] After the dynamic reconfiguration of the system is completed, the actual residual capability of the system is evaluated by calculating the residual capability of the resources and the residual capability of the functions in combination with the reliability index, and the reliability calculation and analysis of the dynamic reconfiguration system are completed.

[0010] Optionally, the tasks, sub-tasks, functions and resources in the dynamic reconfiguration system are modeled by using resource operation capability support vectors, comprising: modeling and describing the relationship between the tasks, sub-tasks, functions and resources by using the resource operation capability support vectors through a layer-by-layer progressive manner, and obtaining the logical association between the tasks and the resources; wherein the task is composed of multiple sub-tasks, the sub-task is supported by multiple functions, and the function is supported by multiple resources.

[0011] Optionally, the resource vector group and the task demand matrix are constructed to obtain the initial demand constraint model of the system, comprising:

[0012] According to the logical association between the tasks and the resources, the capability vector, the function demand support set and the resource support set are used to construct the task demand matrix;

[0013] The task demand matrix is mapped and compared with the resource vector group, and the task demand matrix is decomposed layer by layer through the linear representation relationship between the system elements, so that the system task demand is mapped to the function demand and the resource support step by step.

[0014] Optionally, when the relationship between the resource vector group and the task demand matrix changes, change analysis is performed through mathematical modeling, the existence and uniqueness of the capability-demand constraint solution are judged, a new resource configuration scheme is generated, and system reconfiguration is completed, comprising:

[0015] For the task demand matrix, a current resource vector group for satisfying the current task demand is obtained by solving a configuration vector and analyzing the capability-demand constraint solution, and a system reconstruction mathematical model is constructed according to the current resource vector group to find a new resource configuration scheme;

[0016] When the capability-demand constraint solution has a unique solution, the current resource vector group satisfies the current task, and the configuration vector gives a resource configuration scheme;

[0017] When the capability-demand constraint solution has infinite solutions, the configuration vector is optimized, a plurality of linearly independent resource vector groups are found in the resource vector group, and an optimal resource vector group for satisfying the current task demand is selected;

[0018] When the capability-demand constraint solution has no solution, the current resource vector group cannot satisfy the current task, and a new resource vector group and task demand matrix are constructed by adjusting the system resource or system demand.

[0019] Optionally, when the capability-demand constraint solution has infinite solutions and there is no optimal resource vector group satisfying the current task demand, in the system reconstruction mathematical model, each linearly independent resource vector group generates a resource vector space, and in the resource vector space, the total resource cost of each resource vector group is calculated according to the use cost of each resource; the calculation formula of the total resource cost is:

[0020]

[0021] Wherein, f i (λ i ) is a cost function corresponding to the resource vector group, λ i is a linear combination coefficient of the resource vector group;

[0022] The objective function and the constraint condition are introduced into the total resource cost, and the minimum total resource cost value is selected as the optimal solution of the system dynamic reconstruction.

[0023] Optionally, the formula of the objective function is:

[0024]

[0025] The formula of the constraint condition is:

[0026] B·X≥d c

[0027] X≤N

[0028] Wherein, d c is a task demand, N is a demand resource scale vector, and X is a hierarchical configuration facing demand.

[0029] Optionally, after the system dynamic reconstruction is completed, the actual remaining capability of the system is obtained by calculating the remaining capability of the resources and the remaining capability of the functions, and the reliability calculation and analysis of the dynamically reconstructed system are completed, including:

[0030] In the capability vector model of the resources, a norm for measuring the synthesized capability vector of the resources is introduced, and a weight vector for measuring the importance of each synthesized capability vector of the resources is set, and then the total capability level of the system is defined by using the norm and the weight vector; the definition formula of the total capability level of the system is:

[0031]

[0032] wherein Cs is the total resource cost, c is the norm, and w is the weight vector;

[0033] The support relationship of the resources to the tasks is calculated to obtain the remaining capability and energy level of the resources and the remaining capability and energy level of the functions; the expression of the support relationship is:

[0034]

[0035] wherein R is a resource vector group, X and Y are the hierarchical configurations facing the demand, is the total resource demand vector of the functions, is the total resource demand vector of the tasks;

[0036] The remaining capability and energy level of the resources are obtained by comparing X with the scale vector of the resources, and the remaining capability and energy level of the functions are obtained by comparing Y with the scale vector of the functions.

