CN105868858A - Method for optimizing track of engraving machine - Google Patents

Abstract

The invention provides a method for optimizing the track of an engraving machine. The method includes the steps of converting a given processing track into a generalized traveling salesman model, solving the generalized traveling salesman model by using an improved ant colony algorithm, and obtaining the shortest path and the corresponding processing track sequence. After the track data are converted into a mathematics model that can be optimized through an algorithm, the ant colony algorithm is employed to optimize the track data, and the ordering optimization of given track is realized. Convergence speed is accelerated, and the precision of the optimal solution is ensured. The method is flexible and practical, and the processing efficiency of an engraving machine is substantially improved.

Description

A kind of track optimizing method for engraving machine

Technical field

The present invention relates to a kind of track optimizing method for engraving machine, including to the processing method of given trace with right The optimization method of the track after process.

Background technology

Engraving machine is exactly the equipment carrying out carving of installing machines to replace manual labor, and the processing that it is equivalent to a kind of brill, milling combines sets Standby.Engraving machine on the market is all directly to perform by the engraving track of CAM Software Create, and its working (machining) efficiency is deeply raw by CAM software The impact of the quality of the machining locus become, and these CAM software is typically by software company's exploitation of specialty, the production of engraving machine Business is difficult to be changed the quality of machining locus by CAM.Therefore, engraving machine system is when performing machining locus, first to machining locus It is optimized process just to seem and be highly desirable to.

Summary of the invention

It is an object of the invention to provide a kind of track optimizing method for engraving machine, including the process to given trace Method and the optimization method to the track after processing.Wherein the processing method to given trace solves and is converted into by track data The mathematical model problem of algorithm optimization can be carried out, the optimization method of the track after processing is solved the sequence to given trace Optimization problem.The present invention has the most practical feature, it is intended to substantially increase the working (machining) efficiency of above-mentioned engraving machine system.

The purpose of the present invention is realized by following technical proposals:

A kind of track optimizing method for engraving machine, including step:

(1) given machining locus is changed into broad sense travelling salesman's model；

(2) use and improve broad sense travelling salesman's model corresponding to ant colony optimization for solving, obtain shortest path and corresponding Machining locus order.

Further, described step (1) specifically includes:

(11) using the upper interpolated point of all given machining locus as " city " in broad sense travelling salesman's model, every track Preserve as " group of cities " and number；

(12) calculate the distance between each city, and ask for the neighborhood in each city according to the distance between each city；

(13) current city is initialized to the pheromone in these neighborhood cities.(size of pheromone concentration characterizes The distance in path) strengthen.

Further, described step (2) specifically includes:

(21) parameters improved in ant group algorithm is initialized, including: Formica fusca number m, tradition ratio p of ant, city Group number N_group, circulation total degree N, preliminary stage cycle-index be N ', initial time each routing information cellulose content τ_o, information inspire Factor-alpha, expectation heuristic factor β, pheromone intensity Q, volatility coefficient ρ；

(22) initializing ant colony, every Formica fusca k randomly chooses a group of cities, then randomly chooses one in this metropolitan county This city, as starting point, is joined Formica fusca path path by individual city_kIn, and by all cities in the group of cities at Formica fusca place City joins taboo list tabu_kIn；

(23) every Formica fusca is according to self affiliated ant kind and residing stage, calculates all general up to the transfer in city Rate, and with roulette rule select next up to city, and add it to Formica fusca path path_kIn, by Formica fusca place Group of cities in all cities join taboo list tabu_kIn；

(24) calculate the idle stroke in the path of every Formica fusca, select Formica fusca that in current iteration, idle stroke is the shortest as Excellent Formica fusca is also made comparisons with the optimum Formica fusca in all previous iteration, selects wherein optimum Formica fusca, updates the pheromone on each path, and Carry out next iteration；

(25) if iterations is N, then terminate algorithm, export optimal result, otherwise repeat step (23) and step (24)。

Further, the more new regulation updating the pheromone on each path described in step (24) is: by optimum Formica fusca Path definition is optimal path, if optimal path, then strengthens the pheromone on this path, and the letter on other paths Breath element then carries out volatilization to a certain extent according to volatilization factor.

