CN106200650A - Based on method for planning path for mobile robot and the system of improving ant group algorithm - Google Patents

Abstract

The present invention relates to a kind of method for planning path for mobile robot based on improvement ant group algorithm and system, this method for planning path for mobile robot comprises the steps: step S1, environmental modeling；Step S2, initial information element distributes；And step S3, optimum path search, and output optimal path；Tradition ant group algorithm in the past is improved by the present invention for initialization information element distribution aspect so that the optimizing of Formica fusca can be made guiding by Formica fusca at the very start, and preconvergence speed is substantially accelerated；Reasonable selection to initial parameter simultaneously, the such as selection of pheromone volatilization factor, result is made to be unlikely to be absorbed in locally optimal solution or be difficult to form optimal solution, and Pheromone update mode is made rational modification, can be prevented effectively from and be absorbed in local optimum and improve machine task efficiency and functional reliability.

Description

Based on method for planning path for mobile robot and the system of improving ant group algorithm

Technical field

The present invention relates to mobile robot intelligent algorithm technical field, be specifically related to a kind of based on the shifting improving ant group algorithm Mobile robot paths planning method.This method effectively prevent locally optimal solution and overcomes the problem that convergence rate is slow.

Background technology

Mobile robot path planning refers to the planning of the movement locus of robot, i.e. specifies mobile robot at a tool Having the environment of barrier, give concrete beginning and end simultaneously, under given appreciation condition, mobile robot is according to giving Task avoiding obstacles, search for from the optimal path of origin-to-destination, i.e. require to spend the shortest time, run the shortest road Footpath or consume minimum energy.At present, many intelligent algorithms are applied in mobile robot path planning, including neutral net Method, ant group algorithm, Artificial Potential Field Method, particle algorithm, genetic algorithm, fuzzy reasoning method etc..

Ant group algorithm is the behavior looked for food according to ant colony, and the heuristic intelligent search finding optimal path in appointment figure is calculated Method.Ant group algorithm demonstrates stronger advantage in combinatorial optimization problem, is a reinforcement learning system, has distributed Estimated performance, and there is stronger robustness, it is easy to merge with other optimized algorithms.From proposing up till now, many both at home and abroad This algorithm has been done numerous studies by scholar, is applied to numerous areas, and achieves great successes, typical case application such as travelling Business, assignment problem, scheduling problem, Filled function, network route etc..

Summary of the invention

It is an object of the invention to provide a kind of method for planning path for mobile robot and system, to solve existing ant group algorithm It is easily trapped into local optimum, the speed of service slow, in the slow-footed problem of information processing preconvergence.

In order to solve above-mentioned technical problem, the invention provides a kind of method for planning path for mobile robot, including as follows Step:

Step S1, environmental modeling；

Step S2, initial information element distributes；And

Step S3, optimum path search, and output optimal path.

Further, environmental modeling in described step S1, i.e.

Local environment information according to mobile robot uses grid environmental modeling.

Further, the described local environment information according to mobile robot uses the method for grid environmental modeling to include:

The mobile robot of utilization is from belt sensor group collecting work environmental information, and carries out Map building；Wherein, will be mobile Robot and each city, as particle, will be moved robot and barrier and model according to two-dimensional coordinate system.

Further, in described step S2 initial information element distribution, i.e. initial information element according to Origin And Destination line near Regional concentration is relatively big, and the principle that the pheromone concentration of two diagonal zones that beginning and end line is relative is less is allocated.

Further, optimum path search in described step S3, and output optimal path；I.e.

Optimum path search is carried out, after completing once to circulate, to the pheromone on each city access path according to ant group algorithm Carry out real-time update, after reaching maximum iteration time, export optimal path.

Further, described ant group algorithm carries out the method for optimum path search and includes:

Step Sa, arranges initial parameter, including ant colony scale m, pheromone significance level factor-alpha, the important journey of heuristic function Degree factor-beta, pheromone volatilization factor ρ, pheromone release total amount Q, and set maximum iteration time, iterations initial value；

Step Sb, is randomly placed on Formica fusca starting point, and randomly chooses next node to be visited, and wherein Formica fusca is from node I transfers to the computing formula of the probability of node j, i.e. transition probability:

