CN103092204B - A kind of Robotic Dynamic paths planning method of mixing - Google Patents

Abstract

The invention discloses a kind of Robotic Dynamic paths planning method of mixing, the method can be applied in environmental information part known and there is unknown dynamic and static state barrier simultaneously when.First obtain global path by a kind of genetic algorithm as Global Planning for above-mentioned situation, then carry out sector planning with the Artificial Potential Field Method improved.In sector planning, invention also contemplates that the Velocity-acceleration information of robot and barrier, so have good effect to process active path planning, in addition also contemplate the effect of Global motion planning path to sector planning.The present invention realizes simply, and the robot path obtained comparatively is optimized, and shows good dirigibility in the application simultaneously.

Description

A kind of Robotic Dynamic paths planning method of mixing

Technical field

The present invention relates to a kind of paths planning method, particularly relate to the paths planning method of a kind of mobile robot in the environment that there is dynamic barrier and static-obstacle thing.

Background technology

The planning of mobile robot's real-time route and navigation are one of key elements reflecting robot autonomous ability, are also one of more scabrous problems.The planning of the planning that robot path planning is mainly divided into environmental information known and environmental information the unknown.Adopt segregation reasons for the former, the path obtained is more excellent more, and the latter is many, and employing is planned online, embodies the real-time of path planning.

In recent years the method for many path plannings study by people.The method of main path planning can be divided into two classes---the method (AI) of artificial intelligence and Artificial Potential Field Method (APF).The method that the former mainly uses has genetic algorithm (GA), fuzzy logic control (FLC) and artificial neural network (ANN), and often comparatively complex calculation speed is also comparatively slow for these methods.And the latter is used widely due to its terseness and rapidity in robot path planning, what its basic thought was impact point in environment to its attractive force and barrier forms a kind of potential field environment to its repulsive force.In dynamic environment, Artificial Potential Field Method mainly contains two kinds for the thinking solving planning problem, a kind of resolving ideas is by propositions such as Fujimura, main thought is as a dimension using temporal information, dynamic disorder rule translation is static programming, but limitation is that the track of dynamic barrier needs known in advance.Another kind of solution is by propositions such as Ko and Lee, main thought is that the velocity information of barrier is incorporated in repulsive potential force function, Ge and Cui has done further improvement on this basis, and the benefit of the method is the track not needing to know in advance barrier, so have good real-time.

The real-life situation situation that component environment information is known often, as under the environment of plant, some operating modes are known, some operating modes are unknown.The path only obtained with above-mentioned online paths planning method is more for optimizing, only comparatively slow by above-mentioned off-line planning method computing velocity, compares and is difficult to process dynamic barrier.

Summary of the invention

The object of the invention is to overcome the deficiencies in the prior art, a kind of Robotic Dynamic paths planning method of mixing is provided.

The object of the invention is to be achieved through the following technical solutions: a kind of Robotic Dynamic paths planning method of mixing comprises the steps:

Step 1: utilize vision sensor to obtain environmental information, comprise the positional information of known quiescent state obstacle information in environment, impact point information and robot self;

Step 2: the environmental information obtained in step 1 represents with Grid Method, obtains grating map, the size of grid depends on planning precision;

Step 3: the path planning grating map genetic algorithm obtained in step 2 being carried out to the overall situation, obtain an overall path, this path is a broken line;

Step 4: extract break and global object point, starting point in the broken line obtained from step 3 as the key point needed for sector planning, these key points are the localized target point of sector planning;

Step 5: utilize the dynamic barrier of mobile robot's periphery and the key point described in step 4 as localized target point, adopt Artificial Potential Field Method (APF) to construct potential field environment, also add the line segment that is formed by connecting by the adjacent key point obtained in step 4 in potential field environment to the attractive force potential field of robot simultaneously;

Step 6: be subject to attractive force and repulsive force, the move under influence of making a concerted effort in the potential field environment that robot constructs in step 5, carry out localized target planning;

Step 7: judge whether the current position of robot arrives the localized target point described in step 4, if arrive localized target point, upgrading impact point is that next key point turns to step 5 to re-construct local potential field environment as localized target point;

Step 8: if current localized target point is that the global object point of environment is then when method after robot arrival global object point terminates;

The invention has the beneficial effects as follows, carry out segregation reasons, plan online circumstances not known known environment, combine both advantages, optimization was both compared in the path after planning, and also had stronger dirigibility to the process of the unknown barriers such as dynamic barrier.The method that the present invention proposes is more applicable for the path planning of actual automatic factory environment.

Accompanying drawing explanation

Fig. 1 is the process flow diagram of the Robotic Dynamic paths planning method of mixing;

Tu2Shi robot is subject to repulsion vector plot.

Embodiment

Describe the present invention in detail below in conjunction with accompanying drawing, object of the present invention and effect will become more obvious.