[0037] Optionally, after the system dynamic reconstruction is completed, the actual remaining capability of the system is obtained by calculating the remaining capability of the resources and the remaining capability of the functions, and the reliability calculation and analysis of the dynamically reconstructed system are completed, and further including:

[0038] The difference between the capability of the resources and the expectation is calculated to measure the effectiveness of the resources; the expression of the effectiveness of the resources is:

[0039]

[0040] wherein ASR(i) is the effectiveness of the resources, δ r (e r ) is the mapping function of the resource defects;

[0041] The output function set is used to measure the effectiveness of the functions; the expression of the effectiveness of the functions is:

[0042]

[0043] FI = {F d1 ,F d2 ,…,F dm}

[0044] F di = (fd1, fd2, …, fd n )

[0045] F O = {F c1 ,F c2 ,…,F cm}

[0046] F ci = (fc1, fc2, …, fc n )

[0047] Wherein, A F is the function effectiveness, F O is the existing output function set of the system, F I is the input function set, and F ci is the function.

[0048] The actual remaining capacity of the system is calculated by using a multiplication calculation method, and the expression of the actual remaining capacity is as follows:

[0049]

[0050] Wherein, A T is the actual remaining capacity, A R is the resource effectiveness, and A F is the function effectiveness.

[0051] The application provides a task reliability modeling and analysis method based on a dynamic reconstruction system, and the following technical effects are disclosed:

[0052] 1. The application effectively describes the dynamic reconstruction behavior of the system by constructing a resource capacity and task demand equation, which not only provides a quantitative basis for the reconstruction decision of the system, but also allows real-time adjustment in actual operation to optimize the efficiency of resource allocation and task execution, and enables the system to maintain efficient operation in a changing environment and has stronger adaptability.

[0053] 2. The application provides a new method for comprehensively evaluating the task reliability of the system by calculating the resource remaining capacity and the function remaining capacity, which not only considers the influence of a single resource or function, but also comprehensively reflects the interaction between the two, ensuring the comprehensiveness and accuracy of the reliability analysis, and this feature makes the method applicable to various complex systems and can effectively support the formulation of key decisions.

[0054] 3. This invention, through the dynamic reconfiguration behavior of the system, fully considers the interrelationships between resources, functions, and tasks, establishing a comprehensive modeling and analysis framework. This approach enables a more accurate assessment of the task reliability of reconfigurable systems in dynamic environments, providing effective theoretical support and practical guidance for related applications.

[0055] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 This is a schematic diagram of the method flow provided in an embodiment of the present invention;

[0058] Figure 2 A flowchart illustrating the dynamic reconfiguration and reliability assessment provided in this embodiment of the invention;

[0059] Figure 3 This is a schematic diagram of the requirements structure of a dynamic reconfiguration system provided in an embodiment of the present invention. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] like Figure 1 As shown, this invention provides a method for task reliability modeling and analysis based on a dynamically reconfigurable system, including:

[0063] 1. Utilizing resource operation capability support vectors, model the tasks, subtasks, functions, and resources in the dynamically reconfigurable system, constructing resource vector sets and task requirement matrices to obtain the initial requirement constraint model of the system. This includes:

[0064] 1.1 Mathematical Modeling of Reconfigurable Systems

[0065] like Figures 2-3 As shown, in the effective organization of a hierarchical, dynamically reconfigurable system, there are three main determining factors: resources, functions, and tasks. Formalizing these three factors is a prerequisite for measuring the reliability of system tasks. The resource integration platform contains a large number of public resources and some dedicated resources, which are connected and communicate via a communication bus, such as the integrated core processor, sensors, and communication equipment in an electronic system. The function integration platform uniformly organizes and manages the system's dedicated functions, logical functions, organizational functions, and application functions, which can be categorized according to their impact on system tasks as critical functions, important functions, basic functions, and auxiliary functions. The system management integration platform organizes and manages system tasks and professional tasks to provide logical guarantees, and these tasks can be categorized as core tasks, basic tasks, guarantee tasks, and auxiliary tasks.