$P_{ij}^k=\left\{\begin{array}{c}\frac{{\[{\mathrm{\τ}}_{ij}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{ij}\left(t\right)\]}^{\mathrm{\β}}}{\underset{s\∈{\mathrm{allow}}_k}{\Σ}{\[{\mathrm{\τ}}_{is}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{is}\left(t\right)\]}^{\mathrm{\β}}},s\∈{\mathrm{allow}}_k\\ 0,s\∉{\mathrm{allow}}_k\end{array}\right.---\left(1\right);$

In formula (1),

allow_k(k=1,2 ..., m) it is the set of Formica fusca k node to be visited, and when starting, allow_kIn have (n-1) Individual element, i.e. includes other all nodes in addition to Formica fusca k starting point, propelling over time, allow_kIn element by The fewest, until being empty, i.e. have access to impact point complete；η_ijT () is heuristic function, represent that Formica fusca transfers to node j from node i Expected degree, computing formula is as follows:

${\mathrm{\η}}_{ij}\left(t\right)=\frac{1}{d_{ij}}---\left(2\right);$

In formula (2), d_ijRepresenting the distance between node i and node j, computing formula is as follows:

$d_{ij}=\sqrt{{(x_i-x_j)}^2+{(y_i-y_j)}^2}---\left(3\right);$

Step Sc, after having calculated internodal transition probability, uses the next joint to be visited of roulette method choice Point；

Step Sd, path walked to Formica fusca carries out local message element amount and updates, dense to weaken the pheromone in path of having passed by Degree, it is as follows that local message element measures more new formula:

τ_ij(t+1)=(1-ε) τ_ij(t)+ετ₀(4)；

In formula (4): ε is local message element volatility coefficient, τ₀=λ/d_ij, λ is constant；

Step Se, when all Formica fuscas all complete one take turns iteration after, pheromone concentration is carried out the overall situation renewal, more new formula is such as Under:

τ_ij(t+1)=(1-ρ) τ_ij(t)+ω₁ρΔ₁τ_ij+ω₂ρΔ₂τ_ij(5)；

${\mathrm{\Δ}}_1{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{worst}-L_{best})/L_{best},L_i\∈L_{best}\\ Q/L_i,L_i\∉L_{best}\end{array}\right.---\left(6\right);$

${\mathrm{\Δ}}_2{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{best}-L_{worst})/L_{worst},L_i\∈L_{worst}\\ Q/L_i,L_i\∉L_{worst}\end{array}\right.---\left(7\right);$

In formula (5), (6), (7),

ρ is global information element volatility coefficient；ω₁、ω₂For weight coefficient, and ω₁+ω₂=1；L_bestOptimum for current iteration Path；L_worstFor current iteration worst path length；L_iIt is the i-th paths length；Q is ant colony pheromone total amount；

Step Sf, it may be judged whether reach maximum iteration time；I.e.

If being not reaching to maximum iteration time, go to step Sc；

Otherwise, terminate iteration, export optimal solution, be the Formica fusca route map through optimal path.

Another aspect, present invention also offers a kind of mobile robot path planning system.

Described mobile robot path planning system includes:

Environmental modeling unit, uses grid environmental modeling according to the local environment information of mobile robot；

Path planning unit, initial information element distributes, and by optimum path search to export optimal path.

Further, described environmental modeling unit includes: with the mobile robot of environment information acquisition sensor；

Described mobile robot collecting work environmental information, and carry out Map building；Wherein, will mobile robot and each City, as particle, will be moved robot and barrier and model according to two-dimensional coordinate system.

Further, described path planning unit is distributed by initial information element and optimum path search is to export optimal path；Its Middle initial information element distributes, i.e. initial information element is relatively big according to Origin And Destination line near zone concentration, and beginning and end is even The principle that the pheromone concentration of two diagonal zones that line is relative is less is allocated；And optimum path search is with the optimum road of output Footpath, i.e. carries out optimum path search according to ant group algorithm, after completing once to circulate, carries out the pheromone on each city access path Real-time update, after reaching maximum iteration time, exports optimal path.

