CN104360910A - Equipment distribution method for detecting equipment network on basis of particle swarm algorithm - Google Patents

Abstract

The invention relates to an equipment distribution method for a detecting equipment network on the basis of a particle swarm algorithm. The method comprises the following steps that 1, a monitoring task model among detecting equipment, a detecting target and a deterring time interval is built, detectable monitoring task sequences of equipment are generated, and in addition, simulation parameters are initialized; 2, the corresponding relationship between each particle and the detectable monitoring task sequences of the equipment is built, and in addition, the time interval of the detecting task is regulated according to elements of each particle; 3, the fitness function of each particle is calculated according to the mapping relationship between each particle and a monitoring task, and in addition, the local optimal solution and the global optimal solution of a particle swarm are updated; 4, whether the number of iterations reaches preset conditions or not is judged, if so, the monitoring task sequence corresponding to the global optimal particle is output, if not, the parameters of the particle swarm are updated, and the second to fourth steps are repeated until the conditions are met. The equipment distribution method has the advantages that the problem of task distribution when a plurality of pieces of detecting equipment are networked for monitoring a plurality of targets is effectively solved, and the optimization of the equipment work state of the detecting equipment network is realized.

Description

Based on the device allocation method of the detecting devices net of particle cluster algorithm

Technical field

The present invention relates to Space Object Detection technical field, relate more specifically to a kind of device allocation method of the detecting devices net based on particle cluster algorithm, the optimization of device resource when being applicable to the multiple extraterrestrial target of many equipment combined detection distributes.

Background technology

What detecting devices distribution technique solved is how to optimize the detection resource allocation problem distributing the multiple extraterrestrial target of multiple detecting devices networking combined detection.Over nearly ten or twenty year, extraterrestrial target quantity sharply increases, and causes great threat to the safety of spacecraft in-orbit.In order to ensure the safety of spacecraft in-orbit, detecting devices is on the increase, and needs each equipment of Optimized Operation, gives full play to the detection performance of each equipment, the more target in-orbit of detection.Being optimized scheduling to the detectable situation of all devices is a discrete combination problem, and when equipment and destination number increase, one-tenth geometric series increases and becomes a unsolvable problem of classic method by this optimization problem complexity.

Particle group optimizing (Particle Swarm Optimization, PSO) be a kind of new optimum theory occurred in recent years, the method to be looked for food migrating with clustering behavior and a kind of global random searching algorithm based on swarm intelligence of proposing (for example, see article " Particle swarm optimization " in process by simulation flock of birds, publish in Proceedings of the 1995IEEE International Conference on Neural Networks, 1995:1942-1948).

The continuity of particle cluster algorithm to optimization object function and problem definition does not have particular/special requirement, makes this algorithm application more extensive; There is no center control constraints, more have robustness; Adopt the information sharing mode of non-immediate to realize cooperation, have more extendibility; Random search essence, makes it more difficultly be absorbed in local optimum.Therefore, particle cluster algorithm is for complexity, and particularly the optimization computational problem of multimodal higher-dimension has very strong superiority.

The feature that particle cluster algorithm has easy understanding, easily realizes, all has stronger ability of searching optimum to non-linear, multiple peak problem, gets the attention when processing some optimization problems in scientific research and engineering practice.How being applied in the scheduling of detecting devices monitoring task will be the technical issues that need to address of the present invention.

Summary of the invention

In view of this, one of fundamental purpose of the present invention is to provide a kind of detecting devices networking equipment distribution method based on particle cluster algorithm, so that when to the scheduling of monitoring task resource, realize obtaining faster the monitoring task scheduling result of optimization, the computation complexity overcoming conventional combination method is high and the optimization process of many equipment multi-objective problem is consumed to the problem of a large amount of time and computational resource deficiency.

For achieving the above object, technical solution of the present invention is:

Based on a device allocation method for the detecting devices net of particle cluster algorithm, it is characterized in that: comprise the steps:

Step 1, according to the detection relation of detecting devices to extraterrestrial target, set up detecting devices collection, monitoring task model between detection of a target collection and detection time Interval Set, generate the detectable monitoring task sequence of equipment, and initialization simulation parameter;

Step 2, set up the corresponding relation of the task sequence that each particle and equipment can be monitored, and the time interval of element adjustment detection mission according to each particle;

Step 3, calculate the fitness function of each particle according to particle each described in step 2 and the mapping relations of monitoring task, and upgrade locally optimal solution and the globally optimal solution of described population;

Step 4, judge whether iterations reaches and preset threshold value, as satisfied condition, exporting the monitoring task sequence that the globally optimal solution that calculates according to step 3 is corresponding, otherwise upgrading population parameter, and jumping to step 2.

