CN111452052A - Navigation positioning and task planning method of robot for files - Google Patents

Abstract

The invention discloses a navigation positioning and task planning method of a robot for files, which comprises the following steps (1) of preprocessing: defining coordinate axes and initial parameters of a file rack area; (2) the robot starts from an initial position and passes through a file area; (3) judging whether the robot has passed through the file shelf with the largest number in the task; (4) judging whether the task is an executable task or not according to the task matrix to be executed, (5) decomposing the task to form a new task matrix; (6) establishing an objective function, and optimizing a new task matrix; (7) and executing the task. The invention can solve the problem that the robot is difficult to identify when the file rack is opened, so that the robot can monitor the change of the external environment in real time, effectively and quickly adapt to the changed external environment, is suitable for different maps, has wide applicability and can realize efficient human-computer interaction between the client and the robot.

Description

Navigation positioning and task planning method of robot for files

Technical Field

The invention relates to a navigation positioning and task planning method, in particular to a navigation positioning and task planning method of a robot for files.

Background

The number of files stored in the file room is thousands, and a large amount of manpower and material resources are often consumed when the files are accessed, so that the exertion of the professional technical ability of a file administrator is restricted, the available value of the files is reduced, and the automation level of file management is urgently needed to be improved. In recent years, in the file industry, it has been widely attempted to perform the checking and access of files by using a robot instead of a human.

The file robot is a highly automated file device, generally comprising a robot moving chassis, an operation mechanical arm and functional equipment, wherein the robot moving chassis is responsible for realizing walking, replaces the functions of human legs, and realizes autonomous navigation walking and target identification; the mechanical arm is equivalent to the arm of a person and is responsible for driving the functional equipment to do spatial motion; functional equipment such as a mechanical arm and various detection equipment is used for realizing automatic counting and access.

At present, an archive robot generally adopts an intelligent vehicle (AGV) based on an S L AM technology, the equipment can identify the surrounding environment through intelligent training to realize autonomous navigation in a fixed environment, when the environment changes, if a target is not at the original fixed position after moving, retraining or programming is needed, and the self-adaptability is poor.

In practice, files are stored in densely packed file racks as soon as they are stored, usually each file rack is closely packed together, and when a file is to be accessed, the corresponding target compact rack is opened. The method is limited by the current technology and cost factors, but when a plurality of file racks are opened, the number of each working area cannot be guaranteed, and the width of each working area cannot be guaranteed to be consistent, so that the positions of each file rack and a working channel thereof are random, the positions of the working areas are changed, the target identification of the robot is difficult, and the human-computer interaction of a client is difficult. In order to solve the problems, the invention provides a target position control method for a robot to be self-adaptive to the environment.

Disclosure of Invention

The purpose of the invention is as follows: the invention aims to provide an efficient and environment-adaptive navigation positioning and task planning method of a file robot, so as to solve the problems.

The technical scheme is as follows: the invention discloses a navigation positioning and task planning method of a robot for files, which comprises the following steps:

(1) pretreatment: defining coordinate axes and initial parameters of a file rack area; the robot acquires a task instruction of a target file and defines an executable task matrix as TE;

(2) starting from an initial position, the robot penetrates through a file area, detects the number of left and/or right file racks, the width of a working area and whether the file racks are opened or not, and records coordinate parameters;

(3) judging whether the robot has passed through the file shelf with the largest number in the task;

(4) judging whether the task is an executable task or not according to the task matrix to be executed, if so, forming an executable task matrix and executing the step (5), otherwise, deleting the task;

(5) rearranging the task execution sequence according to the map of the archive region, and decomposing the tasks according to the number of tasks that the robot can process once to form a new task matrix;

(6) establishing an objective function, and optimizing a new task matrix;

(7) and (4) placing the robot at an initial position, and executing the task according to the task matrix optimized in the step (6).

Has the advantages that: compared with the prior art, the invention has the following remarkable advantages:

(1) the invention can solve the problem that the robot is difficult to identify when the file rack is opened, so that the robot can monitor the change of the external environment in real time, effectively and quickly adapt to the changed external environment, is suitable for different maps, has wide applicability and can realize efficient human-computer interaction between the client and the robot.

(2) The task planning method of the invention enables the robot to be more efficient and orderly in the task execution process, and improves the task execution efficiency by rearranging and optimizing the tasks issued by the operators.