Further, described ant group algorithm carries out the method for optimum path search and includes:

Step Sa, arranges initial parameter, including ant colony scale m, pheromone significance level factor-alpha, the important journey of heuristic function Degree factor-beta, pheromone volatilization factor ρ, pheromone release total amount Q, and set maximum iteration time, iterations initial value；

Step Sb, is randomly placed on Formica fusca starting point, and randomly chooses next node to be visited, and wherein Formica fusca is from node I transfers to the computing formula of the probability of node j, i.e. transition probability:

$P_{ij}^k=\left\{\begin{array}{c}\frac{{\[{\mathrm{\τ}}_{ij}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{ij}\left(t\right)\]}^{\mathrm{\β}}}{\underset{s\∈{\mathrm{allow}}_k}{\Σ}{\[{\mathrm{\τ}}_{is}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{is}\left(t\right)\]}^{\mathrm{\β}}},s\∈{\mathrm{allow}}_k\\ 0,s\∉{\mathrm{allow}}_k\end{array}\right.---\left(1\right);$

In formula (1),

allow_k(k=1,2 ..., m) it is the set of Formica fusca k node to be visited, and when starting, allow_kIn have (n-1) Individual element, i.e. includes other all nodes in addition to Formica fusca k starting point, propelling over time, allow_kIn element by The fewest, until being empty, i.e. have access to impact point complete；η_ijT () is heuristic function, represent that Formica fusca transfers to node j from node i Expected degree, computing formula is as follows:

${\mathrm{\η}}_{ij}\left(t\right)=\frac{1}{d_{ij}}---\left(2\right);$

In formula (2), d_ijRepresenting the distance between node i and node j, computing formula is as follows:

$d_{ij}=\sqrt{{(x_i-x_j)}^2+{(y_i-y_j)}^2}---\left(3\right);$

Step Sc, after having calculated internodal transition probability, uses the next joint to be visited of roulette method choice Point；

Step Sd, path walked to Formica fusca carries out local message element amount and updates, dense to weaken the pheromone in path of having passed by Degree, it is as follows that local message element measures more new formula:

τ_ij(t+1)=(1-ε) τ_ij(t)+ετ₀(4)；

In formula (4): ε is local message element volatility coefficient, τ₀=λ/d_ij, λ is constant；

Step Se, when all Formica fuscas all complete one take turns iteration after, pheromone concentration is carried out the overall situation renewal, more new formula is such as Under:

τ_ij(t+1)=(1-ρ) τ_ij(t)+ω₁ρΔ₁τ_ij+ω₂ρΔ₂τ_ij(5)；

${\mathrm{\Δ}}_1{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{worst}-L_{best})/L_{best},L_i\∈L_{best}\\ Q/L_i,L_i\∉L_{best}\end{array}\right.---\left(6\right);$

${\mathrm{\Δ}}_2{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{best}-L_{worst})/L_{worst},L_i\∈L_{worst}\\ Q/L_i,L_i\∉L_{worst}\end{array}\right.---\left(7\right);$

In formula (5), (6), (7),

ρ is global information element volatility coefficient；ω₁、ω₂For weight coefficient, and ω₁+ω₂=1；L_bestOptimum for current iteration Path；L_worstFor current iteration worst path length；L_iIt is the i-th paths length；Q is ant colony pheromone total amount；

Step Sf, it may be judged whether reach maximum iteration time；I.e.

If being not reaching to maximum iteration time, go to step Sc；

Otherwise, terminate iteration, export optimal solution, be the Formica fusca route map through optimal path.

The invention has the beneficial effects as follows, tradition ant group algorithm in the past is done by the present invention for initialization information element distribution aspect Improve so that the optimizing of Formica fusca can be made guiding by Formica fusca at the very start, preconvergence speed is substantially accelerated；Simultaneously to initially The reasonable selection of parameter, the such as selection of pheromone volatilization factor ρ so that result is unlikely to be absorbed in locally optimal solution or be difficult to Form optimal solution, and Pheromone update mode is made rational modification, can be prevented effectively from and be absorbed in local optimum and improve machine Task efficiency and functional reliability.

Accompanying drawing explanation

The present invention is further described with embodiment below in conjunction with the accompanying drawings.

Fig. 1 present invention improves ant group algorithm flow chart；

Fig. 2 tradition ant group algorithm route map；

Fig. 3 tradition ant group algorithm respectively contrasts with average distance for beeline；

Ant group algorithm path optimizing figure after Fig. 4 improvement；

After Fig. 5 improves, ant group algorithm respectively contrasts with average distance for beeline.

Detailed description of the invention

In conjunction with the accompanying drawings, the present invention is further detailed explanation.These accompanying drawings are the schematic diagram of simplification, only with The basic structure of the illustration explanation present invention, therefore it only shows the composition relevant with the present invention.

