CN103745116A - DIMA (distributed integrated modular avionics) system evaluation and optimization method - Google Patents

Abstract

The invention belongs to the technical field of aviation equipment design and discloses a DIMA (distributed integrated modular avionics) system evaluation method. The method includes: performing physical layer modeling, and mapping hardware equipment in a DIMA system to a mounting space to create a two-dimensional physical layer mapping vector; performing functional layer modeling, and mapping subtasks in the DIMA system to the hardware equipment according the known system-task-subtask tree to create a two-dimensional functional layer mapping vector; performing priority ranking on multiple performance evaluation indexes, and building a DIMA system multi-target performance description function set; substituting the physical layer mapping vector and the functional layer mapping vector into the DIMA system multi-target performance description function set to obtain a DIMA system evaluation result. The invention further discloses a DIMA system optimization method.

Description

Distributed comprehensively modularized avionics DIMA system evaluation method and optimization method

Technical field

The invention belongs to air equipment design field, be specifically related to a kind of distributed comprehensively modularized avionics DIMA system evaluation method and optimization method.

Background technology

Along with day by day improving with complicated various of aircraft (particularly civilian big-and-middle-sized aircraft) function, it is day by day complicated and huge that airborne electronic equipment system becomes, transmission and processing to information on machine have brought very large challenge, mass data needs safely, in time, accurately, intactly transmit and process, aircraft airborne system (particularly civil aircraft airborne electronic equipment system architecture) is directly connected to means, strategy and the method for the airborne electronic equipment system integration, integration test and checking, is one of aircraft development key technical problem.

Early stage airborne equipment is discrete, each equipment independently completes function separately, equipment room does not have or seldom has the exchange of information and sharing of resource, the development of digital technology makes the contact of airborne equipment more and more, mobile system becomes increasingly complex, particularly along with the development of supercomputing technology, on machine, information processing presents synthesization, modularization, standardized trend, bus network technology provides high reliability for mobile system, the network foundation of high integrality and high transfer rate, be widely used at present modern airliner, as airliner Boeing B787 and Air Passenger A380.With combining closely of bus network technology, mobile system framework has centralized architecture and two kinds of modes of distributed structure/architecture, the Typical Representative of centralized architecture is B787, various standardized processing modules concentrate in same rack to be processed, be called comprehensively modularized avionics system (IMA), the Typical Representative of distributed structure/architecture is Air Passenger A380, various standardized processing modules are dispersed in by function in the processing rack of zones of different on machine, be called distributed modular avionics (distributed Integrated Modular Avionics, DIMA) system.Than IMA, DIMA is placed on the position of close terminal (as various types of sensors, actuator etc.) calculating and process the ground that distributes, and the distributed structure of DIMA all may be brought improvement in various degree at aspects such as weight, volume, power consumption and system complexity, the securities of aircraft.

Due to distributed frame and modular equipment and the treatment mechanism of DIMA, different design proposals can be brought different performances, therefore needs DIMA to carry out modeling to assess the performance of certain system, and further optimizes according to demand and obtain optimal design.

with Frank Thielecke in document " Supporting the Design of Distributed Integrated Modular Avionics Systems with Binary Programming " and " Multi-objective mapping optimization for distributed Integrated Modular Avionics " by duty mapping is carried out to modeling to different equipment with by device map to a four different equipment compartment to DIMA, and for weight, just dressing up multiple targets such as basis and operation cost is optimized, utilize overall 0-1 integer multiple objective programming to solve a series of Pareto optimal solutions.But, in document, directly give locking equipment, there is no to cover the process of selecting from multiple optional equipments; Performance index while being optimized in document concentrate on weight and all kinds of cost from the angle of economy, considered, lack from safety perspective and consider a prior class reliability index avionics system design; Constraint condition, performance index all in document all only limit to linear function, and can completely is not described all kinds of situations in practical engineering application; And in document, solve a series of Pareto optimal solutions, and ignored the feature of avionics system self in aerospace applications, different performance index (objective function) particularly reliability index have different priority, should give consideration in optimization.

Summary of the invention

The present invention one of is intended to solve the problems of the technologies described above at least to a certain extent or at least provides a kind of useful business to select.For this reason, one object of the present invention is to propose a kind of thoughtful distributed comprehensively modularized avionics DIMA system evaluation method comprehensively.Another object of the present invention is to propose the distributed comprehensively modularized avionics DIMA system optimization method that a kind of optimum results is good.

For achieving the above object, according to the distributed comprehensively modularized avionics DIMA system evaluation method of the embodiment of the present invention, comprise the following steps: carry out Physical layer modeling, the hardware device in described DIMA system is mapped in installing space, to create the Physical layer mapping vector x of binary _ins; Carry out functional layer modeling, according to known system-task-subtask tree, the subtask in described DIMA system is mapped in described hardware device, to create the functional layer mapping vector x of binary _pros; Multiple Performance Evaluating Indexes are carried out to prioritization, and set up the multiobject performance specification function set of DIMA system { f _goal1(x _ins, x _pros), f _goal2(x _ins, x _pros) ... f _goalNum(x _ins, x _pros), subscript goalNum represents target number; By described Physical layer mapping vector x _inswith functional layer mapping vector x _prosthe multiobject performance specification function set of DIMA system { f described in substitution _goal1(x _ins, x _pros), f _goal2(x _ins, x _pros) ... f _goalNum(x _ins, x _pros) obtain the assessment result of DIMA system.

In one embodiment of the invention, described Physical layer mapping vector $x_{\mathrm{ins}}={[x_{{\mathrm{<figure-callout\; id="1"\; label="D"\; filenames="HDA0000458144240000022.png"\; state="\{\{state\}\}">D</figure-callout>}}_1,{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA0000458144240000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_1},x_{{\mathrm{<figure-callout\; id="1"\; label="D"\; filenames="HDA0000458144240000022.png"\; state="\{\{state\}\}">D</figure-callout>}}_1,L_2},...,x_{{\mathrm{<figure-callout\; id="1"\; label="D"\; filenames="HDA0000458144240000022.png"\; state="\{\{state\}\}">D</figure-callout>}}_1,L_{N_{\mathrm{Loc}}}},...,x_{D_{N_{\mathrm{Dev}}},{\mathrm{<figure-callout\; id="1"\; label="L\; N"\; filenames="HDA0000458144240000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_{N_{}}},...,x_{D_{N_{\mathrm{Dev}}},L_{N_{\mathrm{Loc}}}}]}^T,$ Wherein N _locfor the number of described installing space, N _devfor the number of described hardware device, when i described hardware device is mapped in j described installing space, $x_{D_i,L_j}=1,$ Otherwise $x_{D_i,L_j}=0.$

In one embodiment of the invention, described Physical layer mapping vector x _insneed to meet following two constraint conditions:

The first constraint condition is that same hardware device can not be installed in multiple installing spaces simultaneously, and mathematic(al) representation is: A _{ins_sgl}x _ins≤ b _{ins_sgl}, wherein

b _{ins_sgl}=[1,1 ..., 1] ^tand be 1 × N _devrank matrix, I _loc=[1,1 ..., 1] and be N _loc× 1 rank matrix;

The second constraint condition is no more than the installation resource that this installing space provides for the installation resource of all hardware devices consume in same installing space, and mathematic(al) representation is: A _{ins_rsc}x _ins≤ b _{ins_rsc}, wherein $A_{\mathrm{ins}\_\mathrm{rsc}}=\left(\begin{array}{c}{A}_{\mathrm{ins}\_\mathrm{rscl}}\\ .\\ .\\ .\\ A_{\mathrm{ins}\_\mathrm{rscN}}\end{array}\right),b_{\mathrm{ins}\_\mathrm{rsc}}=\left(\begin{array}{c}{b}_{\mathrm{ins}\_\mathrm{rscl}}\\ .\\ .\\ .\\ b_{\mathrm{ins}\_\mathrm{rscN}}\end{array}\right),$ Subscript ins_rsc1 to ins_rscN represents different installation resource types, take ins_rscx represent x class installation resource wherein 1≤x≤N and x as integer, $A_{\mathrm{ins}\_\mathrm{rscx}}=\left[\begin{array}{cccc}{\mathrm{<figure-callout\; id="1"\; label="s\; D"\; filenames="HDA0000458144240000022.png"\; state="\{\{state\}\}">s</figure-callout>}}_{D_{}},0,...,0& s_{D_2,{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA0000458144240000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_1},0,...,0& ...& s_{D_{N_{\mathrm{Dev}}},{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA0000458144240000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_1},0,...,0\\ 0,{\mathrm{<figure-callout\; id="1"\; label="s\; D"\; filenames="HDA0000458144240000022.png"\; state="\{\{state\}\}">s</figure-callout>}}_{D_{}},...,0& 0,s_{D_2,L_2},...,0& ...& 0,s_{D_{\mathrm{Dev}},L_2},...,0\\ ...& ...& ...& ...\\ 0,...,{\mathrm{<figure-callout\; id="1"\; label="s\; D"\; filenames="HDA0000458144240000022.png"\; state="\{\{state\}\}">s</figure-callout>}}_{D_{}}& 0,...,s_{D_2,L_{N_{\mathrm{Loc}}}}& ...& 0,...,s_{D_{N_{\mathrm{Dev}}},L_{N_{\mathrm{Loc}}}}\end{array}\right],$

a _{ins_rscx}in element

be the x class installation resource that i equipment consumes while being installed to j installing space, b _{ins_rscx}in element

represent the x class installation resource that j installing space provides.

In one embodiment of the invention, described in carry out functional layer modeling, according to known system-task-subtask tree by the duty mapping in described DIMA system in described hardware device, to create the functional layer mapping vector x of binary _prosspecifically comprise: according to system-task-subtask tree, determine that the processing resource type that described DIMA system needs is altogether pros_rscN kind, and determine that it is N that consumption n class is processed the n class subtask number of resource _tasknindividual, wherein 1≤n≤pros_rscN and n are integer; According to described x _insand the self property of hardware device, the number that is identified for providing n class to process the n kind equipment of resource is N _devnindividual, wherein 1≤n≤pros_rscN and n are integer; Described functional layer mapping vector

wherein part corresponding to n class subtask is $x_n={[x_{T_n^{\left(1\right)},D_n^{\left(1\right)}},x_{T_n^{\left(1\right)},D_n^{\left(2\right)}},...,x_{T_n^{\left(1\right)},D_n^{\left(N_{\mathrm{Devn}}\right)}},...,x_{T_n^{\left(N_{\mathrm{Taskn}}\right)},D_n^{\left(1\right)}},...,x_{T_n^{\left(N_{\mathrm{Taskn}}\right)},D_n^{\left(N_{\mathrm{Devn}}\right)}}]}^T,$ If t n class subtask is assigned on k n kind equipment, move, remember x _nin element otherwise

wherein 1≤t≤N _tasknand t is integer, 1≤k≤N _devnand k is integer.

In one embodiment of the invention, described functional layer mapping vector x _prosneed to meet following two constraint conditions, the 3rd constraint condition be same n class subtask indivisible, only can be assigned on a n kind equipment and complete, and there is not the situation of Lou selecting subtask, mathematic(al) representation is: A _{pros_sgl}x _pros=b _{pros_sgl}.Wherein

i _n=[1,1 ..., 1] and be N _devn× 1 rank matrix, b _{pros_sgl}=[1,1 ..., 1] ^tand be

rank matrix; The 4th constraint condition is no more than the processing resource that this hardware device provides for the summation of processing resource consumption corresponding to the subtask moved in same hardware device, and mathematic(al) representation is: A _{pros_rsc}x _pros≤ b _{pros_rsc}, wherein b _{pros_rsc}in the upper limit of certain class resource select in being shone upon by Physical layer and the summation of such resource that this kind equipment of installing provides determines, by described x _insdetermine.

For achieving the above object, according to the distributed comprehensively modularized avionics DIMA system optimization method of the embodiment of the present invention, comprise the following steps: propose the initialization scheme of distributed comprehensively modularized avionics DIMA system, and according to the appraisal procedure described in claim 1-5 any one, calculate the assessment result of initialization scheme; Multiple performance parameters are carried out to the order of priority of non-increasing, the performance parameter that priority resolution is high is optimization direction; On the basis of initialization scheme, according to the layer sorting method with tolerant, change multiple-objection optimization into single goal optimization and dwindle gradually optimal solution set to unique value or all objective optimizations and solve complete.

In sum, appraisal procedure of the present invention and optimization method, for the feature of aerospace applications, are realized DIMA system physical layer and functional layer modeling, set up the multiple objective function with hierarchical priority, complete Performance Evaluation and multi-objective optimization design of power to DIMA system, there is following characteristics and advantage:

(1) the invention provides a kind of method of distributed modular avionics system (DIMA) modeling and optimization that is adapted to aerospace applications, provide covering equipment selection, installation and system-task that DIMA designs-subtask tree to set up and be assigned to the whole process modeling of equipment, from reliability and validity two aspects, it is carried out to performance evaluation, and for the multiple goal that has hierarchical priority, complete the optimized design of DIMA.

(2) to the distributing uniform of equipment, utilize the supply of different resource and the formal modeling of consumption to describe physical layer equipment selection, installation and functional layer task.

(3) mapping of Physical layer and functional layer is that the combined action embodiment of the flow processs such as equipment is set up and be assigned to equipment selection, installation and system-task in DIMA design-subtask tree, set it as optimized variable, realized the comprehensive consideration to DIMA design each side.

(4) foundation of performance index function has covered reliability and the validity two class indexs of aerospace applications, and gives respectively different priority according to practical engineering application.