Wherein, in described step 1, the set of definition detecting devices is A={ α ₁..., α _m, detection of a target set is B=(β ₁..., β _n, detectable time interval set is ∑=(ε ₁..., ε _k, time interval ε _k(1≤k≤K) is expressed as time period [s _k, e _k], the equipment that this time interval is corresponding and target are respectively α _{m, k}(α _{m, k}∈ A) and β _{n, k}(β _{n, k}∈ B), the interval corresponding monitoring set of tasks of raw definition uphole equipment, the detection of a target and detection time is Ω ₀={ ω ₁..., ω _k, wherein ω _k=(α _{m, k}, β _{n, k}, κ _k, η _k), κ _kand η _kbe respectively monitoring task ω _kstart time and end time, and definition particle vector X=(x ₁..., x _k), wherein x _k(1≤k≤K) is expressed as the execution preferred number of a kth monitoring task, definition particle rapidity vector V=(v ₁..., v _k); And initialization simulation parameter: iterations is chosen for K, and generate 5K particle, a 5K particle rapidity and 5K monitoring set of tasks one_to_one corresponding, and carry out initialization, other parameter initialization is particular value.

Wherein, in described step 2, set up the corresponding relation of the task sequence that the element of each particle and detecting devices can be monitored, and comprise according to the step that the time interval of parameter to detection mission of each particle adjusts:

1. choose the monitoring task that the sequence number of the element that kth is large in a particle is corresponding, is set to the start time of this time interval the start time of monitoring task, wherein k is positive integer, increases progressively from 1,1≤k≤K;

2. determine to monitor job end time according to the requirement of the minimum detection duration of monitoring task;

3. upgrade the start time and the end time that there are the time interval of common factor with the time interval chosen, form the new monitoring task sequence corresponding with this particle, continue next element repeat step 1. ~ 3. until all elements of this particle has all processed;

4. continue to choose next particle repeat step 1. ~ 3., until all particle process complete.

Wherein, in described step 2, in time interval adjustment process, the condition of demand fulfillment is:

A, each radar equipment can only monitor a target in-orbit simultaneously;

B, each monitoring task must perform in corresponding time interval, monitoring task ω _kthe time span of (1≤k≤K) can not be less than the shortest time requirement T (ω of task _k), and arbitrary time interval belongs to and only belongs to the interval subset of pot life of some monitoring tasks;

C, target beta _nmonitoring time length summation in (1≤n≤N) each time interval is not less than the shortest detection requirement μ of this target call _n.

Wherein, fitness function F (A, B, the Ω of described particle t in step 3 _t) computing method be:

F(A，B，Ω _t)＝D _eff，t·D _eqi，t(1)

Wherein D _{eff, t}for the monitoring task scheduling benefit that particle is corresponding, D _{eqi, t}for equipment use equilibrium degree in the monitoring task sequence that particle is corresponding, 1≤t≤5K, circular is as follows:

The monitoring task scheduling benefit D that a, particle t are corresponding _{eff, t}be defined as

$D_{\mathrm{eff},t}=\underset{{\mathrm{\β}}_n\∈B}{\mathrm{\Σ}}{\mathrm{\ρ}}_n\·f(\underset{{\mathrm{\ω}}_{t,k}{\∈\mathrm{\Ω}}_t}{\mathrm{\Σ}}\mathrm{\φ}({\mathrm{\ω}}_{t,k},{\mathrm{\β}}_n)({\mathrm{\η}}_{t,k}-{\mathrm{\κ}}_{t,k}),{\mathrm{\μ}}_n)---\left(2\right)$

In formula, ρ _nfor the preferred number of target beta n monitoring; ω _{t, k}it is the kth monitoring task in t monitoring series of task; φ (ω _{t, k}, β _n) be discriminant function, as monitoring task ω _{t, k}be used for monitoring objective β _ntime φ (ω _{t, k}, β _n)=1, otherwise φ (ω _{t, k}, β _n)=0; κ _{t, k}and η _{t, k}represent ω respectively _{t, k}start time and the end time; F (t, μ) represents monitoring task scheduling utility function, describes the integrated scheduling usefulness of actual execution time l under minimum monitoring time μ condition of target, is defined as

$f(l,\mathrm{\&mu;})=\left\{\begin{array}{cc}0& l<\mathrm{\&mu;}\\ 1+\mathrm{\&sigma;}\frac{l-\mathrm{\&mu;}}{\mathrm{\&mu;}}& l\&GreaterEqual;\mathrm{\&mu;}\end{array}\right.---\left(3\right)$