[0066] 1.11 Modeling of Basic System Elements

[0067] The basic elements in the system include task M. y Subtask T k Function k1 Resource R k Four types. A reconfigurable system has y Missions {M1, M2, ..., M}. y}, where M x ={T j ,T2,···,T k Subtask M x ={T j ,T2,···,T k} and also by Function k1 Function kt The supporting, functional process consists of m resources {R} j ,R2,···,R k Operational support can be represented by resource operation capability support vectors to depict the essence of task requirements. This progressively layered modeling clarifies the logical relationship between tasks and resources, providing a clear and structured foundation for subsequent mathematical modeling and analysis.

[0068] 1.12 Establish the initial requirement constraint model of the system

[0069] A functional requirement can be represented by a capability vector, D f =(c i :n i ,…c j :n j The requirements of subtask T can be expressed as follows: The requirements of task M can be expressed as By analyzing the task requirement tree, the set of capabilities that support a task requirement can also be expressed by a capability vector, i.e., d. m =(n1·c1,n2·c2,…,n n ·c n ).

[0070] The system task requirements are: Functional Requirements Support Set Resource support is integrated but:

[0071] because:

[0072]

[0073] thus:

[0074]

[0075] D = B·A·R

[0076] in:

[0077]

[0078] Expanding the demand matrix D yields:

[0079]

[0080] Where, d i If the demand vectors are at each level, then This is the total demand vector.

[0081] Based on the above modeling process, an initial demand matrix for the system was constructed. By mapping and comparing the demand matrix with the resource matrix, the specific resource capacity requirements of the system at the current moment can be quantified. This analysis provides a rigorous theoretical basis for assessing task reliability and lays the foundation for resource allocation optimization and dynamic system reconfiguration.

[0082] Capability deployment is a process of decomposing system task requirements layer by layer through the linear representation relationships between system elements. That is, it's the process of expressing d through R, from expectation to the vector group [r1 r2…r ... m ] T To express d as d <= λ1r1 + λ2r2 + ... + λ m r m ; System task requirement D should be derived from the system's effective function set F = {F} l ,F2,…,F x Linear or partially linear representation. System functional design requirements. The system effective resource support set R = {R1:n1, R2:n2, …, R m :n m} is linearly tabulated or partially linearly tabulated. In this way of hierarchical linear development, the system task demand is mapped to the functional demand and resource support level by level, realizing the capability decomposition from the high-level task target to the bottom-level resource support, and providing a clear mathematical basis for task reliability evaluation and resource allocation optimization.

[0083] In summary, the system element modeling, the resource vector group R and the task demand matrix D are completed in this step, and the system constraint that can meet the normal execution of the task is given, that is, the system functional design demand l ,F2,…,F x} should be linearly tabulated or partially linearly tabulated. The system functional design demand The system effective resource support set R = {R1:n1, R2:n2, …, R m :n m} is linearly tabulated or partially linearly tabulated. However, the resource vector group R of the system will change with the operation of the system, which leads to the fact that the constraint cannot or partially cannot be met, and further leads to the fact that the system task reliability decreases from the initial state. When these quantities change, the resource vector to which they belong naturally also changes.