(5) multiple-objection optimization method for solving has utilized the priority of different performance index in aerospace applications, provides the optimal design of realistic engineering application demand.

Additional aspect of the present invention and advantage in the following description part provide, and part will become obviously from the following description, or recognize by practice of the present invention.

Accompanying drawing explanation

Above-mentioned and/or additional aspect of the present invention and advantage accompanying drawing below combination is understood becoming the description of embodiment obviously and easily, wherein:

Fig. 1 is the process flow diagram of the distributed comprehensively modularized avionics system appraisal procedure of the embodiment of the present invention;

Fig. 2 is the schematic diagram of the distributed comprehensively modularized avionics system appraisal procedure of the embodiment of the present invention;

Fig. 3 is the exemplary plot of the DIMA physical layer architecture of the embodiment of the present invention;

Fig. 4 is the exemplary plot of DIMA system-task-subtask tree of the embodiment of the present invention;

Fig. 5 is the process flow diagram of the distributed comprehensively modularized avionics system appraisal procedure of the embodiment of the present invention.

Embodiment

Describe embodiments of the invention below in detail, the example of described embodiment is shown in the drawings, and wherein same or similar label represents same or similar element or has the element of identical or similar functions from start to finish.Below by the embodiment being described with reference to the drawings, be exemplary, be intended to for explaining the present invention, and can not be interpreted as limitation of the present invention.

The present invention aims to provide a kind of distributed comprehensively modularized avionics (DIMA) System Assessment Method and optimization method that is adapted to aerospace applications, provide equipment selection, installation and system-task to DIMA-subtask tree to set up and be assigned to the modeling of equipment, carry out from many aspects performance evaluation, and according to the feature of aerospace applications, for multiple goal, complete the optimal design of DIMA.

As depicted in figs. 1 and 2, distributed comprehensively modularized avionics (DIMA) System Assessment Method of one embodiment of the invention can comprise the following steps:

Step S11. carries out Physical layer modeling, the hardware device in DIMA system is mapped in installing space, to create the Physical layer mapping vector x of binary _ins.

Particularly, with reference to Fig. 3, the installation space that is distributed in diverse location that the physical layer model of DIMA is provided by aircraft, all avionics hardware devices, the avionics bus network that connects them and other lead-in wires and avionics hardware device form to the corresponding relation (mapping) of installing space.

Step S12. carries out functional layer modeling, sets the duty mapping in DIMA system in hardware device, to create the functional layer mapping vector x of binary according to known system-task-subtask _pros.

Wherein, the functional layer of DIMA is called the Task Tree and the subtask that form and is formed to the corresponding relation (mapping) of operated in avionics hardware device by division, the necessary peripheral hardware of the task of avionics system and corresponding subtask.Tree can be with reference to figure 4 in system-task-subtask.

Step S13. carries out prioritization to multiple Performance Evaluating Indexes, and sets up the multiobject performance specification function set of DIMA system { f _goal1(x _ins, x _pros), f _goal2(x _ins, x _pros) ... f _goalNum(x _ins, x _pros), subscript goalNum represents target number.

For example, according to the actual demand of avionics system, can set up a series of objective functions of reliability and two priority levels of validity, the objective function priority level entirety of reliability level is higher than validity, and concrete objective function also gives priority flag simultaneously.

Step S14. shines upon vector x by Physical layer _inswith functional layer mapping vector x _prosthe multiobject performance specification function set of substitution DIMA system { f _goal1(x _ins, x _pros), f _goal2(x _ins, x _pros) ... f _goalNum(x _ins, x _pros) obtain the assessment result of DIMA system.

Concrete flow implementation is as follows:

1, DIMA Physical layer modeling

In the Physical layer modeling of DIMA, be arranged on the Aerial Electronic Equipment (hereinafter to be referred as " equipment ") of the standard modular in the installing space that aircraft provides, such as electronics bay, rack etc., be divided into multiclass standard device, for example but be not limited only to computing module, data memory module, I/O interface module etc., each kind equipment is divided into multiple models by its different performance and parameter, and these equipment are realized and being interconnected by the bus network of the fixed interface access aircraft in installing space; Other are arranged on the external unit (hereinafter to be referred as " peripheral hardware ") of aircraft non-modularization everywhere as required, such as sensor, actuator etc., by lead-in wire, be connected to the standard device of installing space, as data memory module, I/O interface module etc. is uploaded Information Monitoring and distribution control command etc.

In the physical layer model of DIMA, the parameter of installing space mainly comprises: residing positional information in aircraft, upper limit of the carrying capacity such as such as weight, volume, rack backboard slot, refrigerating capacity, bus network access providing for equipment etc., this carrying capacity quantizing according to Practical Project situation is called " installation resource " hereinafter for short, the major parameter of equipment comprises: the unit procurement cost of all kinds of each model device that may provide, maintenance cost etc., the maximum quota of all kinds of each model device, if and equipment is installed in installing space, the weight that can take, volume, rack backboard slot, refrigerating capacity, the quantity of the installation resource such as bus access capability etc., and be similar to the installation resource that installing space provides, each kind equipment provides quantification such " processing resource " for functional layer, for example computational resource, storage resources, I/O resource etc., the equipment of different model provides the processing resource of varying number simultaneously, the major parameter of peripheral hardware comprises: all kinds of peripheral hardwares are installed to the positional information in aircraft according to its concrete function, the outer required wire length that is set to each installing space being determined by its positional information and airframe, unit weight, the unit procurement cost of this peripheral hardware lead-in wire, safeguards replacement cost etc.

In the physical layer model of DIMA, from all kinds of each model device that may provide, select and be installed to the flow process (hereinafter to be referred as " Physical layer mapping ") of a certain installing space, carry out by the following method modeling.For expressing conveniently, the installing space that aircraft is provided is designated as

, wherein N _locbe the number of installing space, the equipment that may provide is designated as

, wherein N _devfor the number of devices likely providing, it is the summation of all kinds of each model number of devices.Physical layer mapping is by a N _loc× N _devthe vector x of dimension _insdescribe,

X _insonly containing " 0 " and " 1 " two kinds of elements, if wherein

represent that i equipment is installed to j installing space, otherwise if

represent that this installation does not occur.All possible x _insvalue has covered all possible Physical layer mapping situation.Physical layer mapping x _inscomprehensive reflection in DIMA system hardware device select and the result of installing in installing space.At x _insan example in, each equipment at most only may be installed to a space, ( ) in only have at most one 1, but also likely certain alternative equipment be not selected into system, so corresponding all elements is all 0.This constraint condition is described below by inequality group,

A _{ins_sgl}x _ins≤ b _{ins_sgl}(2) wherein

as previously mentioned, there is the upper limit for the installation resource that equipment provides in installing space, therefore all kinds of resource consumptions that are arranged on all devices of a certain installing space all can not exceed the resource upper limit that this installing space can provide, this constraint is also described by inequality group, for example: for first kind installation resource, its constraint inequality is A _{ins_rsc1}x _ins≤ b _{ins_rsc1}, wherein