In formula, parameter σ is monitoring redundancy execution utilization coefficient, and span is [0,1];

Equipment use equilibrium degree D in the monitoring task sequence that b, particle t are corresponding _{eqi, t}be defined as

$D_{\mathrm{eqi},t}=\frac{{\left(\underset{{\mathrm{\&theta;}}_m>0}{\mathrm{\&Pi;}}(1+{\mathrm{\&theta;}}_m\&CenterDot;U_A({\mathrm{\&alpha;}}_m,{\mathrm{\&Omega;}}_t))\right)}^{\frac{1}{M_0}}}{{\left(\underset{{\mathrm{\&theta;}}_0>0}{\mathrm{\&Pi;}}(1+\frac{{\mathrm{\&theta;}}_m}{M_0}\&CenterDot;(1+\underset{{\mathrm{\&theta;}}_i>0}{\mathrm{\&Sigma;}}\frac{1}{{\mathrm{\&theta;}}_i}-\frac{M_0}{{\mathrm{\&theta;}}_m}))\right)}^{\frac{1}{M_0}}}---\left(4\right)$

In formula, θ _m(θ _m>=0) be the preferential coefficient of performance of uphole equipment, if θ _m=0, then represent the utilization factor not needing to consider this equipment; M ₀for equipment preferred number is greater than 0 equipment number, U _a(α _m, Ω _t) be at monitoring task Ω _tdistribution method under equipment α _mutilization factor, be defined as

$U_A({\mathrm{\&alpha;}}_m,{\mathrm{\&Omega;}}_t)=\mathrm{sign}\left({\mathrm{\&theta;}}_m\right)\&CenterDot;\frac{{\mathrm{\&Sigma;}}_{{\mathrm{\&omega;}}_{t,k}{\&Element;\mathrm{\&Omega;}}_t}\mathrm{\&psi;}({\mathrm{\&omega;}}_{t,k},{\mathrm{\&alpha;}}_m)({\mathrm{\&eta;}}_{t,k}-{\mathrm{\&kappa;}}_{t,k})}{{\mathrm{\&Sigma;}}_{\underset{{\mathrm{\&alpha;}}_i\&Element;A}{{\mathrm{\&theta;}}_i>0}{\mathrm{\&Sigma;}}_{{\mathrm{\&omega;}}_{t,k}{\&Element;\mathrm{\&Omega;}}_t}}\mathrm{\&psi;}({\mathrm{\&omega;}}_{t,k},{\mathrm{\&alpha;}}_i)\left({\mathrm{\&eta;}}_{t,k}{\mathrm{\&kappa;}}_{t,k}\right)}---\left(5\right)$

Wherein ψ (ω _{t, k}, α _m) be discriminant function, as monitoring task ω _{t, k}by equipment α _mψ (ω during monitoring _{t, k}, α _m)=1, otherwise ψ (ω _{t, k}, α _m)=0; Sign (.) is sign function, is defined as

$\mathrm{sign}\left(x\right)=\left\{\begin{array}{cc}1& x>0\\ 0& x=0\\ -1& x<0\end{array}\right.---\left(6\right)$

Wherein, in step 4 described in preset threshold value be the max calculation that presets time condition.

Known based on technique scheme, method of the present invention compared with prior art advantage is: the present invention makes full use of the feature that particle cluster algorithm can realize from calculating combinatorial problem optimizing fast, particle cluster algorithm is incorporated in the distribution of detecting devices networking equipment, the Optimized Operation of the device resource to multiple detecting devices networking can be realized fast, greatly improve equipment allocative efficiency.

Accompanying drawing explanation

Fig. 1 is the process flow diagram of the detecting devices networking equipment distribution method based on particle cluster algorithm of the present invention.

Embodiment

For making the object, technical solutions and advantages of the present invention clearly understand, below in conjunction with specific embodiment, and with reference to accompanying drawing, the present invention is described in further detail.