Drawings

FIG. 1 is a schematic structural diagram of a robotic device;

FIG. 2 is a schematic diagram of a file area;

FIG. 3 is a schematic view of the opening of a file rack in the file area;

fig. 4 is a process of building the map matrix shown in fig. 3 by the robot.

Detailed Description

The technical scheme of the invention is further explained by combining the attached drawings.

The invention is carried out with a robot device as shown in fig. 1.

The robot 1 is provided with a distance measuring instrument group 2, the distance measuring instrument comprises four laser distance measuring instruments, namely X1, X2, Y1 and Y2, the measuring ranges of the four laser distance measuring instruments are 0-50 m, the measuring accuracy is 1mm, and the response frequency is 10Hz at most. Also, X1, X2, Y1, and Y2 represent readings from four sensors.

Fig. 2 is a schematic layout of a large archive, the outermost black box is the wall of the warehouse, and the lower part of the diagram is the robot parking area. The storage has four file rack groups, and the files are assembled in the storage state and are opened only when being accessed. As shown in fig. 1, the right side of the file rack is provided with N file racks, wherein the number 1 to 8 file racks are combined together, the number 4 file rack is an immovable file rack, the number 9 to N-1 file racks are combined together, the number N file rack is opened, and the number 13 file rack is an immovable file rack; as shown in the figure, the left side of the drawing has M file frames, No. 1-8 file frames are one group, No. 9-M file frames are another group, and No. 4 and No. 12 file frames are fixed. A larger passageway is reserved between the two groups of file racks, for example, a larger passageway is reserved between the No. 8 rack and the No. 9 rack on the left side in the figure, and a larger main channel passageway is also reserved between the No. 8 and the No. 9 file racks on the left side and the right side in the figure 1. In different archives, the number of the archives in each group is not necessarily equal, nor is 8, and the positions of the archives fixing frames in each group are different, some are foremost and some are rearmost, and are determined by each archives during design and construction.

For more representative purposes, the steps are explained by taking the case that the robot is placed in the middle, and the invention is also applicable to the case that the robot moves between the left and right file racks and the wall.

(1) Pretreatment: defining coordinate axes and initial parameters of a file rack area; and defining a task matrix according to the task instruction of the robot for acquiring the target file.

(11) Defining a file rack width matrix

The dimension of each shelf in the Y direction is called the width, denoted shelf number WK, K1, 2,3,4 … …, as WA1 denotes the width of the left side shelf No. 1. Generally, the dimensions of the file racks in an archive are not too great in each category, and usually are at most three or four. Defining a width matrix

W1 ═ WA1 WA2 … … … … WAM ] indicates the left M shelf widths in the figure;

w2 ═ WB1 WB2 … … … … WBN indicates the right N shelf widths in the figure.

(12) The actual value of the width of each shelf is in error from the nominal value and is noted as Δ w.

(13) Defining a file rack position matrix:

the initial position of each file rack is represented by the position of the file rack parallel to the center line of the X axis, as shown in the figure, the initial position of the No. 5 file rack on the right side is PB5, and the positions of the file racks on the two sides are respectively recorded by a matrix:

PA[M]＝[PA1 PA2………PAM],PB[N]＝[PB1 PB2………PBN]

(14) defining a file rack map variable:

define a 2M +1 bit map variable: MA (011111111011111111110000000000000000)

Define a 2N +1 bit map variable: MB is [ 01111111101111111110000000000000000 ]

In the variables, 0 represents a gap and 1 represents a file rack. For example: 101 denotes a situation where there is a gap between two file shelves, the above formula is an initial state, and the file shelves are all closed. The last 0 in the matrix is artificially added and does not represent the number of spaces.

(15) Defining a distance matrix of a file rack in a working area:

DA[M-1]＝[L1 L2………………Lm-1],DB[N-1]＝[L1 L2………………..Ln-1]

each element of the distance matrix DA, DB represents the distance between two adjacent shelves, in mm, with the initial value being all zero, for example, the first element L1 of the DB matrix represents the distance between the right 1 and 2 shelves, as shown in fig. 1, L1 ═ 0, which indicates that the right 1 and 2 archive shelves are not open, and the distance is 0.