The invention discloses a kind of method for planning path for mobile robot based on improvement ant group algorithm and system, its planning Step is: be modeled mobile work robot environment；According to nodal distance information, pheromone concentration is initialized, give Formica fusca provides and guides；This ant group algorithm is utilized to allow mobile robot carry out route searching；According to optimum results output mobile machine The motion path of people.The present invention is prevented effectively from locally optimal solution by improving Pheromone update mode, and overcomes tradition ant The problem that group's algorithm the convergence speed is slow.

By the following examples 1 and embodiment 2 this method for planning path for mobile robot and system are carried out launch explanation.

Embodiment 1

As it is shown in figure 1, the present embodiment 1 provides a kind of method for planning path for mobile robot, including:

Step S1, environmental modeling；

Step S2, initial information element distributes；And

Step S3, optimum path search, and output optimal path.

As environmental modeling one preferred embodiment, environmental modeling in described step S1, i.e. according to mobile machine The local environment information of people uses grid environmental modeling.Concrete, the described local environment information according to mobile robot uses The method of grid environmental modeling includes: utilize mobile robot from belt sensor group (such as but not limited to photographic head, sonar ring, Infrared sensor) collecting work environmental information, and carry out Map building；Wherein, robot and each city will be moved as matter Point, will move robot and barrier and model according to two-dimensional coordinate system, and grid environment will be carried out coordinate process.

Initial information element distribution in described step S2, i.e.

Initial information element is relatively big according to Origin And Destination line near zone concentration, two that beginning and end line is relative The principle that the pheromone concentration of diagonal zones is less is allocated.

As optimum path search in described step S3, and the detailed description of the invention of output optimal path.

Optimum path search is carried out, after completing once to circulate, to the pheromone on each city access path according to ant group algorithm Carry out real-time update, after reaching maximum iteration time, export optimal path.

As it is shown in figure 1, the ant group algorithm flow chart added up to described in this method for planning path for mobile robot is specific as follows:

Step Sa, arranges initial parameter, including ant colony scale m, pheromone significance level factor-alpha, the important journey of heuristic function Degree factor-beta, pheromone volatilization factor ρ, pheromone release total amount Q, and set maximum iteration time, iterations initial value, its Whether middle initial value can be set to 1, maximum iteration time or be set as required, and reach home relevant with Formica fusca.

Step Sb, is randomly placed on starting point by Formica fusca, randomly chooses next node to be visited according to certain probability, its The computing formula that middle Formica fusca transfers to the probability of node j, i.e. transition probability from node i is:

$P_{ij}^k=\left\{\begin{array}{c}\frac{{\[{\mathrm{\τ}}_{ij}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{ij}\left(t\right)\]}^{\mathrm{\β}}}{\underset{s\∈{\mathrm{allow}}_k}{\Σ}{\[{\mathrm{\τ}}_{is}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{is}\left(t\right)\]}^{\mathrm{\β}}},s\∈{\mathrm{allow}}_k\\ 0,s\∉{\mathrm{allow}}_k\end{array}\right.---\left(1\right);$

In formula (1), allow_k(k=1,2 ..., m) it is the set of Formica fusca k node to be visited, during beginning, allow_kIn There is (n-1) individual element, i.e. include other all nodes in addition to Formica fusca k starting point, propelling over time, allow_kIn Element gradually decreases, until being empty, i.e. has access to impact point complete；α is the pheromone significance level factor, and its value is the biggest, represents The concentration of pheromone role in transfer is the biggest；β is the heuristic function significance level factor, and its value is the biggest, represents and inspires letter Number effect in transfer is the biggest, i.e. Formica fusca can be transferred to apart from short node with bigger probability；η_ijT () is heuristic function, Representing that Formica fusca transfers to the expected degree of node j from node i, computing formula is as follows:

${\mathrm{\η}}_{ij}\left(t\right)=\frac{1}{d_{ij}}---\left(2\right);$

In formula (2), d_ijRepresenting the distance between node i and node j, computing formula is as follows:

$d_{ij}=\sqrt{{(x_i-x_j)}^2+{(y_i-y_j)}^2}---\left(3\right);$

Step Sc, after having calculated internodal transition probability, uses the next joint to be visited of roulette method choice Point；

Step Sd, path walked to Formica fusca carries out local message element amount and updates, dense to weaken the pheromone in path of having passed by Degree, it is as follows that local message element measures more new formula:

τ_ij(t+1)=(1-ε) τ_ij(t)+ετ₀(4)；

In formula (4): ε is local message element volatility coefficient, τ₀=λ/d_ij, λ is constant；

Step Se, when all Formica fuscas all complete one take turns iteration after, pheromone concentration is carried out the overall situation renewal, more new formula is such as Under:

τ_ij(t+1)=(1-ρ) τ_ij(t)+ω₁ρΔ₁τ_ij+ω₂ρΔ₂τ_ij(5)；

${\mathrm{\Δ}}_1{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{worst}-L_{best})/L_{best},L_i\∈L_{best}\\ Q/L_i,L_i\∉L_{best}\end{array}\right.---\left(6\right);$

${\mathrm{\Δ}}_2{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{best}-L_{worst})/L_{worst},L_i\∈L_{worst}\\ Q/L_i,L_i\∉L_{worst}\end{array}\right.---\left(7\right);$

In formula (5), (6), (7),

ρ is global information element volatility coefficient；ω₁、ω₂For weight coefficient, and ω₁+ω₂=1；L_bestOptimum for current iteration Path；L_worstFor current iteration worst path length；L_iIt is the i-th paths length；Q is ant colony pheromone total amount；

Step Sf, it may be judged whether reach maximum iteration time (judging whether Formica fusca reaches home)；I.e.

If being not reaching to maximum iteration time, go to step Sc；

Otherwise, terminate iteration, export optimal solution, be the Formica fusca route map through optimal path.

Embodiment 2

On the basis of embodiment 1, the present embodiment 2 additionally provides a kind of mobile robot path planning system, including:

Environmental modeling unit, uses grid environmental modeling according to the local environment information of mobile robot；

Path planning unit, initial information element distributes, and by optimum path search to export optimal path.

Wherein, described environmental modeling unit includes: with the mobile robot of environment information acquisition sensor；

Described mobile robot collecting work environmental information, and carry out Map building；Wherein, will mobile robot and each City, as particle, will be moved robot and barrier and model according to two-dimensional coordinate system.

Described path planning unit is distributed by initial information element and optimum path search is to export optimal path；Wherein

Initial information element distributes, i.e. initial information element is relatively big according to Origin And Destination line near zone concentration, starting point and The principle that the pheromone concentration of two diagonal zones that terminal line is relative is less is allocated；And

Optimum path search, to export optimal path, i.e. carries out optimum path search according to ant group algorithm, after completing once to circulate, to respectively Pheromone on the access path of individual city carries out real-time update, after reaching maximum iteration time, exports optimal path.

Further, described ant group algorithm carries out the method for optimum path search and includes:

Step Sa, arranges initial parameter, including ant colony scale m, pheromone significance level factor-alpha, the important journey of heuristic function Degree factor-beta, pheromone volatilization factor ρ, pheromone release total amount Q, and set maximum iteration time, iterations initial value；

Step Sb, is randomly placed on Formica fusca starting point, and randomly chooses next node to be visited, and wherein Formica fusca is from node I transfers to the computing formula of the probability of node j, i.e. transition probability:

$P_{ij}^k=\left\{\begin{array}{c}\frac{{\[{\mathrm{\τ}}_{ij}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{ij}\left(t\right)\]}^{\mathrm{\β}}}{\underset{s\∈{\mathrm{allow}}_k}{\Σ}{\[{\mathrm{\τ}}_{is}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{is}\left(t\right)\]}^{\mathrm{\β}}},s\∈{\mathrm{allow}}_k\\ 0,s\∉{\mathrm{allow}}_k\end{array}\right.---\left(1\right);$

In formula (1),

allow_k(k=1,2 ..., m) it is the set of Formica fusca k node to be visited, and when starting, allow_kIn have (n-1) Individual element, i.e. includes other all nodes in addition to Formica fusca k starting point, propelling over time, allow_kIn element by The fewest, until being empty, i.e. have access to impact point complete；η_ijT () is heuristic function, represent that Formica fusca transfers to node j from node i Expected degree, computing formula is as follows:

${\mathrm{\η}}_{ij}\left(t\right)=\frac{1}{d_{ij}}---\left(2\right);$

In formula (2), d_ijRepresenting the distance between node i and node j, computing formula is as follows:

$d_{ij}=\sqrt{{(x_i-x_j)}^2+{(y_i-y_j)}^2}---\left(3\right);$

Step Sc, after having calculated internodal transition probability, uses the next joint to be visited of roulette method choice Point；