The invention provides a kind of detecting devices networking equipment distribution method based on particle cluster algorithm, utilize particle cluster algorithm theory can realize the feature of combinatorial problem optimizing fast, particle cluster algorithm is incorporated in detecting devices networking Optimized Operation.As shown in Figure 1, be the implementing procedure figure of a kind of detecting devices networking equipment distribution method based on particle cluster algorithm of the present invention, specifically comprise following 4 steps:

(1) defining detecting devices set is A={ α ₁..., α _m, detection of a target set is B={ β ₁..., β _n, detectable time interval set is ∑={ ε ₁..., ε _k, time interval ε _k(1≤k≤K) is expressed as time period [s _k, e _k], the equipment that this time interval is corresponding and target are respectively α _{m, k}(α _{m, k}∈ A) and β _{n, k}(β _n, _k∈ B), the interval corresponding monitoring set of tasks of definition uphole equipment, the detection of a target and detection time is Ω ₀={ ω ₁..., ω _k, wherein ω _k=(α _{m, k}, β _{n, k}, κ _k, η _k).Definition particle vector X=(x ₁..., x _k), wherein x _k(1≤k≤K) is expressed as the execution preferred number of a kth monitoring task, definition particle rapidity vector V=(v ₁..., v _k).Other simulation parameter of initialization: monitoring task shortest length collection is { T ₁..., T _k, target detection the shortest monitoring duration collection is { μ ₁..., μ _k, iterations is chosen for K, generates 5K particle { X ₁..., X _5K, a 5K particle rapidity { V ₁..., V _5Kand 5K monitoring set of tasks { Ω ₁..., Ω _5Kone_to_one corresponding, by t (1≤t≤5K) individual particle X _twith speed V _tbe initialized as $X_t=\{x_{t,1}^0,...,x_{t,K}^0\}$ With $V_t=\{v_{t,1}^0,...,v_{t,K}^0\},$ Wherein with $v_{t,k}^0(1\&le;k\&le;K)$ All be set to equally distributed random number between (0,1), and t (1≤t≤5K) individual monitoring series of task is initialized as Ω _t=Ω ₀, the preferred number set { ρ of each target monitoring of initialization ₁..., ρ _n, the preferential coefficient of performance set { θ of initialization apparatus ₁..., θ _m;

(2) carry out adjustment according to the time interval of element to detection of each particle and upgrade monitoring task, in time interval adjustment process, need the several supposed premise conditions set:

A, each radar equipment can only detect a target in-orbit simultaneously;

The monitoring task of b, t (1≤t≤5K) individual particle must perform in corresponding time interval, monitoring task ω _{t, k}the time span of (1≤k≤K) can not be less than T _k, and arbitrary time interval belongs to and only belongs to the interval subset of pot life of some monitoring tasks;

C, target beta _nmonitoring time length summation in (1≤n≤N) each time interval is not less than the shortest detection requirement of this target call _{μ n}.

Determine that the time interval of the element of each particle to detection carries out adjustment and upgrade monitoring task agent containing 4 steps:

1. the sequence number i of the element that kth is large in t particle is chosen _{s, k}corresponding time interval adjust, wherein t and k is positive integer, increases progressively from 1,1≤t≤5K, 1≤k≤K;

2. task is monitored the corresponding start time and the end time determination be divided into following three steps:

1) task is monitored start time be set to time interval start time, namely

$k_{i_{t,k}}=s_{i_{t,k}};$

2) job end time timing is really monitored, to ensure that this monitoring task execution time is long as far as possible on the one hand, after ensureing to distribute this monitoring task simultaneously, time interval as much as possible is had to carry out monitoring task in the time interval set of device conflict or goal conflict, so the end time of monitoring task is

${\mathrm{\&eta;}}_{i_{t,k}}=\min \{e_{i_{t,k}},\min \{e_h-T_h{|}_{e_h-T_h\&GreaterEqual;{\mathrm{\&kappa;}}_h+T_h}^{{\mathrm{\&epsiv;}}_h{\&Element;\mathrm{\&Gamma;}}_{i_{t,k}}}\left\}\right\}---\left(1\right)$

Wherein, for the time end time, represent monitoring series of task Ω _tin with time interval there is the set of occuring simultaneously and belonging to the time interval of same equipment;

3) update time Interval Set in start time of all time intervals and end time.

3. upgrade the start time and the end time that there are the time interval of common factor with the time interval chosen, form the new monitoring task sequence corresponding with t particle, continue next element repeat step 1. ~ 3. until all elements of this particle has all processed;

4. continue to choose next particle repeat step 1. ~ 3., until all particle process complete.