Further, enhancing or the volatilization of described pheromone is determined by equation below:

${\mathrm{\τ}}_{ij}(t+n)=\{\begin{array}{cc}(1-\mathrm{\ρ})\·{\mathrm{\τ}}_{ij}\left(t\right)+{\mathrm{\Δ\τ}}_{ij}\left(t\right)& (i,j)\∈{\mathrm{Path}}_{\min }\\ {\mathrm{\τ}}_{ij}\left(t\right)& other\end{array},$

${\mathrm{\Δ\τ}}_{ij}\left(t\right)=\{\begin{array}{cc}Q/L& (i,j)\∈{\mathrm{Path}}_{\min }\\ 0& other\end{array},$

Wherein, τ_ij(t+n) represent be engraved in when t+n path (i, j) on the amount of pheromone, ρ ∈ [0,1) for volatilization system Number, 1-ρ is the residual coefficients of pheromone, Δ τ_ij(t) be this circulation in path (i, j) on pheromone increment, Path_minFor Optimum Formica fusca path, Q is pheromone intensity, and L is total length in optimum the walked path of Formica fusca in circulating by this, owing to using Ant- Cycle model and only row during optimal path being updated, so Δ τ_ijT the computing formula of () is as shown in Article 2 formula, Q can be with shadow Ring convergence of algorithm speed.

Further, ant kind described in step (23) include tradition ant, rebel ant with rebellion ant, described tradition ant be up to The hunting zone in city is that the neighborhood of current city is its Formica fusca up to city；Described rebel ant is the search up to city Scope is all cities in addition to city in current Formica fusca taboo list；If the neighborhood of current city does not has optional city City, then tradition ant is converted into rebellion ant, and it becomes in addition to city in current Formica fusca taboo list up to the hunting zone in city All cities；The described residing stage includes preliminary stage and later stage, is preliminary stage when iterations is less than N ', greatly In being later stage equal to N '.

Further, described step (23) calculate all specifically include step up to the transition probability in city:

If Formica fusca be tradition ant, then the neighborhood of current city be its up to city, its transition probability computing formula is:

The neighborhood of city i is U (i), τ '_ijInformation table concentration between (t) strengthened t city i and j, η_ijFor road (i, j) heuristic information of self, its value generally takes η in footpath_ij=1/d_ij, wherein d_ijIt it is the distance between i and j of city；

If not having optional city in the neighborhood of current city, then tradition ant is converted into rebellion ant, and it is up to the model in city Enclosing all cities become in addition to city in current Formica fusca taboo list, the computing formula of its transition probability is:

If current Formica fusca is for rebelling ant, then it is in addition to city in current Formica fusca taboo list up to the hunting zone in city All cities, the computing formula of its transition probability is:

$p_{ij}^k\left(t\right)=\{\begin{array}{cc}\frac{{\[{\mathrm{\τ}}_{ij}\left(t\right)\]}^{\mathrm{\α}}{\[{\mathrm{\η}}_{ij}\left(t\right)\]}^{\mathrm{\β}}}{\underset{s\∈{\mathrm{allowed}}_k}{\Σ}{\[{\mathrm{\τ}}_{is}\left(t\right)\]}^{\mathrm{\α}}{\[{\mathrm{\η}}_{is}\left(t\right)\]}^{\mathrm{\β}}}& j\∈{\mathrm{allowed}}_k\\ 0& other\end{array},$

Wherein, allowed_kPermission for kth Formica fusca shifts city, i.e. except tabu_kOutside all cities, τ_ij(t) The information table concentration strengthened for not the carrying out between i and j of t city；

When all Formica fuscas are all gone through all over complete all groups of cities, then algorithm completes an iteration.