[0084] 2. As shown in Figures 2-3 When the relationship between the resource vector group and the task demand matrix changes, change analysis is carried out through mathematical modeling, the existence and uniqueness of the capability-demand constraint solution are judged, a new resource allocation scheme is generated, and system reconstruction is completed. When the capability-demand constraint solution has a unique solution, a new resource allocation scheme is generated, when the capability-demand constraint solution has infinite solutions, a target function and a constraint condition are introduced into the system reconstruction mathematical model, and the optimal solution is calculated, when the capability-demand constraint solution has no solution, the system resource or demand is adjusted, and a new resource vector group and task demand matrix are constructed. Including

[0085] 2.1 Mathematical modeling of system dynamic reconstruction behavior

[0086] The mathematical modeling step of the system dynamic reconstruction behavior is developed around the dynamic balance between the system resource capability and the task demand. When the system task reliability decreases with time, the relationship between the resource vector group R and the task demand matrix D changes, and through mathematical modeling analysis, a new resource allocation scheme is sought to maintain the overall service capability of the system. After the system dynamic reconstruction behavior, the basic solution system may still be able to meet the following formula.

[0087]

[0088] Y X R ≥ D

[0089] Y X ≤ N

[0090] where the resource capability is organized as R1, R2,..., R m into a capability matrix R, that is, R = (R1, R2,..., R m ), D is a matrix composed of multiple task demand levels, that is, D is contained in the column space of R, X and Y are level configurations facing the demand D, and N is a demand resource size vector. In addition, the constraint condition is Y X ≤ N. The solution of the formula has three possibilities: a unique solution, infinite solutions, and no solution. The unique solution indicates that the current system resource capability can exactly meet the demand; the infinite solutions mean that the system can meet the demand through multiple configurations; and the no solution indicates that the system resource can no longer support the current task demand.

[0091] The essence of reconstruction is to find a new current resource vector group from the changed P that can meet the current demand. When the system is reconstructed, another maximal linearly independent group of P needs to be given for D, and the task demand can still be met. Thus, due to the compensation effect of logical reliability, although the physical reliability of the system has degenerated, the service capability (that is, the overall reliability) of the system is not affected.

[0092] The system resource effectiveness state allows D to be represented by the resource vector group P (each component of X needs to meet the system size constraint). If D cannot be represented by P, it means that the current resource platform cannot meet the demand constraint D. Due to the gradual degradation of the physical health of the system, the original balance between resource capability and task demand is broken. In order to establish a new balance, a new resource configuration needs to be found according to the existing resource capability. According to the capability-demand model, when the health condition of the system changes, the change in system capability causes the original capability matrix P to change to P1, which is a matrix similar to P. Under the condition that D is unchanged, the process of establishing a new capability-demand balance is system reconstruction, that is,

[0093] P1 Y ≥ D

[0094] Y ≤ N

[0095] From the mathematical form, solving the above formula can be expected to obtain the configuration vector Y. As described above, the solution of the model is divided into three cases: there is a unique solution, there are infinite solutions, and there is no solution. For a given demand D, when there is a unique solution, it means that the current system capability composition exactly meets the system task demand D. The solution vector Y gives the resource configuration vector. When there are infinite solutions, it means that the current system capability can meet the demand D through multiple configurations, and the solution vector Y is given by the general solution expression. When the above formula has a solution, if r = R(P) = R(P, D), finding a linearly independent group with r number of elements is crucial, which is the support resource set of the system.

[0096] Get the requirements d c The configuration involves finding r linearly independent resource vector groups. When the number of resource types is greater than r, finding a set of r linearly independent resource vector groups is the problem of solving the above equation. Obtaining the optimal configuration from these vectors is the optimization problem of reconfigurable system configuration. The equation is satisfied when the requirements are exactly met.

[0097] In practical solutions, for requirement D, a maximal linearly independent set is not necessarily required; instead, it is sufficient to express a set that satisfies P1·Yd. c A linearly independent set G ≥ θ is sufficient, but it necessarily holds that R(G) <= r. For example, there is... If d c =d1=(2,1), then a maximal linearly independent set of P1 is {R1,R2}, and a configuration is as follows. Meanwhile, 2·R3 T It can also satisfy d c Additionally, if there are

[0098]

[0099] d c =d2=(2,1,0,0)

[0100] The maximal linearly independent set of P1 is (R1, R2, R3), which is the largest linearly independent set of the equation P1·Y=d. c For example, since R(P1) = 3, R(P1,d) c Since ) = 4, there is no solution. However, for d1, we can use {2R1}, {2R3} or {R1,R2} to satisfy the condition.