Step Sd, path walked to Formica fusca carries out local message element amount and updates, dense to weaken the pheromone in path of having passed by Degree, it is as follows that local message element measures more new formula:

τ_ij(t+1)=(1-ε) τ_ij(t)+ετ₀(4)；

In formula (4): ε is local message element volatility coefficient, τ₀=λ/d_ij, λ is constant；

Step Se, when all Formica fuscas all complete one take turns iteration after, pheromone concentration is carried out the overall situation renewal, more new formula is such as Under:

τ_ij(t+1)=(1-ρ) τ_ij(t)+ω₁ρΔ₁τ_ij+ω₂ρΔ₂τ_ij(5)；

${\mathrm{\Δ}}_1{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{worst}-L_{best})/L_{best},L_i\∈L_{best}\\ Q/L_i,L_i\∉L_{best}\end{array}\right.---\left(6\right);$

${\mathrm{\Δ}}_2{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{best}-L_{worst})/L_{worst},L_i\∈L_{worst}\\ Q/L_i,L_i\∉L_{worst}\end{array}\right.---\left(7\right);$

In formula (5), (6), (7),

ρ is global information element volatility coefficient；ω₁、ω₂For weight coefficient, and ω₁+ω₂=1；L_bestOptimum for current iteration Path；L_worstFor current iteration worst path length；L_iIt is the i-th paths length；Q is ant colony pheromone total amount；

Step Sf, it may be judged whether reach maximum iteration time；I.e.

If being not reaching to maximum iteration time, go to step Sc；

Otherwise, terminate iteration, export optimal solution, be the Formica fusca route map through optimal path.

On the basis of embodiment 1 and embodiment 2, by Fig. 2 and Fig. 4, Fig. 3 and Fig. 5 to this mobile robot path planning Method uses the effect of the traditional ant group algorithm after optimizing to illustrate.

According to the optimal route figure of output, Fig. 2 is the optimal route figure of tradition ant group algorithm output, and Fig. 4 is the ant improved Group algorithm output optimal route figure, in conjunction with these two optimal route figures it is seen that, although the road that Fig. 4 shortens relative to Fig. 2 Electrical path length is less, but number of turns significantly reduces, and therefore moves machine task efficiency and is the most just improved, and this is the most just Reach to make up the purpose of tradition ant group algorithm Shortcomings.

Simulation comparison experiment is carried out under simple condition.From Fig. 3 and Fig. 5 it is seen that, tradition ant group algorithm want iteration About 44 times ability approximate convergences are to optimal solution, as long as the ant group algorithm after improvement iterates to about 30 times just can converge to optimum Solving, efficiency has been obviously improved many.Tradition ant group algorithm is done the present invention advantage improved it is clear that ant colony after Gai Jining Algorithm walks many detours by Formica fusca less in early stage, improves the convergence of algorithm early stage, substantially increases the work improving robot Make efficiency and functional reliability.

Method for planning path for mobile robot based on improvement ant group algorithm and system, have quickly receipts in algorithm early stage Holding back property, reduces iterations, improves search efficiency, shortens path, meets artificial planning intention, it is adaptable to mobile robot Independent navigation in a static environment.

With the above-mentioned desirable embodiment according to the present invention for enlightenment, by above-mentioned description, relevant staff is complete Entirely can carry out various change and amendment in the range of without departing from this invention technological thought.The technology of this invention The content that property scope is not limited in description, it is necessary to determine its technical scope according to right.

Claims (9)

1. a method for planning path for mobile robot, it is characterised in that comprise the steps:

Step S1, environmental modeling；

Step S2, initial information element distributes；And

Step S3, optimum path search, and output optimal path.

Method for planning path for mobile robot the most according to claim 1, it is characterised in that

Environmental modeling in described step S1, i.e.

Local environment information according to mobile robot uses grid environmental modeling, and its method includes:

The mobile robot of utilization is from belt sensor group collecting work environmental information, and carries out Map building；Wherein, will mobile machine People and each city, as particle, will be moved robot and barrier and model according to two-dimensional coordinate system.

Method for planning path for mobile robot the most according to claim 2, it is characterised in that

Initial information element distribution in described step S2, i.e.

Initial information element is relatively big according to Origin And Destination line near zone concentration, two diagonal angles that beginning and end line is relative The principle that the pheromone concentration in region is less is allocated.