(3) according to the mapping relations of each particle in step (2) with monitoring task, the implementation procedure calculating the fitness function of t (1≤t≤5K) individual particle is divided into four steps:

1. the monitoring task scheduling benefit D that this particle t is corresponding is calculated _{eff, t}:

$D_{\mathrm{eff},t}=\underset{{\mathrm{\&beta;}}_n\&Element;B}{\mathrm{\&Sigma;}}{\mathrm{\&rho;}}_n\&CenterDot;f(\underset{{\mathrm{\&omega;}}_{t,k}{\&Element;\mathrm{\&Omega;}}_t}{\mathrm{\&Sigma;}}\mathrm{\&phi;}({\mathrm{\&omega;}}_{t,k},{\mathrm{\&beta;}}_n)({\mathrm{\&eta;}}_{t,k}-{\mathrm{\&kappa;}}_{t,k}),{\mathrm{\&mu;}}_n)---\left(2\right)$

In formula, ρ _nfor target beta _nthe preferred number of monitoring; φ (ω _{t, k}, β _n) be discriminant function, as monitoring task ω _{t, k}be used for monitoring objective β _ntime φ (ω _{t, k}, β _n)=1, otherwise φ (ω _{t, k}, β _n)=0; F (t, μ) represents monitoring task scheduling utility function, and describe the actual execution time l of target, the integrated scheduling usefulness under minimum monitoring time μ condition, is defined as

$f(l,\mathrm{\&mu;})=\left\{\begin{array}{cc}0& l<\mathrm{\&mu;}\\ 1+\mathrm{\&sigma;}\frac{l-\mathrm{\&mu;}}{\mathrm{\&mu;}}& l\&GreaterEqual;\mathrm{\&mu;}\end{array}\right.---\left(3\right)$

In formula, parameter σ is monitoring redundancy execution utilization coefficient, and span is [0,1];

2. equipment use equilibrium degree D in monitoring task sequence corresponding to particle is calculated _{eqi, t}:

$D_{\mathrm{eqi},t}=\frac{{\left(\underset{{\mathrm{\&theta;}}_m>0}{\mathrm{\&Pi;}}(1+{\mathrm{\&theta;}}_m\&CenterDot;U_A({\mathrm{\&alpha;}}_m,{\mathrm{\&Omega;}}_t))\right)}^{\frac{1}{M_0}}}{{\left(\underset{{\mathrm{\&theta;}}_0>0}{\mathrm{\&Pi;}}(1+\frac{{\mathrm{\&theta;}}_m}{M_0}\&CenterDot;(1+\underset{{\mathrm{\&theta;}}_i>0}{\mathrm{\&Sigma;}}\frac{1}{{\mathrm{\&theta;}}_i}-\frac{M_0}{{\mathrm{\&theta;}}_m}))\right)}^{\frac{1}{M_0}}}---\left(4\right)$

In formula, θ _m(θ _m>=0) be uphole equipment α _m(1≤m≤M) preferential coefficient of performance, if ρ _m=0, then represent the utilization factor not needing to consider this equipment; M ₀for equipment preferred number is greater than 0 equipment number, U _a(α _m, Ω _t) be at monitoring task Ω _tdistribution method under equipment α _mutilization factor, be defined as

$U_A({\mathrm{\&alpha;}}_m,{\mathrm{\&Omega;}}_t)=\mathrm{sign}\left({\mathrm{\&theta;}}_m\right)\&CenterDot;\frac{{\mathrm{\&Sigma;}}_{{\mathrm{\&omega;}}_{t,k}{\&Element;\mathrm{\&Omega;}}_t}\mathrm{\&psi;}({\mathrm{\&omega;}}_{t,k},{\mathrm{\&alpha;}}_m)({\mathrm{\&eta;}}_{t,k}-{\mathrm{\&kappa;}}_{t,k})}{{\mathrm{\&Sigma;}}_{\underset{{\mathrm{\&alpha;}}_i\&Element;A}{{\mathrm{\&theta;}}_i>0}{\mathrm{\&Sigma;}}_{{\mathrm{\&omega;}}_{t,k}{\&Element;\mathrm{\&Omega;}}_t}}\mathrm{\&psi;}({\mathrm{\&omega;}}_{t,k},{\mathrm{\&alpha;}}_i)\left({\mathrm{\&eta;}}_{t,k}{\mathrm{\&kappa;}}_{t,k}\right)}---\left(5\right)$

Wherein ψ (ω _{t, k}, α _m) be discriminant function, as monitoring task ω _{t, k}by equipment α _mψ (ω during monitoring _{t, k}, α _m)=1, otherwise ψ (ω _{t, k}, α _m)=0; Sign (.) is sign function, is defined as

$\mathrm{sign}\left(x\right)=\left\{\begin{array}{cc}1& x>0\\ 0& x=0\\ -1& x<0\end{array}\right.---\left(6\right)$

3. fitness value V α lue corresponding to this particle is calculated according to fitness function _t, fitness function F (A, B, Ω _t) be defined as:

Vαlue _t＝F(A，B，Ω _t)＝D _eff·De _qi(7)

4. according to the fitness value of all calculating particles, the particle X that in current iteration calculating, in all particles, fitness value is maximum is obtained _{iter, max}with the monitoring series of task Ω of correspondence _{iter, max}, and then count on the maximum particle X of fitness so far _{g, max}with the monitoring series of task Ω of correspondence _{g, max}, wherein X _{iter, max}and X _{g, max}be respectively locally optimal solution and the globally optimal solution of population.