(16) Other error variables are defined:

defining a maximum measurement error of the robot for all the rack spacings, denoted Δ L, L denotes the rack spacing necessary to get the robot into the aisle, defining an error variable ER, the number of tasks that the robot can perform at one time being TM;

(17) the customer tasks are expressed in a matrix:

a user of the archive robot usually issues a task instruction of the robot on the user's client, and if the robot is required to retrieve some archives from the archive, the user retrieves the archives by file names, and the archives are generally stored in different archive shelves in the archive out of order. The client software reads the serial number of the file rack from the database to form a file fetching instruction of the client.

The conference task matrices TA0, TB0 represent the task sequences of the left and right archive racks, respectively, both matrices being column matrices, each element of the matrix representing a work position. For example: TA0[ 13332 ] indicates that some operation task needs to be performed on the left side archive shelf, where workspace No. 3 has three tasks. The position of workspace is provided with 1, 3, 2 threely, and wherein 1 workspace is located between 1,2 archives framves in the left side, and 3 workspaces are located between 3,4 archives framves in the left side, and 2 workspaces are located between 2,3 archives framves in the left side.

The client commands are arranged in the order thought by the client, and the tasks of the left and right file frames are also required to be respectively arranged in ascending order of the working area numbers, such as TA0[ 15552 ] which is changed into TA0[ 12555 ].

Tasks in the same working area in the sequence are merged and represented, and matrixes TN1 and TN2 are defined to represent the number of tasks, TN 1m represents the number of TA 0m, TN 2m represents the number of TB 0m, and dimensions of TA1 and TB1 and SDA and SDB are obtained.

For example: after TA0 merge, TA0 ═ 125, [ TN1 ═ 113 ], TA0[3] ═ 5, and TN1[3] ═ 3, indicating that there are 3 tasks in left work area No. 5. TA0 has three elements, 1,2, and 5, and SDA is 3.

(18) Defining task matrixes TA and TB and executable tasks TE:

an SD × 4 matrix TE is defined, column 1 indicating the tasks to be performed, indicated by the work area number, column 2 indicating the number of tasks, column 3 indicating the position of the work area, column 4 indicating the left and right side of the work, 0 indicating the left side, and 1 indicating the right side.

The procedure of the pretreatment step is detailed:

in this embodiment, the client instruction is: [ left 16, right 2, left 12, right 16, left 8, right 16, left 6, right 16, left 12, right 16, left 6, right 5, left 6, right 16, left 14, right 16, left 3, right 16, left 1, left 5, right 16, left 1, right 16]

Expressed and ordered by a matrix:

TA0＝[1、1、3、5、5、5、5、5、5、5、5、5、5、5、5、6、6、6、8、12、12、14、16]

TB0＝[2、5、16、16、16、16、16、16、16、16、16]

merging the same working areas:

examples are: TA1 ═ 13568121416%

TN1＝[2 1 12 3 1 2 1 1]

TB1＝[2 5 16]

TN2＝[1 1 9]

Examples are: TA1 has 8 elements, SDA ═ 8, TB1 has 3 elements, SDB ═ 3

In the present embodiment, if the above tasks are all executable, then

TE＝[(1，2，P1，0)(3，1，P3，0)(5，12，P5，0)(6，3，P6，0)(8，1，P8，0)(12，2，P12，0)(14，1，P14，0)(16，1，P16，0)(2，1，P2，1)(5，1，P5，1)(16，9，P16，1)]

(2) The robot starts from the initial position, passes through the file area, detects the number of left and/or right file racks, the width of the workspace and whether a file rack is open or not, and records X, Y the coordinate parameters.

(201) Initializing variables, wherein m is equal to n is equal to 1, a file rack counter K1 is equal to 0, and K2 is equal to 0; when the mobile robot starts from the home position, the detected values X1 and X2, which respectively represent the distances from the robot to the left and right walls, are denoted as X10 and X20. The value of Y1 decreases as the mobile robot moves forward.

(202) The mobile robot intermittently moves forward, continuously detects X1 and X2, when X1 is not equal to X10, the detection indicates that the file rack is detected, and the value of record Y1 is Y10, namely: y10 ═ Y1, pa (m) ═ Y2+ W1(m)/2, while X10 ═ X1;

(203) the mobile robot intermittently moves forward, continuously detects X1 and X2, when X2 is not equal to X20, the detection indicates that the file rack is detected, and the value of record Y1 is Y20, namely: y20 ═ Y1, pb (n) ═ Y2+ W2(n)/2, while X20 ═ X2;