Method for planning path for mobile robot the most according to claim 3, it is characterised in that

Optimum path search in described step S3, and output optimal path；I.e.

Carry out optimum path search according to ant group algorithm, after completing once to circulate, the pheromone on each city access path is carried out Real-time update, after reaching maximum iteration time, exports optimal path.

Method for planning path for mobile robot the most according to claim 4, it is characterised in that

Described ant group algorithm carries out the method for optimum path search and includes:

Step Sa, arranges initial parameter, including ant colony scale m, pheromone significance level factor-alpha, heuristic function significance level because of Sub-β, pheromone volatilization factor ρ, pheromone release total amount Q, and set maximum iteration time, iterations initial value；

Step Sb, is randomly placed on Formica fusca starting point, and randomly chooses next node to be visited, and wherein Formica fusca turns from node i The computing formula moving on to the probability of node j, i.e. transition probability is:

$P_{ij}^k=\left\{\begin{array}{c}\frac{{\[{\mathrm{\τ}}_{ij}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{ij}\left(t\right)\]}^{\mathrm{\β}}}{\underset{s\∈{\mathrm{allow}}_k}{\Σ}{\[{\mathrm{\τ}}_{is}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{is}\left(t\right)\]}^{\mathrm{\β}}},s\∈{\mathrm{allow}}_k\\ 0,s\∉{\mathrm{allow}}_k\end{array}\right.---\left(1\right);$

In formula (1),

allow_k(k=1,2 ..., m) it is the set of Formica fusca k node to be visited, and when starting, allow_kIn have (n-1) individual unit Element, i.e. includes other all nodes in addition to Formica fusca k starting point, propelling over time, allow_kIn element gradually subtract Few, until being empty, i.e. have access to impact point complete；η_ijT () is heuristic function, represent that Formica fusca transfers to the phase of node j from node i Prestige degree, computing formula is as follows:

${\mathrm{\η}}_{ij}\left(t\right)=\frac{1}{d_{ij}}---\left(2\right);$

In formula (2), d_ijRepresenting the distance between node i and node j, computing formula is as follows:

$d_{ij}=\sqrt{{(x_i-x_j)}^2+{(y_i-y_j)}^2}---\left(3\right);$

Step Sc, after having calculated internodal transition probability, uses the next node to be visited of roulette method choice；

Step Sd, path walked to Formica fusca carries out local message element amount and updates, to weaken the pheromone concentration in path of having passed by, office Pheromone amount more new formula in portion's is as follows:

τ_ij(t+1)=(1-ε) τ_ij(t)+ετ₀(4)；

In formula (4): ε is local message element volatility coefficient, τ₀=λ/d_ij, λ is constant；

Step Se, when all Formica fuscas all complete one take turns iteration after, pheromone concentration is carried out the overall situation renewal, more new formula is as follows:

τ_ij(t+1)=(1-ρ) τ_ij(t)+ω₁ρΔ₁τ_ij+ω₂ρΔ₂τ_ij(5)；

${\mathrm{\Δ}}_1{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{worst}-L_{best})/L_{best},L_i\∈L_{best}\\ Q/L_i,L_i\∉L_{best}\end{array}\right.---\left(6\right);$

${\mathrm{\Δ}}_2{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{best}-L_{worst})/L_{worst},L_i\∈L_{worst}\\ Q/L_i,L_i\∉L_{worst}\end{array}\right.---\left(7\right);$

In formula (5), (6), (7),

ρ is global information element volatility coefficient；ω₁、ω₂For weight coefficient, and ω₁+ω₂=1；L_bestFor current iteration optimal path Length；L_worstFor current iteration worst path length；L_iIt is the i-th paths length；Q is ant colony pheromone total amount；

Step Sf, it may be judged whether reach maximum iteration time；I.e.

If being not reaching to maximum iteration time, go to step Sc；

Otherwise, terminate iteration, export optimal solution, be the Formica fusca route map through optimal path.

6. a mobile robot path planning system, it is characterised in that including:

Environmental modeling unit, uses grid environmental modeling according to the local environment information of mobile robot；

Path planning unit, initial information element distributes, and by optimum path search to export optimal path.

Mobile robot path planning system the most according to claim 6, it is characterised in that

Described environmental modeling unit includes: with the mobile robot of environment information acquisition sensor；

Described mobile robot collecting work environmental information, and carry out Map building；Wherein, will mobile robot and each city As particle, robot and barrier will be moved and model according to two-dimensional coordinate system.