$b_{\mathrm{ins}\_\mathrm{rscl}}={[r_s^{\left(L_1\right)},r_s^{\left(L_2\right)},...,r_s^{\left(L_{N_{\mathrm{Loc}}}\right)}]}^T.$

A _{ins_rsc1}in element be i equipment such resource consumption while being installed to j installing space, b _{ins_rsc1}in element

represent such resource that j installing space provides.Similarly, other installation resource constraints are also described by inequality group, and all inequality resource constraints synthesize by splicing

A _{ins_rsc}x _ins≤ b _{ins_rsc}(3) wherein, $A_{\mathrm{ins}\_\mathrm{rsc}}=\left(\begin{array}{c}{A}_{\mathrm{ins}\_\mathrm{rscl}}\\ .\\ .\\ .\\ A_{\mathrm{ins}\_\mathrm{rscN}}\end{array}\right),b_{\mathrm{ins}\_\mathrm{rsc}}=\left(\begin{array}{c}{b}_{\mathrm{ins}\_\mathrm{rscl}}\\ .\\ .\\ .\\ b_{\mathrm{ins}\_\mathrm{rscN}}\end{array}\right),$ Subscript ins_rsc1 to ins_rscN represents different installation resource types.Above-mentioned formula (2) and formula (3) are in Physical layer mapping model due to two groups of intrinsic constraints of its physical significance, remove this in addition, according to the demand of practical engineering application, other the mapping for Physical layer also can join in model, for example in system, may require certain two equipment not to be installed in same installing space, specifically describe as form is as g _ins(x _ins)≤0 contain x _insequation or the inequality of function, this constraint function is not limited to linear function.Vector x _inswith a series of constraint specifications the model of DIMA Physical layer mapping, itself and aforesaid installing space, equipment and peripheral hardware have formed the physical layer model of DIMA.

Implementing during Physical layer modeling, the present invention is using parameters such as the installing space in practical application, equipment and peripheral hardwares as initial conditions, to relevant parameter assignment in model, and creates the mapping vector x of corresponding dimension _ins, vector x _insthe initial value of middle element is 0, if in practical application equipment to be installed to installing space be corresponding element assignment 1 in vector.

2, DIMA functional layer modeling

In the functional layer modeling of DIMA, the relation of task, subtask in DIMA system is described by tree structure.Consider by a series of processing procedure, to complete specific function separately in DIMA, for example monitoring, control, communication, navigation, amusement etc., functional layer model is divided into different " task " by the processing procedure of actual DIMA system by different effects, task is by a series of " subtask " and call corresponding peripheral hardware and realize, subtask is defined as in model as realizing no longer further processing procedure or the software package of segmentation of specific purpose, this processing procedure may need to operate in the standard module equipment of multiple kinds, but all processing concentrate at most in an equipment and complete in same class equipment.Subtask is also the least unit that in functional layer model, the different disposal process from DIMA is shone upon to distinct device.Whole DIMA system is root node, the more than different task for dividing, above a series of subtasks for realizing this task.Fig. 4 is an example for system-task-subtask tree of DIMA systemic-function layer.

In the functional layer model of DIMA, by equipment in DIMA physical layer model, selected to have determined the operable all kinds of each model device of the task in functional layer and all kinds of processing resource summation with the result of installing, the parameter that belongs to the subtask of a certain task mainly comprises: the various kinds of equipment resource quantity that this subtask need consume or take and the peripheral hardware that need to call etc., establish in addition a certain subtask and need consume i class processing resource, for this subtask of the convenient note of statement is one " i class resource subtask ", hereinafter to be referred as i class subtask.A task may consume multiclass resource simultaneously, and therefore this task adheres to the subtask of multiple types separately.

Be similar to the physical layer model of DIMA, in the functional layer model of DIMA, subtask be assigned to the flow process (hereinafter to be referred as " functional layer mapping ") of moving in its required distinct device, also by binary vector x _prosmethod be described

$x_{\mathrm{pros}}={[x_1^T,x_2^T,...,x_{\mathrm{pros}\_\mathrm{rscN}}^T]}^T$

Wherein subscript 1 to pros_rscN represents different processing resource types, also corresponding different subtask type and device type, and part corresponding to i class subtask is

N _devifor shining upon x by Physical layer _insthe number of the equipment that i class resource is provided determining, hereinafter to be referred as i kind equipment, and has

the equipment sum of all classes equals Physical layer mapping x _insthe number of middle nonzero element.N _taskifor the sum of i class subtask in DIMA functional layer, because a task may adhere to multiclass subtask separately simultaneously, so

be more than or equal to general assignment number.If fruit x _iin element

representing that j i class subtask is assigned on k i kind equipment moves.Functional layer mapping x _proscomprehensive reflection the foundation of system-task in DIMA system-subtask tree and the task result to devices allocation mapping.

By the definition of subtask, an i class subtask only can be assigned on an i kind equipment, and this intrinsic constraint condition is provided by equation:

A _{pros_sgl}x _pros=b _{pros_sgl}wherein,

About this equation, it should be explained that, because all subtasks must be assigned on certain equipment, complete, do not exist certain subtask not to be selected into, therefore got rid of the situation that the equation left side equals 0 on mathematical description, so this constraint condition is equation, but not containing the inequality of "≤".

Be similar to the restriction of space resources in Physical layer mapping, the processing resource constraint of functional layer is also by inequality group A _{pros_rsc}x _pros≤ b _{pros_rsc}provide, wherein b _{pros_rsc}in the upper limit of certain class resource select in being shone upon by Physical layer and the summation of such resource that this kind equipment of installing provides determines.In the functional layer model of DIMA, except two kinds of above-mentioned inherent constraints, other can be by form as g according to practical engineering application _pros(x _pros)≤0 contain x _prosequation or the inequality of function provide.

When implementing functional layer modeling, according to practical engineering application, set up the tree structure of system-task-subtask, using all kinds of task parameters as initial conditions, to relevant parameter assignment in model, and create the mapping vector x of corresponding dimension _pros, vector x _prosthe initial value of middle element is 0, if stipulated in practical application, certain task is processed in particular device is corresponding element assignment 1 in vector.And b in processing resource constraint _{pros_rsc}by undetermined parameter, represented.

3, set up the multiple objective function with hierarchical priority for DIMA

The Performance Evaluating Indexes of DIMA system and objective function are divided into reliability index and Validity Index two classes.Reliability index refers to guarantee the important indicator of aircraft basic function and security, error probability, mistake tolerance, reconstruct and the bit error rate of communicating by letter in specific subsystem realizes with function of such as system, the delay time and jitter of network etc.Validity Index comprises the power consumption, weight considered from economic angle, just dresses up basis, operation and maintenance cost etc. and extra available service quality etc. assurance aircraft basic function and security requirement.According to the feature of aerospace applications, reliability index has higher priority with respect to Validity Index.