(204) the mobile robot continues to move forwards, continuously detects X1 and X2, and when X1 is not equal to X10, the mobile robot indicates that the robot has moved over the left file rack, and detects that the left file rack is opened in the working area. The Y1 value at this time needs to be recorded, so the last Y10 is stored by Y11, i.e., Y11 is changed to Y10; the current Y1 value is then recorded again as Y10, i.e.: y10 ═ Y1, and Y11 ═ Y11-Y1 were calculated, and represent the total width when several shelves were combined;

(205) if Y11-W1(m) | > (K1+5) × Δ W, namely the width of one shelf subtracted from the total width margin is larger than a measurement error, Y11-Y11-W1 (m), m-m +1, PA (m) -PA (m-1) + [ W1(m-1) + W1(m) ]/2, and K1-K1 +1 continue to circulate the fifth step, or else, the number of the detected left-side file frames is calculated, m file frames are detected at present, K1 is made to be 0, and the sixth step is shifted;

(206) from the matrix MA, m 1 are counted from left to right, whether the position behind the mth 1 is zero or not is detected to identify whether the mobile robot walks to the position of the original aisle or not, if the position is not zero, since the left archive shelf is detected to be opened at the moment, 0 is inserted to indicate the mark behind the 1, and meanwhile, the last 0 in the MA is deleted to ensure the number of digits of the MA.

(207) The mobile robot continues to move forwards, continuously detects X1 and X2, and when X2 is not equal to X20, the mobile robot indicates that the robot has moved over the file rack on the right side, and detects that the work area way opened by the file rack on the right side. The Y1 value at this time needs to be recorded, so the last Y20 is stored by Y21, i.e., Y21 is changed to Y20; the current Y1 value is then recorded again as Y20, i.e.: y20 ═ Y1:

(208) if Y21-W2(m) | > (K1+5) × Δ W, Y21 is Y11-W2 (n), n is n +1, PB (n) -PA (n-1) + [ W1(n-1) + W1(n) ]/2, and K2 is K2+1, continuing to circulate the eighth step, otherwise, indicating that the number of the left-side file frames is detected to be calculated, detecting n file frames at present, making K2 equal to 0, and then going to the ninth step;

(209) from the matrix MB, n1 are counted from left to right, whether the position behind the 1 is zero or not is detected to identify whether the mobile robot walks to the position of the original aisle or not, if not, since the right-side file shelf is detected to be opened at the moment, 0 is inserted to indicate the mark behind the 1, and the last 0 in the MB is deleted at the same time to ensure the digit of the MB.

(210) The mobile robot intermittently moves forward, continuously detects X1 and X2, and when X1 is not equal to X10, the mobile robot indicates that the robot has moved through the working channel in the middle of the left file rack and detects the left file rack again. The Y1 value at this time needs to be recorded, so the last Y10 is stored by Y11, i.e., Y11 is changed to Y10; the current Y1 value is then recorded again as Y10, i.e.: y10 ═ Y1: the distance the work area is open is calculated and stored in matrix D1, D1[ m ] (Y11-Y1) in order to record this channel width.

(211) The mobile robot intermittently moves forward, continuously detects X1 and X2, and when X2 is not equal to X20, the robot is indicated to have passed through the working channel in the middle of the right file rack and detect the right file rack again. The Y1 value at this time needs to be recorded, so the last Y20 is stored by Y21, i.e., Y21 is changed to Y20; the current Y1 value is then recorded again as Y20, i.e.: y20 ═ Y1: the distance the work area is open is calculated and stored in matrix D2, D2 n ═ Y21-Y1, to record this channel width.

Step (2) can sufficiently solve the situation that a plurality of file shelves shown in fig. 3 are opened, and when only one shelf is opened in each area, the space is easy to find, for example, a working area channel is arranged between the 16 th shelf and the N th shelf on the right side in the figure. However, this situation is often inconvenient, firstly, the robot has high working efficiency, and the file rack is inconvenient to open and close once after completing one job, and especially under the multitask condition, the file racks are objectively required to be opened simultaneously; secondly, there is not accurate position control device in present file frame system because of cost reason, has opened different quantity shelves respectively also because of various reasons during the in-service use, and the distance between the shelf is uncertain. Because a plurality of shelves are opened, the position of a working area is uncertain before each work, and if the robot cannot correctly judge a target working area, the robot can fail to work or even cause faults such as collision and the like. Therefore, the robot needs to recognize that those spaces are accessible, and the left and right side distances of the robot are measured, updated, and recorded in real time in step (2), so that the above problems can be overcome.