[0101] In summary, after modeling the mathematical model of the system's dynamic reconfiguration in this step, the capacity-demand constraint solutions are classified into three cases: unique solution, infinitely many solutions, and no solution. When there is a unique solution, the system capacity exactly meets the task requirement D, and the solution vector Y gives the resource allocation scheme. When there are infinitely many solutions, the goal of the system's dynamic reconfiguration is to select the scheme with the highest system reliability from multiple configurations through optimization calculations. When there is no solution, it indicates that the current system resources cannot meet the task requirements, and resources or requirements need to be adjusted to restore balance. Based on this, an objective function and constraints are further introduced to provide a computational basis for achieving the optimal solution for the system's dynamic reconfiguration.

[0102] 3. For example Figures 2-3As shown, when the capability-demand constraint solution has infinitely many solutions, if there is no solution (i.e., there are infinitely many solutions and there is no optimal resource vector set that satisfies the current task requirements), then the task requirements need to be downgraded and solved again. When solving again, if there is a unique solution, then there is no need to solve again, and the reconstruction is completed by directly outputting the solution.

[0103] In the system reconfiguration mathematical model, a resource vector space is generated for each linearly independent group of resource vectors. Within this space, the total resource cost for each resource vector group is calculated based on the usage cost of each resource. Specifically, this includes:

[0104] The resource vector group P is an m×n matrix (R1,R2,…,Rm), d c For the current synthesis task requirements, if R(P) = R(P,d) c When ) = m, [P,d c [They are linearly dependent, while P is linearly independent, because P exactly satisfies d] c Therefore, P is called the minimum system configuration.

[0105] Given a set of n-dimensional vectors {R1, R2, ..., R...} m The vector space it generates is as follows:

[0106]

[0107] This means that a given set of resource vectors can span a resource vector space. However, for the current task requirements, the coordinates (λ1, λ2, ..., λ) that can be satisfied... m The number of coordinates is finite. When there is more than one coordinate that satisfies the current task requirements, the system's dynamic reconfiguration behavior aims to find the optimal solution. From a mathematical perspective, if the coordinates (λ1, λ2, ..., λ...) are finite... m The formula The smaller the value, the fewer resources are used, and the better the solution. From an engineering perspective, considering resource costs, the cost of using each resource is different, so different configurations directly affect the total cost. Let the resource cost be represented by a vector w = (w1, w2, ..., w...). m Then the total resource cost is:

[0108]

[0109] The resource vector set P is an m×n matrix (R1, R2, ..., Rm). If the capacity-demand equation has a solution, and n < m, then the column vectors are necessarily linearly dependent, and the system resources are in a state of configurable redundancy. R(P) = r, then the number of combinations of the maximal linearly independent set of P is C(r, m), which is the number of unconstrained system configurations. These configurations may all satisfy the task requirement d. cTherefore, for these C(r,m) configurations, we can use the lowest cost as the evaluation criterion. Considering generality, the cost function is given here in vector form.

[0110] F c =(f1(n1),f2(n2),…,f m (n m ))

[0111] Then there is

[0112]

[0113] Among them, f i (λ i ) is R i The corresponding cost function, λ i For R i The linear combination coefficients.

[0114] The problem can be described as follows:

[0115] Objective function:

[0116]

[0117] Constraints:

[0118] B·X≥d c

[0119] X≤N

[0120] For each maximally linearly independent set, calculate its total cost C; the configuration with the smallest C value is the optimal configuration. This reduces to an optimization problem, which can be solved using integer programming.

[0121] 4. For example Figures 2-3 As shown, after the system dynamic reconfiguration is completed, the remaining resource capacity and functional capacity are calculated, and the actual remaining capacity of the system is evaluated by combining reliability indicators, thus completing the reliability calculation and analysis of the dynamically reconfigured system.