The performance index of DIMA system are described as Physical layer mapping x _inswith functional layer mapping x _prosfunction f (x _ins, x _pros), the performance that the designs of concentrated expression DIMA system physical layer and functional layer reach.If this target function is linear function, can be expressed as

but be not limited to linear function.Actual according to engineering application, set up respectively a series of reliability indexs and Validity Index, can be designated as for simplicity { f _i(x _ins, x _pros).Priori in being applied by engineering has corresponding priority to evaluate to each index simultaneously, reliability index relative reliability index has higher priority level, in the inner different index of one deck priority by the distribution of accurate weights or utilize the statement of similar " extremely important ", " important ", " generally ", " inessential ", " very inessential " to carry out mark.

Below respectively take computing module loading index and just the dress indicator of costs as example, the foundation of DIMA system performance index (objective function) is described.When large amount of complex calculation task moves for a long time in DIMA computing module, may affect the error probability of system treatment scheme, therefore with computing module loading index, be used for describing the loading condition of all computing module entirety in DIMA.By single computing module load definition, be $f_{\mathrm{CBur}}^{C_j}=b_{C_j}\log \left(b_{C_j}\right),$ Wherein $b_{C_j}=f_{C_j}\left(x_{\mathrm{pros}}\right)=f_{C_j}^T{A}_{\mathrm{pros}\_\mathrm{rsc}}{x}_{\mathrm{pros}}$ For x _prosfunction, represent the computational resource of all tasks consumption that operate in j computing module, and have

the computing module load of entire system is each computing module load summation

computing module loading index belongs to reliability index, is therefore divided into corresponding higher priority tier, and according to practical engineering application, in higher priority tier, gives " important " priority flag.Just the dress indicator of costs is used for describing the installation cost of DIMA devices in system, peripheral hardware and supporting lead-in wire, is defined as f _insCost(x _ins, x _pros)=c _dev+ c _per+ c _wire, be x _insand x _prosfunction, wherein c _devand c _perrepresent respectively first this summation of dressing up of all individual equipments and peripheral hardware, the equipment that peripheral hardware wire length is operated in by peripheral hardware positional information, the corresponding task of functional layer and the positional information of this equipment place installing space of Physical layer obtain, c _wirefor just dressing up this summation by the lead-in wire that unit length lead-in wire is just dressed up originally and wire length is calculated.Just the dress indicator of costs belongs to Validity Index, is therefore divided into corresponding lower priority layer, and according to practical engineering application, gives " generally " priority flag in lower priority layer.

While implementing to set up for the multiple objective function with hierarchical priority of DIMA, conventional index can off-line be set up and preserves called, also can select according to the actual requirements part index number, adjustment priority evaluation and set up new index etc.

4, DIMA system performance assessment

This stage judges according to the actual requirements carries out the assessment of DIMA system performance.For DIMA system performance assessment, according to DIMA system to be assessed, complete respectively Physical layer modeling and functional layer modeling, to each parameter assignment set up Physical layer mapping and functional layer is shone upon vector x respectively _insand x _pros, bring the performance index function f in a flow process into _i(x _ins, x _pros) can to this DIMA system, carry out Performance Evaluation from every side.

As from the foregoing, the first Physical layer modeling of distributed comprehensively modularized avionics (DIMA) System Assessment Method, rear functional layer modeling, wherein Physical layer covers from all kinds of optional equipments selection equipment and is installed to the process of installing space, and the 4th class constraint of functional layer is relevant with the selection result of Physical layer various kinds of equipment.Except Physical layer and functional layer are separately corresponding two classes constraints, constraint condition that can also be additional according to practical engineering application, and this additional constraint condition is not limited to linear restriction.The finally Performance Evaluation to DIMA system, comprises reliability index and the large class of Validity Index two, and gives priority flag.The present invention proposes a kind of reliability index especially, i.e. computing module load.

As shown in Figure 5, distributed comprehensively modularized avionics (DIMA) system optimization method of one embodiment of the invention can comprise the following steps:

Step S51. proposes the initialization scheme of distributed comprehensively modularized avionics DIMA system, and according to the assessment result that discloses appraisal procedure calculating initialization scheme above.

Step S52. carries out the order of priority of non-increasing to multiple performance parameters, the performance parameter that priority resolution is high is optimization direction.

Step S53., on the basis of described initialization scheme, according to the layer sorting method with tolerant, changes multiple-objection optimization into single goal optimization and dwindles gradually optimal solution set to unique value or all objective optimizations and solve complete.

The output of the DIMA system multi-objective optimization design of power that the present invention proposes, is not a series of Pareto optimal solutions, but has utilized the optimum solution of index priority prior imformation.Its detailed process is as follows.

For the multi-objective optimization design of power of DIMA, with Physical layer and functional layer mapping vector

for optimized variable, be constrained to constraint condition with intrinsic in Physical layer and functional layer and that set up according to practical engineering application m, be designated as g _j(x)≤0, j=1,2 ..., m, the performance index of the k with priority evaluation of setting up in an above flow process are that objective function carries out multiple-objection optimization, this multi-objective optimization question is described as:

$\left\{\begin{array}{c}\min \left\{u_i(w_i,f_i\left(x\right))\right\},i=\mathrm{1,2},...,k\\ s.t.x\&Element;G=\left\{x\right|g_j\left(x\right)\&le;0,j=\mathrm{1,2},...,m\}\end{array}\right.$

Wherein u _i(w _i, f _i(x)), i=1,2 ..., it is performance index function f that k is illustrated in optimization aim while being optimized _i(x) and priority evaluate w _ithe function of (accurate weights or priority flag as previously mentioned); g _j(x)≤0, j=1,2 ..., the multiple constraint condition that m comprises constraint intrinsic in DIMA model Physical layer and functional layer and sets up according to practical engineering application.

For the feature of DIMA multiple goal performance index and priority, below take linear weighted function method and a kind of with tolerant layer sorting method the multi-objective optimization design of power as example explanation DIMA.