Further, described step (23) also includes:

If the stage residing for Formica fusca is preliminary stage, according to formula:

$p_{ij}^{k^{*}}\left(t\right)=1-\frac{\sqrt{\sqrt{2}a+a^2-2\sqrt{2}{\mathrm{ap}}_{ij}^k\left(t\right)+0.5}-\frac{\sqrt{2}}{2}}{a}-p_{ij}^k\left(t\right)$

Transition probability is amplified, whereinFor amplify before with amplify after transition probability,Be a coefficient with the change of algorithm cycle-index, n-th (n ∈ [0, N ')) secondary circulation time a value be a_n, Then have:

$a_n=-\frac{\sqrt{2}}{2}+\frac{\sqrt{2}n}{2N^{\′}},n\∈\[n,N^{\′}\],$

I.e. algorithm iteration number of times is less than N ', is not amplified transition probability if greater than equal to N ', to accelerate algorithm Convergence rate.

Compared to existing technology, after the present invention by being converted into the mathematical model that can carry out algorithm optimization by track data, Use the ant group algorithm improved that track data is optimized, solve the sorting consistence problem to given trace, add rapid convergence Also ensure that the precision of optimal solution while speed, there is the most practical feature, substantially increase the working (machining) efficiency of engraving machine.

Accompanying drawing explanation

Fig. 1 is that in the present invention, engraving machine system adds the schematic diagram producing idle stroke man-hour.

Fig. 2 is the schematic diagram of the engraving machine performing machining locus in the present invention.

Detailed description of the invention

Below in conjunction with drawings and Examples, the present invention is described in further detail, but embodiments of the present invention are not It is limited to this.

The engraving machine machining locus generated in CAM software is input in host computer, host computer to machining locus Processing mode is as follows:

A kind of track optimizing method for engraving machine, including step:

(1) using the upper interpolated point of all machining locus as " city " in broad sense travelling salesman's model, row are deposited in In table (list) data structure, will be numbered according to the sequence number in its place list；

(2) calculate the distance between each city, and deposit in matrix mat_dist, the element mat_dist in matrix (i, j) represents the Euclidean distance between city i and city j, and wherein i, j are the numbering in the city in (1), and according to each city it Between distance ask for the neighborhood in each city, several cities that i.e. distance current city distance is the shortest, and by the volume in these cities Number deposit in the list_neighborhood list of current city；

(3) to the pheromone concentration of current city to these neighborhood cities, (size of pheromone concentration characterizes the remote of path Closely) strengthen；

(4) parameters improved in ant group algorithm is initialized, including: Formica fusca number m, ratio p of tradition ant, group of cities Number N_group, circulating total degree N, preliminary stage cycle-index is N ', initial time each routing information cellulose content τ_o, information inspire because of Sub-α, it is desirable to heuristic factor β, pheromone intensity Q, volatility coefficient ρ；

(5) initializing ant colony, every Formica fusca k randomly chooses a group of cities, then randomly chooses one in this metropolitan county This city, as starting point, is joined Formica fusca path path by individual city_kIn, and by all cities in the group of cities at Formica fusca place City joins taboo list tabu_kIn；

(6) every Formica fusca is according to self affiliated ant kind and residing stage, calculates all transition probabilities up to city, And with roulette rule select next up to city, and add it to Formica fusca path path_kIn, and by Formica fusca place All cities in group of cities join taboo list tabu_kIn, described ant kind includes tradition ant, rebels ant and rebellion ant, described Tradition ant is to be its Formica fusca up to city up to the neighborhood that hunting zone is current city in city；Described rebel ant is can The hunting zone reaching city is all cities in addition to city in current Formica fusca taboo list；If the neighborhood of current city does not has Have optional city, then tradition ant is converted into rebellion ant, and it becomes except in current Formica fusca taboo list up to the hunting zone in city All cities outside city；The described residing stage includes preliminary stage and later stage, when iterations is less than N ' is Preliminary stage, is later stage more than or equal to N ', calculates and all specifically includes step up to the transition probability in city:

(6.1) if Formica fusca be tradition ant, then the neighborhood of current city be its up to city, its transition probability calculate public affairs Formula is:

The neighborhood of city i is U (i), τ '_ijInformation table concentration between (t) strengthened t city i and j；η_ijFor road (i, j) heuristic information of self, its value generally takes η in footpath_ij=1/d_ij, wherein d_ijCity i, the distance between j；

(6.2) if not having optional city in the neighborhood of current city, then tradition ant is converted into rebellion ant, and it is up to city Scope become all cities in addition to city in current Formica fusca taboo list, the computing formula of its transition probability is:

(6.3) if current Formica fusca is for rebelling ant, then it is except city in current Formica fusca taboo list up to the hunting zone in city All cities outside city, the computing formula of its transition probability is:

$p_{ij}^k\left(t\right)=\{\begin{array}{cc}\frac{{\[{\mathrm{\τ}}_{ij}\left(t\right)\]}^{\mathrm{\α}}{\[{\mathrm{\η}}_{ij}\left(t\right)\]}^{\mathrm{\β}}}{\underset{s\∈{\mathrm{allowed}}_k}{\Σ}{\[{\mathrm{\τ}}_{is}\left(t\right)\]}^{\mathrm{\α}}{\[{\mathrm{\η}}_{is}\left(t\right)\]}^{\mathrm{\β}}}& j\∈{\mathrm{allowed}}_k\\ 0& other\end{array},$

Wherein allowed_kPermission for kth Formica fusca shifts city, i.e. except tabu_kOutside all cities, τ_ij(t) be The information table concentration that not carrying out between i and j of t city is strengthened；

(6.4) if the stage residing for Formica fusca is that (i.e. algorithm iteration number of times is less than N ', if greater than equal to N ' then to preliminary stage Transition probability is not amplified), for accelerating convergence of algorithm speed, can be according to formula:

$p_{ij}^{k^{*}}\left(t\right)=1-\frac{\sqrt{\sqrt{2}a+a^2-2\sqrt{2}{\mathrm{ap}}_{ij}^k\left(t\right)+0.5}-\frac{\sqrt{2}}{2}}{a}-p_{ij}^k\left(t\right)$

Transition probability is amplified, whereinFor amplify before with amplify after transition probability,Be a coefficient with the change of algorithm cycle-index, n-th (n ∈ [0, N ')) secondary circulation time a value be a_n, Then have:

$a_n=-\frac{\sqrt{2}}{2}+\frac{\sqrt{2}n}{2N^{\′}},n\∈\[n,N^{\′}\],$

(6.5) all go through all over complete all groups of cities when all Formica fuscas, then algorithm completes an iteration；

(7) calculate the idle stroke in the path of every Formica fusca, select Formica fusca that in current iteration, idle stroke is the shortest as Excellent Formica fusca is also made comparisons with the optimum Formica fusca in all previous iteration, selects wherein optimum Formica fusca, and updates the pheromone on each path, If optimal path (path of i.e. optimum Formica fusca), then the pheromone on this path is strengthened, and on other paths Pheromone then carries out volatilization to a certain extent, τ according to volatilization factor_ij(t+n) represent be engraved in when t+n path (i, j) on The amount of pheromone, its more new regulation is as follows:

${\mathrm{\τ}}_{ij}(t+n)=\{\begin{array}{cc}(1-\mathrm{\ρ})\·{\mathrm{\τ}}_{ij}\left(t\right)+{\mathrm{\Δ\τ}}_{ij}\left(t\right)& (i,j)\∈{\mathrm{Path}}_{\min }\\ {\mathrm{\τ}}_{ij}\left(t\right)& other\end{array},$