(4) according to the result of calculation of step (3), judge whether iterations iter reaches the threshold value preset, such as max calculation time condition, if do not met, upgrades speed and the particle element of t (1≤t≤5K) individual particle by following formula:

$\begin{array}{c}{V}_t^{\mathrm{iter}+1}=\mathrm{\&chi;}(V_t^{\mathrm{iter}}+c_1\&times;r_1\&times;(X_{\mathrm{iter},\max }-X_t^{\mathrm{iter}})+c_2\&times;r_2\&times;(X_{g,\max }-X_t^{\mathrm{iter}}))\\ x_t^{\mathrm{iter}+1}=X_t^{\mathrm{iter}}+V_t^{\mathrm{iter}+1}\end{array}---\left(8\right)$

In formula, c ₁and c ₂for constant, generally get c ₁=c ₂=2.05, r ₁and r ₂be the random number between 0 ~ 1, χ is contraction factor, is defined as

$\mathrm{\&chi;}=\frac{2}{|2-\mathrm{\&xi;}-\sqrt{{\mathrm{\&xi;}}^2-4\mathrm{\&delta;}}|}---\left(9\right)$

Wherein ξ=c ₁+ c ₂>=4.

Then repeating step (2) ~ (4) until satisfy condition, then exporting monitoring task sequence corresponding to global optimum's particle as satisfied condition, the final Optimized Operation realized detecting devices networking monitoring task.

Below by emulation, the inventive method is verified.What emulate in experiment is two multiple targets in-orbit of detecting devices detection, equipment 1 is positioned at north latitude 40 degree of east longitudes 116 degree, equipment 2 is positioned at north latitude 30 degree of east longitudes 106 degree, investigative range is 60 ° ~ 180 °, orientation, 30 ° ~ 60 °, latitude, the detection of a target is that the target RCS of orbit altitude below 2000 kilometers is greater than 1m ²satellite in orbit, what orbital tracking was selected is the Two-type line that August 18, NASA provided, by pass by analyze obtain two equipment in 1800 seconds of UTC time 2014-08-1800:00:00 ~ 2014-08-1800:30:00 detectable target information respectively in table 1 and 2, have 43 targets can be detected, if each target cumulative time length is no less than 60 seconds, in each time interval, the time span of monitoring is no less than 60 seconds, the preferred number of target 25676 is set to 5, the preferred number of other 42 targets is all set to 1, the preferential coefficient of performance of two detecting devicess is 1, iteration of simulations number of times is set to 100 times.Table 3 and table 4 be not for optimize, and equipment 1 and equipment 2 directly carry out the result of following the tracks of according to time sequencing, and table 5 and table 6 are respectively the result of detection of the equipment 1 after networking optimization and equipment 2.

The detectable target information of table 1 equipment 1

The detectable target information of table 2 equipment 2

The detectable target information of table 3 subsequent detection equipment 1

The detectable target information of table 4 subsequent detection equipment 2

The detection information of equipment 1 after table 5 two equipment network optimizations

The detection information of equipment 2 after table 6 two equipment network optimizations

As can be seen from the contrast of table 3 above and table 5 and table 4 and table 6: detecting devices 1, do not optimize and adopt in chronological order detection time can only detect 12 targets, 19 targets can be detected after optimization, detecting devices 2 do not optimize and adopt in chronological order detection time can only detect 10 targets, 12 targets can be detected after optimization, illustrate that the detection performance of individual equipment has had obvious lifting; Consider two equipment, detectable 21 targets altogether can be found out before optimization, and 30 targets can be detected after optimizing, so be a kind of algorithm that can be used for actual multiple equipment network equipment and distribute based on the detecting devices networking equipment distribution method of particle cluster algorithm.