As shown in Figure 1, the Robotic Dynamic paths planning method of the present invention's mixing comprises the steps:

Step 1: utilize vision sensor to obtain environmental information, comprise the positional information of known quiescent state obstacle information in environment, impact point information and robot self.

This step can adopt Liu Mingshuo. based on the flexible mechanical arm track following of binocular vision odometry. and in Zhejiang University academic dissertation .2011.04, the method for chapter 3 obtains the positional information of Obstacle Position and robot self.

Step 2: the environmental information obtained in step 1 represents with Grid Method, obtains grating map, the size of grid depends on planning precision.

This step can adopt Zhang Handong. the application of Lattice encoding new method in robot path planning. and Central China University of Science and Technology's journal: natural science edition, the method for 2007,35 (1): 50 one 53. represents environmental information Grid Method.

Step 3: the path planning grating map genetic algorithm obtained in step 2 being carried out to the overall situation, obtain an overall path, this path is a broken line.

This step can adopt Zhang Handong. the application of Lattice encoding new method in robot path planning. and Central China University of Science and Technology's journal: natural science edition, the method of 2007,35 (1): 50 one 53. uses genetic algorithm finally to obtain a broken line.

Step 4: extract break and global object point, starting point in the broken line obtained from step 3 as the key point needed for sector planning, these key points are the localized target point of sector planning.

Obtain the break of broken line and starting point and impact point, by its number consecutively 1,2,3 ... N, wherein 1 is starting point, and N is impact point, gets these key points as path planning.

Step 5: utilize the dynamic barrier of mobile robot's periphery and the key point described in step 4 as localized target point, adopt Artificial Potential Field Method (APF) to construct potential field environment, also add the line segment that is formed by connecting by the adjacent key point obtained in step 4 in potential field environment to the attractive force potential field of robot simultaneously.

The potential field environment of structure comprises gravitational field function and repulsion field function, and the function of potential field environment is as follows: U (q)=U _att(q)+U _rep(q), wherein U _attq () is gravitational field function U _repq () is repulsion field function, q is robot location's vector.

About gravitational field function, the present invention introduces velocity information and the acceleration information of robot relative target point on the basis of traditional gravitational field function, and stressed in gesture force field of such robot can adjust according to position, speed, acceleration.In addition, in order to follow the tracks of a preferably route, robot, except the attraction being subject to impact point, is subject to the attraction in the path that has existed all the time, i.e. so-called " line potential field ".Gravitational field function after improvement is: U _att(q, v, a)=α _q|| q-q _g|| ^m+ α _v|| v-v _g|| ⁿ+ α _a|| a-a _g|| ^p+ α _l(|| q-q _line|| ^l+ || v-v _line|| ^l+ || a-a ^line|| ^l) wherein q, v, a are respectively the vector of the position of robot, speed, acceleration, α _q, α _v, α _abe scale-up factor with m, n, p, different values represents the weight of robot and impact point relative position information, relative velocity, relative acceleration information in gravitation function, α _lwhat represent with l is the weight of line potential field.α _v, α _aand α _lwhen being zero then, the gravitational field function of improvement is identical with traditional gravitational field function.Q _g, v _g, a _gfor the position of impact point, speed and acceleration.Q _line=(x ₀, r (x ₀)) ^twherein r (x) equation (x that is curve ₀, r (x ₀)) represent coordinate curve arriving the nearest point of mobile robot's current location, along with this coordinate of movement of robot can corresponding change.V _line, a _linefor speed and the acceleration of the point on path, usual desirable v _line=(0,0) ^t, a _line=(0,0) ^trepresent that robot is 0 to the acceleration of the point of the proximal most position on curve and speed.

In like manner repulsion field also comprises the positional information of robot and barrier and relative velocity and relative acceleration information, and such robot comprehensively above information can obtain that to keep away barrier stressed.Concrete formula is as follows: U _rep(q, v, a)=α _q(1/ ρ _obs-1/ ρ ₀)+α _vv _ro+ α _aa _rowherein be respectively the vector of the position of robot, speed, acceleration, ρ ₀represent the safe distance of barrier to robot, only at ρ ₀scope within the machine talent be subject to the repulsive interaction of barrier.ρ _obs=|| q-q _obs|| for the center of robot is to the distance at the center of barrier, α _v, α _aand α _lscale-up factor, v _roand a _rorepresent the phasor difference of the speed of barrier and the speed of robot respectively, the phasor difference of the acceleration speed of barrier and the acceleration of robot.

Step 6: be subject to attractive force and repulsive force, the move under influence of making a concerted effort in the potential field environment that robot constructs in step 5, carry out localized target planning.

Robot is as follows at potential field environment lower stress: F=F _att+ F _repwherein F _attfor robot is subject to attractive force, F _repfor the repulsive force that robot is subject to.