${\mathrm{\Δ\τ}}_{ij}\left(t\right)=\{\begin{array}{cc}Q/L& (i,j)\∈{\mathrm{Path}}_{\min }\\ 0& other\end{array},$

Wherein, and ρ ∈ [0,1) it is volatility coefficient, 1-ρ is the residual coefficients of pheromone, Δ τ_ijT () is path in this circulation (i, j) on pheromone increment, Path_minFor optimum ant path.Owing to using Ant-Cycle model and during only to optimal path Row updates, so Δ τ_ijT the computing formula of () is as shown in Article 2 formula, Q is pheromone intensity, and it can affect the receipts of algorithm Holding back speed, L is total length in optimum the walked path of Formica fusca in circulating by this；

(8) if iterations is equal to N, then terminate algorithm, export optimal result, otherwise repeat step (6) to (7).

After upper computer software has processed track, carried out the machining locus of optimal sequencing, these tracks are input to In the controller of engraving machine as shown in Figure 2, controller is according to machining locus, the driver of output control pulse to servomotor In, driver drives servomotor rotates, the linear module in three coordinate moving platforms that motor drag is coupled, drives solid It is connected in the feed motion of the electro spindle of three coordinate moving platforms, so that workpiece is carried out Carving Machining.In accompanying drawing 2,1 is engraving machine Fuselage, 2 is the electro spindle for cutting workpiece, and its end is provided with cutter, and 3 is three coordinate moving platforms, and 4 is AC servo electricity Machine.

As it is shown in figure 1, in the present embodiment, when engraving machine system performs i-th machining locus, cutter can dropping distance d_i To current machining locus starting point, when cutter goes to the end point of track, cutter can automatic lifting distance r_i, then move away from From m_iAbove the starting point of next track, prepare to process next track.Displacement (the d of cutter during above three_i, r_i, m_i) it is idle stroke L produced in the course of processing_invalid, for all i, d_i, r_iAll keep constant, then perform all tracks The idle stroke that (n bar altogether) produces isAdjust the processing sequence of given trace, thus it is possible to vary m_i, So employing optimized algorithm, can be in the hope ofCan be in the hope of the shortest idle stroke min L in the course of processing_invalid。 Point on every track is considered as a city, and every track is considered as a group of cities, then can will solve minL_invalidProblem turns Turn to solve broad sense traveling salesman problem.

Given track is divided into two classes by the present embodiment, and a class is the polygon closed, the another kind of multi-section-line for opening, By the group of cities in track correspondence broad sense travelling salesman's model, and the city of the corresponding each group of cities of each interpolated point on track, Because there being two class tracks, so there being two class groups of cities, in the open group of cities corresponding to multi-section-line track, it is necessary to and it is only capable of choosing Select two cities of first and last therein, and in the group of cities corresponding to closed polygon track, (only can arbitrarily choose a city One city of energy), go through and be the idle stroke that engraving machine adds man-hour and produces, shortest path institute all over the loop that all groups of cities are formed Corresponding machining locus order is the machining locus order of optimum.

When using the ant group algorithm improved, if track is closed polygon, then every Formica fusca is corresponding through this track During group of cities, optionally one of them city, if track is open multi-section-line, then every Formica fusca is corresponding through this track City time, it is necessary to and only may select 2 corresponding cities of track first and last, calculate every Formica fusca city total of process Distance, for judging the quality of Formica fusca, total distance is the shortest, then Formica fusca is the most excellent.

The present embodiment Formica fusca is divided into tradition ant, rebels ant and rebellion ant, three kinds of Formica fuscas next city in searching route During city, it is respectively adopted different hunting zones, has Local Search, also have global search, because part Formica fusca only can be in current city Neighborhood in the next city of search, so can greatly reduce and search for the time waste brought in all cities, accelerate algorithm The precision of algorithm optimal solution is also ensure that while convergence rate.