(3) Whether the robot has passed through the file shelf with the largest number in the task is judged, and the execution result of the robot is shown in fig. 4. Fig. 4 is an example of fig. 3, which shows a process of establishing a map matrix. On the left side of the diagram is the MA variation, the last row is the final result, on the right side is the MB variation, and the last row is also the final result.

(31) If M and N are satisfied simultaneously, the archive map, namely the archive region, is detected, the step (4) is carried out, otherwise, the step (32) is carried out;

(32) judgment of Y₁And (3) if the safety value is larger than the safety value, if so, returning to the step (2), otherwise, returning the robot to the initial position for standby.

(4) Judging whether the task is an executable task or not according to the task matrix to be executed, if so, forming an executable task matrix and executing the step (5), otherwise, deleting the task;

(41) screening out executable tasks on the left side

When DA [ TA1(m)]<L (m is 1-SDA), it means that the width between file shelves is not enough, the task cannot be executed, and the task is TA₁Deleted from TN at the same time₁Deleting TN₁(m), SDA is SDA-1. Where (m) is the variable index in TA1, and TA1(m) represents the value designated by m, which in turn points to the variable for DA as the index.

For example, when m is 2, TA1(2) is 3, DA (3) is 0, and TA1(2) and TN1(2) are deleted, then:

TA1＝[1 5 6 8 12 14 16]，

TN1＝[2 12 3 1 2 1 1]

SDA＝7。

in addition, it can also be seen that the left archive shelf # 15 and 16 is also open, but no tasks are scheduled

(42) Sifting out right-hand executable tasks

When DB [ TB1(n)]<L (n is 1 to SDB), it means that the width between file shelves is not enough, the task cannot be executed, and the task is switched from TB₁Deleted from TN at the same time₂Deleting TN₂(n), and making SDB ═ SDB-1.

For example, when n is 1 or 2, TB1(1) is 2, TB1(2) is 5, DB (2) is 0, DB (5) is 0, and TB1(1) and TB1(2) are deleted: TB1 ═ 16], TN2 ═ 9], SDB ═ 3-2 ═ 1

(43) Forming an executable task matrix TE:

each row of the TE matrix corresponds to one task information:

wherein pre-SDA behavior:

TE[m，1]＝TA1(m)，TE[m，2]＝TN1(m)，

TE[m，3]＝PA{TA1(m)+W1[TA1(m)]/2+DA[TA1(m)]/2；

TE[m，4]＝0；

m＝1～SDA；

wherein: TE m,1 is an element of the mth row and 1 st column of the matrix TE, the value of which indicates the position of the task, for example 5, which indicates that the task is executed between two shelves numbered 5 and 6, and TE m,2 is an element of the mth row and 2 nd column of the matrix TE, the value of which indicates the number of tasks. TE m,4 is an element of the m-th row and 4-th column of the matrix TE, and has only two values of 0 and 1, 0 indicating a shelf with a task area on the left side in fig. 3 and 1 indicating a shelf on the right side.

The following description is the same, with the difference that: here m is a variable with the maximum value SDA, in order to write tasks on both shelves in one matrix, so the row of the lower matrix element is SDA + n, and the other small difference is stated above, i.e. TE [ SDA + n, 4] is 1, indicating that the task area is on the shelf on the right side;

next there is again SDB line:

TE[SDA+n，1]＝TB1(n)，TE[SDA+n，2]＝TN2(n)，

TE[SDA+n，3]＝PA{TB1(n)+W1[TB1(n)]/2+DB[TA1(n)]/2；

TE[SDA+n，4]＝1；

n＝1～SDB；

for convenience of expression, two pointers p and k are defined to point to rows and columns of the TE matrix, and initial values are set to be 1; thus, one row of TE can be expressed as TE (p), and the element of p rows and k columns in TE can be expressed as TE (p, k).