If in the foundation of upper flow process DIMA performance index, be each objective function f _i(x) equal foregoing weight w of corresponding tax _i>=0 describes its priority, considers to utilize linear weighted function method that multi-objective optimization question is converted into single-object problem $f\left(x\right)=\underset{i}{\mathrm{\&Sigma;}}{w}_i{f}_i\left(x\right).$

If in the foundation of upper flow process DIMA performance index, at reliability and inner each the objective function f of validity two class objective functions _i(x) utilize respectively the statement of foregoing similar " extremely important ", " important ", " generally ", " inessential ", " very inessential " to carry out priority flag, consider that employing is a kind of with tolerant layer sorting method, establish objective function according to the sequence of the non-increasing of its priority to be $f_r^{\left(1\right)}\left(x\right),f_r^{\left(2\right)}\left(x\right),...,f_r^{\left(k_r\right)}\left(x\right),f_e^{\left(1\right)}\left(x\right),f_e^{\left(2\right)}\left(x\right),...,f_e^{\left(k_e\right)}\left(x\right),$ Wherein reliability objectives function $f_r^{\left(1\right)}\left(x\right),f_r^{\left(2\right)}\left(x\right),...,f_r^{\left(k_r\right)}\left(x\right)$ Than validity objective function $f_e^{\left(1\right)}\left(x\right),f_e^{\left(2\right)}\left(x\right),...,f_e^{\left(k_e\right)}\left(x\right)$ Entirety has higher priority, and non-increasing sequence is carried out according to priority evaluation in a class objective function inside.First to first aim function

solve single-object problem,

$\left\{\begin{array}{c}\min f_r^{\left(1\right)}\left(x\right)\\ s.t.x\&Element;G=\left\{x\right|g_j\left(x\right)\&le;0,j=\mathrm{1,2},...,m\}\end{array}\right.$

?

note optimal solution set is

with a tolerance (slack) expansion optimal solution set, obtain

wherein tolerance Δ ₁by decision maker according to

priority, minimum value f ₁ ^*provide with numerical value or ratio with the tolerance of practical engineering application.Then use the optimal solution set of expansion to second target function

solve single-object problem $\left\{\begin{array}{c}\min f_r^{\left(2\right)}\left(x\right)\\ s.t.x\&Element;{\tilde{G}}_1\end{array}\right.,$ Continue successively, to j objective function, note $\min f^{\left(j\right)}\left(x\right)=f_j^{*},$ Optimal solution set is $G_j=\{x\&Element;{\tilde{G}}_{j-1}|f^{\left(j\right)}\left(x\right)=f_j^{*}\},$ Expansion optimal solution set is then use

to f ^(j+1)(x) solve single-object problem until optimal solution set only all solves completely containing a value or all objective functions, finally obtain the optimal solution set of DIMA multi-objective optimization question.Tolerance Δ in the present invention _jvalue can be before solving disposable providing, also can in calculating, according to each step result of calculation interactive mode, provide.

In the multi-objective optimization design of power of enforcement DIMA, for the multi-objective optimization question with hierarchical priority, provide multiple concrete method for solving to select, and can add new multi-objective optimization question method for solving, such as but not limited to linear weighted function method, layer sorting method, layer sorting method with tolerant, maximin method, highest priority method, interactive programming method etc.

In sum, the present invention is directed to the feature of aerospace applications, realize DIMA system physical layer and functional layer modeling, set up the multiple objective function with hierarchical priority, complete Performance Evaluation and multi-objective optimization design of power to DIMA system, there is following characteristics and advantage:

(1) the invention provides a kind of method of distributed modular avionics system (DIMA) modeling and optimization that is adapted to aerospace applications, provide covering equipment selection, installation and system-task that DIMA designs-subtask tree to set up and be assigned to the whole process modeling of equipment, from reliability and validity two aspects, it is carried out to performance evaluation, and for the multiple goal that has hierarchical priority, complete the optimized design of DIMA.

(2) to the distributing uniform of equipment, utilize the supply of different resource and the formal modeling of consumption to describe physical layer equipment selection, installation and functional layer task.

(3) mapping of Physical layer and functional layer is that the combined action embodiment of the flow processs such as equipment is set up and be assigned to equipment selection, installation and system-task in DIMA design-subtask tree, set it as optimized variable, realized the comprehensive consideration to DIMA design each side.

(4) foundation of performance index function has covered reliability and the validity two class indexs of aerospace applications, and gives respectively different priority according to practical engineering application.

(5) multiple-objection optimization method for solving has utilized the priority of different performance index in aerospace applications, provides the optimal design of realistic engineering application demand.

Below with the implementation process of a case introduction distributed comprehensively modularized avionics system appraisal procedure of the present invention and optimization method.For simplifying statement, in example below, only consider that partial parameters, performance index and optimization method are as example, in practical engineering application, other parameters, index and optimization method can be expanded by similar disposal route.

If an aircraft provides the installing space of four distributions for DIMA system, be respectively driving cabin headroom, be positioned at the conditional electronic equipment compartment of driving cabin bottom, and lay respectively at the electronic compartment of waist and afterbody, according to practical engineering application, gather all kinds of installation resource information the quantification that installing space provides, in this example, only consider backboard slot, weight and three kinds of installation resource of refrigerating capacity, design parameter numerical value is as shown in table 1.

Table 1: installing space parameter

If only consider in this example two kind equipments, core processing module (Core Processing Module, CPM) and remote data center module (Remote Data Center, RDC), corresponding provide respectively computing resource and interface are processed resource, according to practical engineering application collecting device parameter, in this example, only consider that installation resource consumption, processing resource, unit just dress up originally and quota, its design parameter numerical value is as shown in table 2.

Table 2: device parameter

In this example, establish according to system function requirement 3 class peripheral hardwares are provided altogether, 2 of every classes are symmetrical is arranged in fuselage both sides, peripheral hardware 1 and peripheral hardware 2 are approaching the position at aircraft middle part, peripheral hardware 3 approaches head, according to installing space position, peripheral hardware position, airframe and lead-in wire cabling requirement, try to achieve all kinds of outer wire length that are set to each installing space, according to practical engineering application, gather the parameter of peripheral hardware, only consider the weight of lead-in wire unit length and just dress up this in this example, the concrete numerical value of parameter is as shown in table 3.

Table 3: outer setting parameter

From equipment and installing space parameter, DIMA system comprises 4 installing spaces, i.e. N _loc=4, system provides maximum 6 CPM modules and 5 RDC modules altogether, is designated as respectively CPM1 to CPM6 and RDC1 to RDC5, i.e. N _dev=11, set up Physical layer mapping vector x _insfor N _loc× N _devthe vector of=44 dimensions, is designated as

If Physical layer two intrinsic constraint condition A that only consideration equipment is selected and installed in this example _{ins_sgl}x _ins≤ b _{ins_sgl}with $\left(\begin{array}{c}{A}_{\mathrm{slot}}\\ A_{\mathrm{mass}}\\ A_{\mathrm{cool}}\end{array}\right)x_{\mathrm{ins}}\&le;\left(\begin{array}{c}{b}_{\mathrm{slot}}\\ b_{\mathrm{mass}}\\ b_{\mathrm{cool}}\end{array}\right),$ Wherein

, I _loc=[1,1,1,1],

a _slotbe the resource constraint matrix of 4 × 44 backboard slot, backboard slot resource consumption when element representation equipment is wherein installed to installing space, b _slotbe 4 dimensional vectors, the backboard slot resource that element representation installing space wherein provides, take driving cabin top installing space as example, i.e. A _slotthe 1st behavior

B _slotfirst element be 30.Other elements in constraint inequality can this similarly provide for example.

So far, the Physical layer modeling flow process in this example completes.