It addition, whole optimization process is divided into two stages, it is preliminary stage when iterations is less than N ', after being more than or equal to N ' Stage phase, at preliminary stage, in order to add rapid convergence, by formula

Transition probability is amplified, during the late stages of developmet, sinks into local optimum, then to prevent the crossing rapid convergence of algorithm Using normal transition probability, this measure also ensure that the accuracy of algorithm optimal solution while ensureing algorithm Fast Convergent.

The above embodiment of the present invention is only for clearly demonstrating example of the present invention, and is not to the present invention The restriction of embodiment.For those of ordinary skill in the field, can also make on the basis of the above description The change of other multi-form or variation.Here without also cannot all of embodiment be given exhaustive.All the present invention's Any amendment, equivalent and the improvement etc. made within spirit and principle, should be included in the protection of the claims in the present invention Within the scope of.

Claims (8)

1. the track optimizing method for engraving machine, it is characterised in that include step:

(1) given machining locus is changed into broad sense travelling salesman's model；

(2) use broad sense travelling salesman's model that improvement ant colony optimization for solving is corresponding, obtain shortest path and corresponding processing thereof Track order.

Track optimizing method for engraving machine the most according to claim 1, it is characterised in that: described step (1) is concrete Including:

(11) using the upper interpolated point of all given machining locus as " city " in broad sense travelling salesman's model, every track conduct " group of cities " preserves and numbers；

(12) calculate the distance between each city, and ask for the neighborhood in each city according to the distance between each city；

(13) current city is strengthened to the pheromone in these neighborhood cities.

Track optimizing method for engraving machine the most according to claim 1, it is characterised in that: described step (2) is concrete Including:

(21) parameters improved in ant group algorithm is initialized, including: Formica fusca number m, tradition ratio p of ant, group of cities number N_group, circulation total degree N, preliminary stage cycle-index be N ', initial time each routing information cellulose content τ_o, information heuristic factor α, expectation heuristic factor β, pheromone intensity Q, volatility coefficient ρ；

(22) initializing ant colony, every Formica fusca k randomly chooses a group of cities, then randomly chooses a city in this metropolitan county This city, as starting point, is joined Formica fusca path path by city_kIn, and all cities in the group of cities at Formica fusca place are added Enter to taboo list tabu_kIn；

(23) every Formica fusca is according to self affiliated ant kind and residing stage, calculates all transition probabilities up to city, and With roulette rule select next up to city, and add it to Formica fusca path path_kIn, and by the city at Formica fusca place All cities in city group join taboo list tabu_kIn；

(24) calculate the idle stroke in the path of every Formica fusca, select the Formica fusca that in current iteration, idle stroke is the shortest as optimum ant Ant is also made comparisons with the optimum Formica fusca in all previous iteration, selects wherein optimum Formica fusca, updates the pheromone on each path, and carry out Next iteration；

(25) if iterations is N, then terminate algorithm, export optimal result, otherwise repeat step (23) and step (24).

Track optimizing method for engraving machine the most according to claim 3, it is characterised in that: described in step (24) more The more new regulation of the pheromone on new each path is: be optimal path by the path definition of optimum Formica fusca, if optimal path, then Strengthening the pheromone on this path, the pheromone on other paths is then carried out to a certain extent according to volatilization factor Volatilization.

Track optimizing method for engraving machine the most according to claim 4, it is characterised in that: the enhancing of described pheromone Or volatilization is determined by equation below:

${\mathrm{\τ}}_{ij}(t+n)=\left\{\begin{array}{cc}(1-\mathrm{\ρ})\·{\mathrm{\τ}}_{ij}\left(t\right)+{\mathrm{\Δ\τ}}_{ij}\left(t\right)& (i,j)\∈{\mathrm{Path}}_{\min }\\ {\mathrm{\τ}}_{ij}\left(t\right)& other\end{array}\right.,$

${\mathrm{\Δ\τ}}_{ij}\left(t\right)=\left\{\begin{array}{cc}Q/L& (i,j)\∈{\mathrm{Path}}_{min}\\ 0& other\end{array}\right.,$

Wherein, τ_ij(t+n) represent be engraved in when t+n path (i, j) on the amount of pheromone, ρ ∈ [0,1) be volatility coefficient, 1-ρ For the residual coefficients of pheromone, Δ τ_ij(t) be this circulation in path (i, j) on pheromone increment, Path_minFor optimum ant Ant path, Q is pheromone intensity, and L is total length in optimum the walked path of Formica fusca in circulating by this.