Mobile robot path planning system the most according to claim 7, it is characterised in that

Described path planning unit is distributed by initial information element and optimum path search is to export optimal path；Wherein

Initial information element distributes, i.e. initial information element is relatively big according to Origin And Destination line near zone concentration, beginning and end The principle that the pheromone concentration of two diagonal zones that line is relative is less is allocated；And

Optimum path search, to export optimal path, i.e. carries out optimum path search according to ant group algorithm, after completing once to circulate, to each city Pheromone on city's access path carries out real-time update, after reaching maximum iteration time, exports optimal path.

Mobile robot path planning system the most according to claim 8, it is characterised in that

Described ant group algorithm carries out the method for optimum path search and includes:

Step Sa, arranges initial parameter, including ant colony scale m, pheromone significance level factor-alpha, heuristic function significance level because of Sub-β, pheromone volatilization factor ρ, pheromone release total amount Q, and set maximum iteration time, iterations initial value；

Step Sb, is randomly placed on Formica fusca starting point, and randomly chooses next node to be visited, and wherein Formica fusca turns from node i The computing formula moving on to the probability of node j, i.e. transition probability is:

$P_{ij}^k=\left\{\begin{array}{c}\frac{{\[{\mathrm{\τ}}_{ij}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{ij}\left(t\right)\]}^{\mathrm{\β}}}{\underset{s\∈{\mathrm{allow}}_k}{\Σ}{\[{\mathrm{\τ}}_{is}\left(t\right)\]}^{\mathrm{\α}}\·{\[{\mathrm{\η}}_{is}\left(t\right)\]}^{\mathrm{\β}}},s\∈{\mathrm{allow}}_k\\ 0,s\∉{\mathrm{allow}}_k\end{array}\right.---\left(1\right);$

In formula (1),

allow_k(k=1,2 ..., m) it is the set of Formica fusca k node to be visited, and when starting, allow_kIn have (n-1) individual unit Element, i.e. includes other all nodes in addition to Formica fusca k starting point, propelling over time, allow_kIn element gradually subtract Few, until being empty, i.e. have access to impact point complete；η_ijT () is heuristic function, represent that Formica fusca transfers to the phase of node j from node i Prestige degree, computing formula is as follows:

${\mathrm{\η}}_{ij}\left(t\right)=\frac{1}{d_{ij}}---\left(2\right);$

In formula (2), d_ijRepresenting the distance between node i and node j, computing formula is as follows:

$d_{ij}=\sqrt{{(x_i-x_j)}^2+{(y_i-y_j)}^2}---\left(3\right);$

Step Sc, after having calculated internodal transition probability, uses the next node to be visited of roulette method choice；

Step Sd, path walked to Formica fusca carries out local message element amount and updates, to weaken the pheromone concentration in path of having passed by, office Pheromone amount more new formula in portion's is as follows:

τ_ij(t+1)=(1-ε) τ_ij(t)+ετ₀(4)；

In formula (4): ε is local message element volatility coefficient, τ₀=λ/d_ij, λ is constant；

Step Se, when all Formica fuscas all complete one take turns iteration after, pheromone concentration is carried out the overall situation renewal, more new formula is as follows:

τ_ij(t+1)=(1-ρ) τ_ij(t)+ω₁ρΔ₁τ_ij+ω₂ρΔ₂τ_ij(5)；

${\mathrm{\Δ}}_1{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{worst}-L_{best})/L_{best},L_i\∈L_{best}\\ Q/L_i,L_i\∉L_{best}\end{array}\right.---\left(6\right);$

${\mathrm{\Δ}}_2{\mathrm{\τ}}_{ij}=\left\{\begin{array}{c}(L_{best}-L_{worst})/L_{worst},L_i\∈L_{worst}\\ Q/L_i,L_i\∉L_{worst}\end{array}\right.---\left(7\right);$

In formula (5), (6), (7),

ρ is global information element volatility coefficient；ω₁、ω₂For weight coefficient, and ω₁+ω₂=1；L_bestFor current iteration optimal path Length；L_worstFor current iteration worst path length；L_iIt is the i-th paths length；Q is ant colony pheromone total amount；

Step Sf, it may be judged whether reach maximum iteration time；I.e.

If being not reaching to maximum iteration time, go to step Sc；

Otherwise, terminate iteration, export optimal solution, be the Formica fusca route map through optimal path.