[0122] In a reconfigurable system, resource capacity can be divided into three parts: 1) activated capacity, i.e., capacity or energy in use; 2) redundant capacity, i.e., some unused capacity that cannot be used by other tasks due to resource allocation, which can also be considered temporarily wasted capacity; 3) surplus capacity, i.e. idle and allocable capacity. Total capacity is the sum of the system's existing available effective capacity; activated capacity is the capacity in use, also known as energy. The energy level of a reconfigurable system is the percentage of activated capacity relative to the total available effective capacity. Measuring surplus capacity can provide decision-making references for system maintenance. The capacity activated by task execution at a certain moment only indicates the capacity level required by that task, while system reliability is related to other system capabilities. Therefore, the total capacity level of the system needs to be considered when examining system reliability.

[0123] A system's remaining capacity can be divided into functional remaining capacity and resource remaining capacity. From a system maintenance perspective, resource remaining capacity is more relevant. During system operation, resource remaining capacity changes dynamically and varies depending on function calls and task execution.

[0124] 4.1 Calculation of Remaining Resource Capacity

[0125] The system's resource vector set is R = {R1, R2, ..., R...} m}, then the capability space is represented as:

[0126]

[0127] The capability space depends on the resource availability of the system itself. The capability space can be a hypercubic space covered by the capability of the Euclidean space determined by the resource vector set. Of course, due to the limitation of resource scale, the resource capability is only a spatial region.

[0128] The resource synthesis capability vector can be obtained from the resource capability vector model as c = (c1, c2, ..., c n To measure c, we must introduce a quantity that measures vectors, using the norm of c, which can be defined as follows: Considering the different levels of importance of the ability items, a weight vector w = (w1, w2, ..., w) is given. n Considering the general case, the overall system capability level can be defined as:

[0129]

[0130] In a reconfigurable system, the relationship between resources and tasks can be expressed by the following formula:

[0131]

[0132] Decompose the equation as:

[0133]

[0134] where, represents the total resource demand vector of tasks, represents the total resource demand vector of functions, solve the equation X and Y are integer programming problems, with the help of X, Y can get the actual support of resources to tasks. By comparing the X and the size vector of resources , the remaining capacity and energy level of resources can be obtained.

[0135] Resource effectiveness measure focuses on the existing capacity of resources, estimates the shortage of expected resources, that is, the gap between the capacity of resource i and the expectation It can be calculated by the following expression:

[0136]

[0137] where, the design capacity is set as the standard , then Then the traditional resource capacity effectiveness can be measured by the equation as follows:

[0138]

[0139] where, δ r (e r ) represents the mapping function of resource defects, which takes value in [0,1] and reflects the relationship between resource r and other resources in the platform.

[0140] The remaining capacity of a single resource is related to the capacity structure of the resource, and the capacity weight should be considered when measuring. For example, if a resource of R1 has a capacity degradation (defect weight bits are 0.15 and 0.2), then the remaining capacity of the resource is estimated to be 0.65 according to the capacity weight.

[0141] 4.2 Function remaining capacity calculation

[0142] By comparing Y and the size vector of functions, the remaining capacity and energy level of functions can be obtained.

[0143]

[0144] The capacity effectiveness of the function platform can be obtained by measuring the function demand of the system from the output function set. The priority order of the function output may affect the results of the function platform effectiveness, especially in the case of low function redundancy.

[0145] Assume that the input function set is F I ={F d1F d2 ,…,F dm}, where F di = (fd1, fd2,..., fd n ); the system existing output function set is F O = {F c1 , F c2 ,..., F cm}, where F ci = (fc1, fc2,..., fc n ). Then, for the function F ci , the following is calculated:

[0146]

[0147] Then, for the entire function platform, the function effectiveness A F can be measured by the following formula:

[0148]

[0149] 4.3 Reconfigurable system task reliability calculation

[0150] The reliability of a reconfigurable system not only depends on the resources (such as hardware, computing power, energy, etc.) that the system has, but also depends on how these resources support the normal operation of system functions. Resource residual capacity reflects the potential of the system to continue to provide services under the current state, while function residual capacity measures how many original functions the system can maintain in the face of failures or abnormal situations. Therefore, the actual residual capacity of the system is jointly determined by the resource residual capacity and the function residual capacity, which in turn affects the reliability of the system.