Above-described specific embodiment; object of the present invention, technical scheme and beneficial effect are further described; be understood that; the foregoing is only specific embodiments of the invention; be not limited to the present invention; within the spirit and principles in the present invention all, any amendment made, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (6)

1., based on a device allocation method for the detecting devices net of particle cluster algorithm, it is characterized in that: comprise the steps:

Step 1, according to the detection relation of detecting devices to extraterrestrial target, set up detecting devices collection, monitoring task model between detection of a target collection and detection time Interval Set, generate the detectable monitoring task sequence of equipment, and initialization simulation parameter;

Step 2, set up the corresponding relation of the task sequence that each particle and equipment can be monitored, and the time interval of element adjustment detection mission according to each particle;

Step 3, calculate the fitness function of each particle according to particle each described in step 2 and the mapping relations of monitoring task, and upgrade locally optimal solution and the globally optimal solution of described population;

Step 4, judge whether iterations reaches and preset threshold value, as satisfied condition, exporting the monitoring task sequence that the globally optimal solution that calculates according to step 3 is corresponding, otherwise upgrading population parameter, and jumping to step 2.

2. the device allocation method of the detecting devices net based on particle cluster algorithm according to claim 1, is characterized in that, in described step 1, the set of definition detecting devices is A={ α ₁..., α _m, detection of a target set is B={ β ₁..., β _n, detectable time interval set is ∑={ ε ₁..., ε _k, time interval ε _hbe expressed as time period [s _k, e _k], the equipment that this time interval is corresponding and target are respectively α _{m, k}(α _{m, k}∈ A) and β _{n, k}(β _{n, k}∈ B), the interval corresponding monitoring set of tasks of definition uphole equipment, the detection of a target and detection time is Ω ₀={ ω ₁..., ω _k, wherein 1≤k≤K, ω _k=(α _{m, k}, β _{n, k}, k _k, η _k), k _kand η _kbe respectively monitoring task ω _kstart time and end time, and definition particle vector X=(x ₁..., x _k), wherein x _hbe expressed as the execution preferred number of a kth monitoring task, 1≤k≤K, definition particle rapidity vector V=(v ₁..., v _k); And initialization simulation parameter: iterations is chosen for K, and generate 5K particle, a 5K particle rapidity and 5K monitoring set of tasks one_to_one corresponding, and carry out initialization, other parameter initialization is particular value.

3. the device allocation method of the detecting devices net based on particle cluster algorithm according to claim 1, it is characterized in that, in described step 2, set up the corresponding relation of the task sequence that the element of each particle and detecting devices can be monitored, and comprise according to the step that the time interval of parameter to detection mission of each particle adjusts:

1. choose the monitoring task that the sequence number of the element that kth is large in a particle is corresponding, is set to the start time of this time interval the start time of monitoring task, wherein k is positive integer, increases progressively from 1,1≤k≤K;

2. determine to monitor job end time according to the requirement of the minimum detection duration of monitoring task;

3. upgrade the start time and the end time that there are the time interval of common factor with the time interval chosen, form the new monitoring task sequence corresponding with this particle, continue next element repeat step 1. ~ 3. until all elements of this particle has all processed;

4. continue to choose next particle repeat step 1. ~ 3., until all particle process complete.

4. the device allocation method of the detecting devices net based on particle cluster algorithm according to claim 1, is characterized in that, in described step 2, in time interval adjustment process, the condition of demand fulfillment is:

A, each radar equipment can only monitor a target in-orbit simultaneously;

B, each monitoring task must perform in corresponding time interval, monitoring task ω _ktime span shortest time that can not be less than task require in (ω _k), and arbitrary time interval belongs to and only belongs to the interval subset of pot life of some monitoring tasks, wherein k is positive integer, 1≤k≤K;

C, target beta _nmonitoring time length summation in each time interval is not less than the shortest detection requirement μ of this target call _n, wherein 1≤n≤N.

5. the device allocation method of the detecting devices net based on particle cluster algorithm according to claim 1, is characterized in that, in step 3 fitness function F (A, B, the Ω of described particle t _t) computing method be:

F(A，B，Ω _t)＝D _eff，t·D _eqi，t(1)

Wherein D _{eff, t}for the monitoring task scheduling benefit that particle is corresponding, D _{eqi, t}for the monitoring task sequence that particle is corresponding

Equipment use equilibrium degree in row, 1≤t≤5K, circular is as follows:

The monitoring task scheduling benefit D that a, particle t are corresponding _{eff, t}be defined as