Can be by attractive force suffered by the gravitation function Tui get robot in step 5 $F_{\mathrm{att}}(q,v,a)=-\▿U_{\mathrm{att}}(q,v,a)=-{\▿}_q{U}_{\mathrm{att}}(q,v,a)-{\▿}_v{U}_{\mathrm{att}}(q,v,a)-{\▿}_a{U}_{\mathrm{att}}(q,v,a)$ Robot is subject to the power F that gravitation has relative position to cause _attq(q), the power F that relative velocity causes _attv(v), the power F that relative velocity causes _atta(a) $F_{\mathrm{attq}}\left(q\right)=-{\∂U}_{\mathrm{att}}(q,v,a)/\∂q={\mathrm{m\α}}_q{\left|\right|q-q_g\left|\right|}^{m-1}{e}_{\mathrm{qrg}}+l{\mathrm{\α}}_l{\left|\right|q-q_{\mathrm{line}}\left|\right|}^{l-1}{e}_{\mathrm{qrline}},$ $F_{\mathrm{attv}}\left(v\right)=-{\∂U}_{\mathrm{att}}(q,v,a)/\∂v$ $={\mathrm{n\α}}_v{\left|\right|v-v_g\left|\right|}^{n-1}{e}_{\mathrm{vrg}}+{\mathrm{l\α}}_l{\left|\right|v-v_{\mathrm{line}}\left|\right|}^{l-1}{e}_{\mathrm{vrline}},$ $F_{\mathrm{atta}}\left(a\right)=-\∂U_{\mathrm{att}}(q,v,a)/\∂a={\mathrm{p\α}}_a{\left|\right|a-a_g\left|\right|}^{p-1}{e}_{\arg }$ $+{\mathrm{l\α}}_{\mathrm{line}}{\left|\right|a-a_{\mathrm{line}}\left|\right|}^{l-1}{e}_{\mathrm{arline}}$ Then F _att(q, v, a)=F _attq(q)+F _attv(v)+F _atta(a), wherein e _qrgrepresent that robot location points to object position, e _vrgrepresent robot speed's vectors directed object velocity, e _argrepresent that robot acceleration points to object acceleration, described in all the other sign synchronization rapid 5.

Can be by repulsion suffered by the repulsion function Tui get robot in step 5 $F_{\mathrm{rep}}(q,v,a)=-{\▿U}_{\mathrm{rep}}(q,v,a)=F_{\mathrm{repq}}\left(q\right)+F_{\mathrm{repv}}\left(v\right)+F_{\mathrm{repa}}\left(a\right).$ F _repq(q), F _repv(v), F _repaa () is respectively robot and barrier relative position, speed, acceleration cause repulsive force. $F_{\mathrm{repq}}={\mathrm{\α}}_q{e}_{\mathrm{or}}/{\mathrm{\ρ}}_{\mathrm{obs}}^2+{\mathrm{\α}}_v{v}_{\mathrm{ro}\⊥}{e}_{\mathrm{ro}\⊥}/{\mathrm{\ρ}}_{\mathrm{obs}}+{\mathrm{\α}}_a{a}_{\mathrm{ro}\⊥}{e}_{\mathrm{ro}\⊥}/{\mathrm{\ρ}}_{\mathrm{obs}},$ F _repv=α _vv _roe _or, F _repa=α _av _roe _or, by F _repqthree α _vv _{ro ⊥}e _{ro ⊥}/ ρ _obs, α _aa _{ro ⊥}e _{ro ⊥}/ ρ _obsbe defined as F respectively _repq1, F _repq2, F _repq3what obtain the repulsion of robot according to above rule makes a concerted effort figure as shown in Figure 2.E _orand e _{ro ⊥}represent that barrier points to the direction of robot respectively, and robot points to barrier by rotating clockwise the direction after right angle, v _{ro ⊥}e _{ro ⊥}and a _{ro ⊥}e _{ro ⊥}represent v _roand a _roat e _{ro ⊥}component on direction.F _repqbe split as e _{ro ⊥}the F in direction _repq2and F _repq3, at e _orthe F in direction _repq1and and F _repqvand F _repqamerge and obtain F _repq3, described in all the other sign synchronization rapid 5.

Step 7: judge whether the current position of robot arrives the localized target point described in step 4, if arrive localized target point, upgrading impact point is that next key point turns to step 5 to re-construct local potential field environment as localized target point.

Robot self position current can be learnt by vision sensor by robot, if robot arrives localized target point then get 1 described in step 4 ... next key point in N, and mark a key point and access.

Step 8: if current localized target point is that the global object point of environment is then when method after robot arrival global object point terminates.