[0151] A T = Ο(A R , A F )

[0152] Wherein, O represents a certain relationship or calculation method, which describes how the system task reliability depends on the combination of resource residual capacity A R and function residual capacity A F . Only by adding or linearly combining the system reliability, the residual capacity of the system may be exaggerated. For example, the function residual is high but the resource is almost exhausted, and the system still cannot maintain normal work.

[0153] The multiplication calculation method used in actual calculation ensures that only when the resource and function reach a certain level, the system reliability will remain high, otherwise if any one decreases significantly, the system reliability will decrease rapidly.

[0154]

[0155] According to the above formula, the task reliability of the reconfigurable system can be calculated.

[0156] Therefore, the application can effectively depict the behavior of the system in the dynamic reconfiguration process, and clearly define the selection and optimization target in the reconfiguration process, thereby providing a quantitative basis for the reconfiguration decision of the system, and can also calculate the residual capability of the resource and the residual capability of the function, thereby ensuring the comprehensiveness and accuracy of the reliability analysis.

[0157] The various embodiments are described in a progressive manner in the specification, and each embodiment focuses on the difference from other embodiments, and the same or similar parts between the various embodiments can be mutually referred to.

[0158] The principles and implementation manners of the application are described by using specific examples in the specification, and the above description of the embodiments is only used to help understand the method of the application and the core idea thereof; meanwhile, for the person skilled in the art, the specific implementation manner and application range of the application can be changed according to the idea of the application. In conclusion, the content of the specification should not be understood as the limitation of the application.

Claims

1. A method for modeling and analyzing task reliability based on a dynamic reconfiguration system, characterized in that, include: By utilizing resource operation capability support vectors, we model the tasks, subtasks, functions, and resources in the dynamic reconfiguration system, construct resource vector sets and task requirement matrices, and obtain the initial requirement constraint model of the system. When the relationship between the resource vector group and the task requirement matrix changes, a change analysis is performed through mathematical modeling to determine the existence and uniqueness of the capability-requirement constraint solution, generate a new resource allocation scheme, and complete the system reconfiguration. Specifically, when the capability-requirement constraint solution has a unique solution, a new resource allocation scheme is generated; when the capability-requirement constraint solution has infinitely many solutions, an objective function and constraints are introduced into the system reconfiguration mathematical model, and the optimal solution is calculated; when the capability-requirement constraint solution has no solution, the system resources or requirements are readjusted, and a new resource vector group and task requirement matrix are constructed. After the system dynamic reconfiguration is completed, the remaining resource capacity and functional capacity are calculated, and the actual remaining capacity of the system is evaluated by combining reliability indicators, thus completing the reliability calculation and analysis of the dynamically reconfigured system.

2. The method of claim 1, wherein, Using resource operation capability support vectors, we model tasks, subtasks, functions, and resources in a dynamic reconfiguration system. This includes: using resource operation capability support vectors, we establish modeling descriptions between tasks, subtasks, functions, and resources in a progressive manner to obtain the logical relationships between tasks and resources. Among them, a task consists of multiple subtasks, a subtask is supported by multiple functions, and a function is supported by multiple resource operations.

3. The method of claim 2, wherein, Construct resource vector sets and task requirement matrices to obtain the initial requirement constraint model of the system, including: Based on the logical relationship between tasks and resources, a task requirement matrix is ​​constructed using capability vectors, functional requirement support sets, and resource support sets. The task requirement matrix is ​​mapped and compared with the resource vector group. Through the linear representation relationship between system elements, the task requirement matrix is ​​decomposed layer by layer so that the system task requirements are mapped to functional requirements and resource support level by level.