Track optimizing method for engraving machine the most according to claim 3, it is characterised in that: ant described in step (23) Planting and include tradition ant, rebel ant and rebellion ant, described traditional ant is equal up to the neighborhood that the hunting zone in city is current city For it up to the Formica fusca in city；Described rebel ant is All cities；If not having optional city in the neighborhood of current city, then tradition ant is converted into rebellion ant, and it is up to city Hunting zone become all cities in addition to city in current Formica fusca taboo list；The described residing stage includes preliminary stage And later stage, it is preliminary stage when iterations is less than N ', is later stage more than or equal to N '.

Track optimizing method for engraving machine the most according to claim 6, it is characterised in that: described step (23) is fallen into a trap Calculate and all specifically include step up to the transition probability in city:

If Formica fusca be tradition ant, then the neighborhood of current city be its up to city, its transition probability computing formula is:

The neighborhood of city i is U (i), τ '_ijInformation table concentration between (t) strengthened t city i and j, η_ijFor path (i, J) heuristic information of self, its value generally takes η_ij=1/d_ij, wherein d_ijIt it is the distance between i and j of city；

If not having optional city in the neighborhood of current city, then tradition ant is converted into rebellion ant, and it becomes up to the scope in city For all cities in addition to city in current Formica fusca taboo list, the computing formula of its transition probability is:

If current Formica fusca is for rebelling ant, then it is the institute in addition to city in current Formica fusca taboo list up to the hunting zone in city Having city, the computing formula of its transition probability is:

$p_{ij}^k\left(t\right)=\left\{\begin{array}{cc}\frac{{\[{\mathrm{\τ}}_{ij}\left(t\right)\]}^{\mathrm{\α}}{\[{\mathrm{\η}}_{ij}\left(t\right)\]}^{\mathrm{\β}}}{\underset{s\∈{\mathrm{allowed}}_k}{\Σ}{\[{\mathrm{\τ}}_{is}\left(t\right)\]}^{\mathrm{\α}}{\[{\mathrm{\η}}_{is}\left(t\right)\]}^{\mathrm{\β}}}& j\∈{\mathrm{allowed}}_k\\ 0& other\end{array}\right.,$

Wherein, allowed_kPermission for kth Formica fusca shifts city, i.e. except tabu_kOutside all cities, τ_ijWhen () is t t Carve the information table concentration that not the carrying out between i and j of city is strengthened；

When all Formica fuscas are all gone through all over complete all groups of cities, then algorithm completes an iteration.

Track optimizing method for engraving machine the most according to claim 3, it is characterised in that: described step (23) is also wrapped Include:

If the stage residing for Formica fusca is preliminary stage, according to formula:

$p_{ij}^{k^{*}}\left(t\right)=1-\frac{\sqrt{\sqrt{2}a+a^2-2\sqrt{2}{\mathrm{ap}}_{ij}^k\left(t\right)+0.5}-\frac{\sqrt{2}}{2}}{a}-p_{ij}^k\left(t\right)$

Transition probability is amplified, whereinFor amplify before with amplify after transition probability,Be a coefficient with the change of algorithm cycle-index, n-th (n ∈ [0, N ')) secondary circulation time a value be a_n, Then have:

$\begin{array}{cc}{a}_n=-\frac{\sqrt{2}}{2}+\frac{\sqrt{2}n}{2N^{\′}}& n\∈\[n,N^{\′}\]\end{array}.$