Complete functional layer modeling below, first according to practical engineering application, set up the tree structure of system-task-subtask, in this example, adopt system-task-subtask tree as shown in Figure 4, and be set in Task Tree in the follow-up Optimizing Flow of this example and can not change, processing resource consumption according to practical engineering application to each subtask and the required peripheral hardware calling carry out assignment, and design parameter is as shown in table 4.

Table 4: task parameters

According to the type of processing resource, in this example, comprise altogether 2 class subtasks, correspond respectively on CPM and RDC and move, be designated as respectively and calculate subtask and interface subtask.In this example, DIMA system comprises 6 calculating subtasks altogether, 6 interface subtasks, and wherein 2 tasks are to calculate subtask and interface subtask simultaneously, simultaneously because Physical layer in this example provides 6 CPM and 5 RDC equipment altogether, functional layer mapping x _prosbe the vector of 6 × 6+6 × 5=66 dimension,

Wherein subscript TaskC1 to TaskC6 represents that 6 are calculated subtask, and TaskR1 to TaskR6 represents 6 interface subtasks.

The inherent constraint condition A of functional layer _{pros_sgl}x _pros=b _{pros_sgl}middle A _{pros_sgl}and b _{pros_sgl}be similar to the realization of Physical layer inherent constraint in this example, process resource constraint $\left(\begin{array}{c}{A}_{\mathrm{CPM}}\\ A_{\mathrm{RDC}}\end{array}\right)x_{\mathrm{pros}}\&le;\left(\begin{array}{c}{b}_{\mathrm{CPM}}\\ b_{\mathrm{RDC}}\end{array}\right)$ Middle A _cPMand A _rDCelement be respectively computational resource and the interface resource consumption of task, be similar to the realization of Physical layer in this example, b _cPMand b _rDCthe upper limit of middle processing resource is shone upon x by Physical layer _insdetermine, this example represents with undetermined parameter in implementing.

In addition, establish according to practical engineering application and require subtask corresponding in task 0A and task 0B, task 1A and task 1B to keep apart, for example subtask 0A_0 and subtask 1A_0 require to move on different CPM equipment, and this additional constraint is by inequality A _{pros_seg}x _pros≤ b _{pros_seg}provide, wherein A _{pros_seg}element in the position assignment that may occur above-mentioned conflict, be 1, b _{pros_seg}=[1,1 ... 1] ^t.

So far, in this example, the modeling of DIMA systemic-function layer completes.

Performance index in this example are selected the load of 1 reliability index computing module, are designated as f ₁(x), it defines as previously mentioned, and 2 Validity Index weight and just dress up this are designated as respectively f ₂and f (x) ₃(x), just dress up this definition as previously mentioned, weight indicator is similar just to be dressed up and originally comprises weight of equipment and lead-in wire weight and only will just fill cost parameter and be changed to weight parameter.Wherein computing module load is in higher prior level, and gives " important " priority flag, weight and just dress up this in lower limited level, and give respectively the priority flag of " extremely important " and " important ", the non-increasing sequence of all indexs is f ₁(x), f ₂(x), f ₃(x).

So far, the multiple objective function that in this example, DIMA has a hierarchical priority has been set up.

If the Physical layer of existing a set of DIMA system design scheme is 2 the 1st type CPM(CPM1 and CPM2) be installed in middle part electronic compartment, 1 the 1st type RDC equipment (RDC2) is arranged on afterbody electronic compartment, 1 the 2nd type RDC equipment (RDC1) is arranged on driving cabin top, accordingly by x _insassignment is if this scheme is in functional layer at the Task Tree adopting as shown in Figure 4, it is as shown in table 5 that its subtask is assigned to equipment scheme, corresponding x _proscarry out corresponding assignment x _pros=[1,0 ..., 0,0,1,0 ..., 0 ... ] ^t.

By the x of assignment _insand x _prosthe performance index function that can utilize a flow process to set up is assessed this DIMA system, calculate this system-computed module load 258.01, weight 204, just dress up basis 2494, this performance index value can compare with the reference performance index of having stored before this, or using this index as being increased in storage with reference to index, in this example, below this index is called to reference index.

Table 5: with reference to task allocation example

Below to this DIMA system with computing module load, weight and installation fee with and respective priority be labeled as objective function and carry out multiple-objection optimization, in this example, select the tolerant layer sorting method of foregoing band to solve, and by decision maker, set computing module load according to practical engineering application, the tolerance of weight, it in this example, is the impact of explanation tolerance, provide the solving result that two groups of different tolerances are set, set computing module load for first group, the tolerance of weight is respectively 5% and 10%, set computing module load for second group, the tolerance of weight is 15%.The layer sorting method tolerant through foregoing band solves multi-objective optimization question, tries to achieve optimal design, its performance index and as shown in table 6 to the normalized result of above-mentioned reference index.Optimization design scheme is to a certain degree being gained in weight and is just being dressed up this in the situation that, larger improvement the higher computing module loading index of priority, meet expected design, two optimal design indexs have also reflected the setting that provides tolerance for different practical engineering application.The specific implementation that reaches optimal design index that calculating provides is exported by Physical layer and functional layer mapping x, then by the mapping relations output of x, is realized the scheme of equipment selection, installation, Task Tree and the task distribution of optimal design index.A system design scheme that reaches optimal design index of exporting in this example is provided by table 7 and table 8.

Table 6: optimal design performance index

Table 7 a: optimal device is selected and mount scheme

Table 8 a: optimal scheduling scheme

In the description of this instructions, the description of reference term " embodiment ", " some embodiment ", " example ", " concrete example " or " some examples " etc. means to be contained at least one embodiment of the present invention or example in conjunction with specific features, structure, material or the feature of this embodiment or example description.In this manual, the schematic statement of above-mentioned term is not necessarily referred to identical embodiment or example.And specific features, structure, material or the feature of description can be with suitable mode combination in any one or more embodiment or example.

Although illustrated and described embodiments of the invention above, be understandable that, above-described embodiment is exemplary, can not be interpreted as limitation of the present invention, those of ordinary skill in the art can change above-described embodiment within the scope of the invention in the situation that not departing from principle of the present invention and aim, modification, replacement and modification.

Claims (6)

1. a distributed comprehensively modularized avionics DIMA system evaluation method, is characterized in that, comprises the following steps:

Carry out Physical layer modeling, the hardware device in described DIMA system is mapped in installing space, to create the Physical layer mapping vector x of binary _ins;

Carry out functional layer modeling, according to known system-task-subtask tree, the subtask in described DIMA system is mapped in described hardware device, to create the functional layer mapping vector x of binary _pros;

Multiple Performance Evaluating Indexes are carried out to prioritization, and set up the multiobject performance specification function set of DIMA system { f _goal1(x _ins, x _pros), f _goal2(x _ins, x _pros) ... f _goalNum(x _ins, x _pros), subscript goalNum represents target number;

By described Physical layer mapping vector x _inswith functional layer mapping vector x _prosthe multiobject performance specification function set of DIMA system { f described in substitution _goal1(x _ins, x _pros), f _goal2(x _ins, x _pros) ... f _goalNum(x _ins, x _pros) obtain the assessment result of DIMA system.