4. The method of claim 3, wherein, When the relationship between the resource vector group and the task requirement matrix changes, a change analysis is performed through mathematical modeling to determine the existence and uniqueness of the capability-demand constraint solution, generate a new resource allocation scheme, and complete system refactoring, including: For the task requirement matrix, by solving the configuration vector and analyzing the capability-requirement constraint solution, a current resource vector set is obtained to meet the current task requirements. Based on the current resource vector set, a system reconstruction mathematical model is constructed to find a new resource configuration scheme. When the capacity-demand constraint solution has a unique solution, the current resource vector group satisfies the current task, and the configuration vector provides a resource configuration scheme; When the capacity-demand constraint solution has infinitely many solutions, the configuration vector is optimized by searching for multiple linearly independent resource vector groups in the resource vector group and selecting the optimal resource vector group that meets the current task requirements. When the capacity-demand constraint solution is unsolvable, the current resource vector set cannot meet the current task. The system resources or system requirements are readjusted, and a new resource vector set and task requirement matrix are constructed.

5. The method of claim 4, wherein, When the capacity-demand constraint solution exists infinite solutions and there is no optimal resource vector group satisfying the current task demand, in the system reconfiguration mathematical model, each linearly independent resource vector group generates a resource vector space, in which the total resource cost of each resource vector group is calculated according to the use cost of each resource; The formula for calculating the total resource cost is: wherein f i (λ i ) is a cost function corresponding to the resource vector group, λ i is a linear combination coefficient of the resource vector group; The objective function and the constraint condition are introduced into the total resource cost, and the minimum total resource cost value is selected as the optimal solution of the system dynamic reconstruction.

6. The task reliability modeling and analysis method based on the dynamic reconstruction system according to claim 5, characterized in that, The formula of the objective function is: The formula of the constraint condition is: B * X ≥ d c X≤N where d c is the task demand, N is the demand resource scale vector, and X is the demand-oriented hierarchical configuration.

7. The method of claim 6, wherein, After the system dynamic reconstruction, the actual residual capacity of the system is evaluated by calculating the residual capacity of the resources and the residual capacity of the functions, and combining the reliability index, and the reliability calculation and analysis of the dynamic reconstruction system are completed, including: A norm for measuring the resource synthesis capability vector is introduced into the resource capability vector model, and a weight vector for measuring the importance of each resource synthesis capability vector is set, and then the total system capability level is defined by using the norm and the weight vector; The definition formula of the total system capability level is: Wherein, Cs is the total resource cost, c is the norm, and w is the weight vector; The support relationship of the resources to the task is calculated to obtain the residual capacity and energy level of the resources and the residual capacity and energy level of the functions; The expression of the support relationship is: Wherein, R is resource vector group, X, Y is demand-oriented hierarchical configuration, is the total resource demand vector of function, is the total resource demand vector of task. By comparing X with the scale vector of the resources, the residual capacity and energy level of the resources are obtained, and by comparing Y with the scale vector of the functions, the residual capacity and energy level of the functions are obtained.

8. The method of claim 7, wherein, After the system dynamic reconstruction, the actual residual capacity of the system is obtained by calculating the residual capacity of the resources and the residual capacity of the functions, and the reliability calculation and analysis of the dynamic reconstruction system are completed, and further including: The gap between the capacity of the resources and the expectation is calculated to measure the effectiveness of the resources; The expression of the resource effectiveness measurement is: where ASR(i) is the resource effectiveness, δ r (e r ) is the mapping function for resource defects; The output function set is used to measure the effectiveness of the functions; The expression of the function effectiveness measurement is: F I = {F d1 , F d2 ,..., F dm} F di = (fd1, fd2,..., fd n ) F O = {F c1 , F c2 ,..., F cm} F ci = (fc1, fc2,..., fc n ) wherein A F is a functional effectiveness, F O is a system existing output function set, F I is an input function set, F ci is a function; The actual residual capacity of the system is calculated by using the multiplication calculation method; The expression of the actual residual capacity is: where A T is the actual remaining capacity, A R is the resource effectiveness, A F is the functional effectiveness.

Images