The present invention is directed to the situation that environmental information part is known, have employed a kind of barrier-avoiding method of mixing, first carry out Global motion planning by genetic algorithm and obtain global path, then modified embedded-atom method carry out sector planning.Add velocity information and acceleration information in the artificial potential function improved, not only can enable robot preferably avoiding obstacles but also robot can be made to arrive impact point to be unlikely to speed excessive.The artificial potential function improved also add line potential field, and for following the tracks of expected path preferably, this method simple possible, can meet the requirement of robot real-time route planning.

Claims (1)

1. a Robotic Dynamic paths planning method for mixing, is characterized in that, comprise the steps:

Step 1: utilize vision sensor to obtain environmental information, comprise the positional information of known quiescent state obstacle information in environment, impact point information and robot self;

Step 2: the environmental information obtained in step 1 represents with Grid Method, obtains grating map, the size of grid depends on planning precision;

Step 3: the path planning grating map genetic algorithm obtained in step 2 being carried out to the overall situation, obtain an overall path, this path is a broken line;

Step 4: extract break and global object point, starting point in the broken line obtained from step 3 as the key point needed for sector planning, these key points are the localized target point of sector planning;

Step 5: utilize the dynamic barrier of mobile robot's periphery and the key point described in step 4 as localized target point, adopt Artificial Potential Field Method (APF) to construct potential field environment, also add the line segment that is formed by connecting by the adjacent key point obtained in step 4 in potential field environment to the attractive force potential field of robot simultaneously; This step is specially: the function of structure potential field environment is as follows: U (q)=U _att(q)+U _rep(q), wherein U _attq () is gravitational field function, U _repq () is repulsion field function, q is robot location's vector;

On the basis of traditional gravitational field function, wherein introduce velocity information and the acceleration information of robot relative target point, stressed in gesture force field of such robot can adjust according to position, speed, acceleration; In addition, in order to follow the tracks of a preferably route, robot, except the attraction being subject to impact point, is subject to the attraction in the path that has existed all the time, and improving its formula is U _att(q, v, a)=α _q|| q-q _g|| ^m+ α _v|| v-v _g|| ⁿ+ α _a|| a-a _g|| ^p+ α _l(|| q-q _line|| ^l+ || v-v _line|| ^l+ || a-a _line|| ^l), wherein, q _g, v _g, a _gfor the position of impact point, speed and acceleration; q _line=(x ₀, r (x ₀)) ^t, the wherein equation of ideal path curve that obtains for Global motion planning of r (x), (x ₀, r (x ₀)) represent coordinate curve arriving the nearest point of mobile robot's current location, along with this coordinate of movement of robot can change accordingly; v _line, a _linefor speed and the acceleration of the point on path; || q-q _g|| ^m, || v-v _g|| ⁿ, || a-a _g|| ^prepresent the potential field information of the relative position between robot and impact point, relative velocity, relative acceleration respectively; M, n, p are positive integer, and be worth larger, the intensity of these potential field information is higher; α _q, α _v, α _abe scale-up factor, represent the weight of the potential field information of above-mentioned relative position, relative velocity, relative acceleration in gravitation function; Similarly, || q-q _line|| ^l, || v-v _line|| ^l, || a-a _line|| ^lrepresent the potential field information of the relative position between robot and ideal path curve, relative velocity, relative acceleration respectively, they with expression line potential field information; L is positive integer, is worth larger, and the intensity of line potential field information is higher; α _lbe the real number between 0 to 1, expression be the weight of line potential field information;

In like manner repulsion field also comprises the positional information of robot and barrier and relative velocity and relative acceleration information, and such robot comprehensively above information can obtain that to keep away barrier stressed; Concrete formula is as follows: U _rep(q, v, a)=α _q(1/ ρ _obs-1/ ρ ₀)+α _vv _ro+ α _aa _ro, wherein q, v, a are respectively the vector of the position of robot, speed, acceleration, ρ ₀represent the safe distance of barrier to robot, only at ρ ₀scope within the machine talent be subject to the repulsive interaction of barrier; ρ _obs=|| q-q _obs|| for the center of robot is to the distance at the center of barrier, α _v, α _aand α _qbe scale-up factor, represent the weight of the potential field information of above-mentioned relative velocity, relative acceleration, relative position in repulsion function, v _roand a _rorepresent the phasor difference of the speed of barrier and the speed of robot respectively, the phasor difference of the acceleration of barrier and the acceleration of robot;

Step 6: be subject to attractive force and repulsive force, the move under influence of making a concerted effort in the potential field environment that robot constructs in step 5, carry out localized target planning;

Step 7: judge whether the current position of robot arrives the localized target point described in step 4, if arrive localized target point, upgrading impact point is that next key point turns to step 5 to re-construct local potential field environment as localized target point;

Step 8: if current localized target point is that the global object point of environment is then when method after robot arrival global object point terminates.