2. distributed comprehensively modularized avionics DIMA system evaluation method according to claim 1, is characterized in that, described Physical layer mapping vector $x_{\mathrm{ins}}={[x_{D_1,L_1},x_{D_1,L_2},...,x_{D_1,L_{N_{\mathrm{Loc}}}},...,x_{D_{N_{\mathrm{Dev}}},L_{N_1}},...,x_{D_{N_{\mathrm{Dev}}},L_{N_{\mathrm{Loc}}}}]}^T,$ Wherein N _locfor the number of described installing space, N _devfor the number of described hardware device, when i described hardware device is mapped in j described installing space, $x_{D_i,L_j}=1,$ Otherwise $x_{D_i,L_j}=0.$

3. distributed comprehensively modularized avionics DIMA system evaluation method according to claim 1 and 2, is characterized in that, described Physical layer mapping vector x _insneed to meet following two constraint conditions:

The first constraint condition is that same hardware device can not be installed in multiple installing spaces simultaneously, and mathematic(al) representation is: A _{ins_sgl}x _ins≤ b _{ins_sgl},

Wherein

b _{ins_sgl}=[1,1 ..., 1] ^tand be 1 × N _devrank matrix, I _loc=[1,1 ..., 1] and be N _loc× 1 rank matrix;

The second constraint condition is no more than the installation resource that this installing space provides for the installation resource of all hardware devices consume in same installing space, and mathematic(al) representation is: A _{ins_rsc}x _ins≤ b _{ins_rsc},

Wherein $A_{\mathrm{ins}\_\mathrm{rsc}}=\left(\begin{array}{c}{A}_{\mathrm{ins}\_\mathrm{rscl}}\\ .\\ .\\ .\\ A_{\mathrm{ins}\_\mathrm{rscN}}\end{array}\right),b_{\mathrm{ins}\_\mathrm{rsc}}=\left(\begin{array}{c}{b}_{\mathrm{ins}\_\mathrm{rscl}}\\ .\\ .\\ .\\ b_{\mathrm{ins}\_\mathrm{rscN}}\end{array}\right),$ Subscript ins_rsc1 to ins_rscN represents different installation resource types, take ins_rscx represent x class installation resource wherein 1≤x≤N and x as integer, $A_{\mathrm{ins}\_\mathrm{rscx}}=\left[\begin{array}{cccc}{s}_{D_1,L_1},0,...,0& s_{D_2,L_1},0,...,0& ...& s_{D_{N_{\mathrm{Dev}}},L_1},0,...,0\\ 0,s_{D_1,L_2},...,0& 0,s_{D_2,L_2},...,0& ...& 0,s_{D_{\mathrm{Dev}},L_2},...,0\\ ...& ...& ...& ...\\ 0,...,s_{D_1,L_{N_{\mathrm{Loc}}}}& 0,...,s_{D_2,L_{N_{\mathrm{Loc}}}}& ...& 0,...,s_{D_{N_{\mathrm{Dev}}},L_{N_{\mathrm{Loc}}}}\end{array}\right],$ $b_{\mathrm{ins}\_\mathrm{rscx}}={[r_s^{\left(L_1\right)},r_s^{\left(L_2\right)},...,r_s^{\left(L_{N_{\mathrm{Loc}}}\right)}]}^T,$ A _{ins_rscx}in element be the x class installation resource that i equipment consumes while being installed to j installing space, b _{ins_rscx}in element

represent the x class installation resource that j installing space provides.

4. according to the distributed comprehensively modularized avionics DIMA system evaluation method described in any one in claim 1-3, it is characterized in that, describedly carry out functional layer modeling, according to known system-task-subtask, set the duty mapping in described DIMA system in described hardware device, to create the functional layer mapping vector x of binary _prosspecifically comprise:

According to system-task-subtask tree, determine that the processing resource type that described DIMA system needs is altogether pros_rscN kind, and determine that it is N that consumption n class is processed the n class subtask number of resource _tasknindividual, wherein 1≤n≤pros_rscN and n are integer;

According to described x _insand the self property of hardware device, the number that is identified for providing n class to process the n kind equipment of resource is N _devnindividual, wherein 1≤n≤pros_rscN and n are integer;

Described functional layer mapping vector

wherein part corresponding to n class subtask is $x_n={[x_{T_n^{\left(1\right)},D_n^{\left(1\right)}},x_{T_n^{\left(1\right)},D_n^{\left(2\right)}},...,x_{T_n^{\left(1\right)},D_n^{\left(N_{\mathrm{Devn}}\right)}},...,x_{T_n^{\left(N_{\mathrm{Taskn}}\right)},D_n^{\left(1\right)}},...,x_{T_n^{\left(N_{\mathrm{Taskn}}\right)},D_n^{\left(N_{\mathrm{Devn}}\right)}}]}^T,$ If t n class subtask is assigned on k n kind equipment, move, remember x _nin element

otherwise

wherein 1≤t≤N _tasknand t is integer, 1≤k≤N _devnand k is integer.

5. according to the distributed comprehensively modularized avionics DIMA system evaluation method described in any one in claim 1-4, it is characterized in that described functional layer mapping vector x _prosneed to meet following two constraint conditions,

The 3rd constraint condition be same n class subtask indivisible, only can be assigned on a n kind equipment and complete, and there is not the situation of Lou selecting subtask, mathematic(al) representation is: A _{pros_sgl}x _pros=b _{pros_sgl}

Wherein

i _n=[1,1 ..., 1] and be N _devn× 1 rank matrix, b _{pros_sgl}=[1,1 ..., 1] ^tand be

rank matrix;

The 4th constraint condition is no more than the processing resource that this hardware device provides for the summation of processing resource consumption corresponding to the subtask moved in same hardware device, and mathematic(al) representation is: A _{pros_rsc}x _pros≤ b _{pros_rsc}, wherein b _{pros_rsc}in the upper limit of certain class resource select in being shone upon by Physical layer and the summation of such resource that this kind equipment of installing provides determines, by described x _insdetermine.

6. a distributed comprehensively modularized avionics DIMA system optimization method, is characterized in that, comprises the following steps:

Propose the initialization scheme of distributed comprehensively modularized avionics DIMA system, and according to the appraisal procedure described in any one in claim 1-5, calculate the assessment result of initialization scheme;

Multiple performance parameters are carried out to the order of priority of non-increasing, the performance parameter that priority resolution is high is optimization direction;

On the basis of described initialization scheme, according to the layer sorting method with tolerant, change multiple-objection optimization into single goal optimization and dwindle gradually optimal solution set to unique value or all objective optimizations and solve complete.

Images