$D_{\mathrm{eff},t}=\underset{{\mathrm{\&beta;}}_n\&Element;B}{\mathrm{\&Sigma;}}{\mathrm{\&rho;}}_n\&CenterDot;f(\underset{{\mathrm{\&omega;}}_{t,k}\&Element;{\mathrm{\&Omega;}}_t}{\mathrm{\&Sigma;}}\mathrm{\&phi;}({\mathrm{\&omega;}}_{t,k},{\mathrm{\&beta;}}_n)({\mathrm{\&eta;}}_{t,k}-k_{t,k}),{\mathrm{\&mu;}}_n)---\left(2\right)$

In formula, ρ _nfor target beta _nthe preferred number of monitoring; ω _{t, k}it is the kth monitoring task in t monitoring series of task; φ (ω _{t, k}, β _n) be discriminant function, as monitoring task ω _{t, k}be used for monitoring objective β _ntime φ (ω _{t, k}, β _n)=1, otherwise φ (ω _{t, k}, β _n)=0; k _{t, k}and η _{t, k}represent ω respectively _{t, k}start time and the end time; F (t, μ) represents monitoring task scheduling utility function, describes the integrated scheduling usefulness of actual execution time l under minimum monitoring time μ condition of target, is defined as

$f(l,\mathrm{\&mu;})=\left\{\begin{array}{cc}0& l<\mathrm{\&mu;}\\ 1+\mathrm{\&sigma;}\frac{l-\mathrm{\&mu;}}{\mathrm{\&mu;}}& l\&GreaterEqual;\mathrm{\&mu;}\end{array}\right.---\left(3\right)$

In formula, parameter σ is monitoring redundancy execution utilization coefficient, and span is [0,1];

Equipment use equilibrium degree D in the monitoring task sequence that b, particle t are corresponding _{eqi, t}be defined as

$D_{\mathrm{eqi},t}=\frac{{\left(\underset{{\mathrm{\&theta;}}_m>0}{\mathrm{\&Pi;}}(1+{\mathrm{\&theta;}}_m\&CenterDot;U_A({\mathrm{\&alpha;}}_m,{\mathrm{\&Omega;}}_t))\right)}^{\frac{1}{M_0}}}{{\left(\underset{{\mathrm{\&theta;}}_m>0}{\mathrm{\&Pi;}}(1+\frac{{\mathrm{\&theta;}}_m}{M_0}\&CenterDot;(1+\underset{{\mathrm{\&theta;}}_i>0}{\mathrm{\&Sigma;}}\frac{1}{{\mathrm{\&theta;}}_i}-\frac{M_0}{{\mathrm{\&theta;}}_m}))\right)}^{\frac{1}{M_0}}}---\left(4\right)$

In formula, θ _m(θ _m>=0) be the preferential coefficient of performance of uphole equipment, if θ _m=0, then represent the utilization factor not needing to consider this equipment; M ₀for equipment preferred number is greater than 0 equipment number, U _a(α _m, Ω _t) be at monitoring task Ω _tdistribution method under equipment α _mutilization factor, be defined as

$U_A({\mathrm{\&alpha;}}_m,{\mathrm{\&Omega;}}_t)=\mathrm{sign}\left({\mathrm{\&theta;}}_m\right)\&CenterDot;\frac{{\mathrm{\&Sigma;}}_{{\mathrm{\&omega;}}_{t,k}\&Element;{\mathrm{\&Omega;}}_t}\mathrm{\&psi;}({\mathrm{\&omega;}}_{t,k},{\mathrm{\&alpha;}}_m)({\mathrm{\&eta;}}_{t,k}-k_{t,k})}{{\mathrm{\&Sigma;}}_{\stackrel{{\mathrm{\&theta;}}_i>0}{{\mathrm{\&alpha;}}_i\&Element;A}}{\mathrm{\&Sigma;}}_{{\mathrm{\&omega;}}_{t,k}\&Element;{\mathrm{\&Omega;}}_t}\mathrm{\&psi;}({\mathrm{\&omega;}}_{t,k},{\mathrm{\&alpha;}}_i)({\mathrm{\&eta;}}_{t,k}-k_{t,k})}---\left(5\right)$

Wherein ψ (ω _{t, k}, α _m) be discriminant function, 1≤m≤M; As monitoring task ω _{t, k}by equipment α _mψ (ω during monitoring _{t, k}, α _m)=1, otherwise ψ (ω _{t, k}, α _m)=0; Sign (.) is sign function, is defined as

$\mathrm{sign}\left(x\right)=\left\{\begin{array}{cc}1& x>0\\ 0& x=0\\ -1& x<0\end{array}\right.---\left(6\right)$

6. the device allocation method of the detecting devices net based on particle cluster algorithm according to claim 1, is characterized in that, presetting threshold value is in step 4 the max calculation time condition preset.