For example:

when m is equal to 1, TE [1, 1] ═ TA1(1) is equal to 1, TE [1, 2] ═ TN1(1) is equal to 2,

TE [1, 3] denotes the region position of the left working region No. 1, and is denoted by P1

TE[1，4]＝0；

TE [1, ═ is (1, 2, P, 1)

The result of TE after the end of the fifteenth step is 8 × 4:

TE＝[(1，2，P1，A)(5，12，P5，A)(6，3，P6，A)(8，1，P8，A)(12，2，P12，A)(14，1，P14，A)(16，1，P16，A)(16，9，P16，B)]

TE(1)＝(1，2，P1，A)，TE(1,3)＝P1

TE(2)＝(5，12，P5，A)TE(2,1)＝5

sixteenth, step: reordering tasks

TE [ m, 3] represents the position of a working area capable of executing tasks, the rows of the TE matrix are reordered according to the numerical value of the TE [ m, 3] in the real-time map, the numerical value is smaller before, and the numerical value is larger after;

the result of this example implementation: TE [ (1, 2, P1, a) (5, 12, P5, a) (6, 3, P6, a) (8, 1, P8, a) (12, 2, P12, a) (14, 1, P14, a) (16, 9, P16, B) (16, 1, P16, a) ]

(5) Rearranging the task execution sequence according to the map of the archive region, and decomposing the tasks according to the number of tasks that the robot can process once to form a new task matrix;

(51) TE [ m, 3] represents the position of a working area capable of executing tasks, the rows of the TE matrix are reordered according to the numerical value of the TE [ m, 3], the numerical value is smaller before, and the numerical value is larger after;

(52) defining the number of tasks which can be processed by the robot at one time as TM, defining the task quantity which exceeds TM as a large task, defining the task quantity which is equal to TM as a full task, and defining the task quantity which is less than TM as a small task;

(53) and (4) splitting the large task into a plurality of full tasks and a small task, and circulating the steps until all the large tasks are split.

If the first task in the queue is a large task, a plurality of full tasks are firstly disassembled, and the rest small tasks are later disassembled. And (4) firstly tearing out a small task from the large tasks at other positions in the queue, and keeping the rest of the small tasks as a plurality of full tasks. Specifically, if there is one big task: (TE1, TE2, TE3, TE4) the number of rows of the new matrix is taken as SDE. Then: TE2 is divided by TM, with the quotient Q and the remainder R.

Thus, the decomposition is the result:

if this task is the first row of the TE matrix, the result is

(TE1，TM，TE3，TE4)…(TE1，TM，TE3，TE4)(TE1，R，TE3，TE4)

The front is Q full tasks (TE1, TM, TE3, TE4) and the back is a small task (TE1, R, TE3, TE4)

If this task is not the first row of the TE matrix, the result is

(TE1，R，TE3，TE4)(TE1，TM，TE3，TE4)…(TE1，TM，TE3，TE4)

The front is a small task (TE1, R, TE3, TE4) and the back is Q full tasks (TE1, TM, TE3, TE 4).

After decomposition, the rows of the original matrix are incremented by Q. SDE ═ SDE + Q.

The implementation process of the step (5) is described in detail:

assuming that the number TM of executable tasks of the robot at a time is 8, more than 8 are large tasks, 8 are full tasks, and less than 8 are small tasks

Original TE [2] ═ 5, 12, P5, a) is decomposed into TE [2] ═ 5, 4, P5, a) and TE [3] ═ 5, 8, P5, a)

Original TE [7] ═ (16, 9, P16, B) decomposes into: TE [8] ═ (16, 1, P16, B) and TE [9] ═ 16, 8, P16, B)

The final task matrix is then:

TE＝[(1，2，P1，A)(5，4，P5，A)(5，8，P5，A)(6，3，P6，A)(8，1，P8，A)(12，2，P12，A)(14，1，P14，A)(16，1，P16，B)(16，8，P16，B)(16，1，P16，A)]

because two large tasks are decomposed into two small tasks, namely two full tasks, the original TE is an 8 × 4 matrix, and the two full tasks become a 10 × 4 matrix after decomposition.

(6) And establishing an objective function and optimizing the new task matrix.

(61) Taking the first task TE [1] at the top as [ TE (1, 1), TE (1, 2), TE (1, 3), TE (1, 4) ], and the task number is TE (1, 2); taking several task combinations TECO (i) in the tasks of the task matrix, and the combination condition can be expressed as: summing column 2 of the TECO matrix, no more than TM-TE (p, 2); wherein TM is the number of tasks that the robot can process at a time, and TE (p, 2) is the element of p rows and 2 columns in TE;

when TE (P, 2) ═ TM, TECO is an empty matrix;

TE (P, 2) < TM, TECO is a set of matrix sequences of unequal number of rows, denoted as TECO (i), of several forms:

there is only one row, i.e. there is only one task in the combination:

TECO(1)＝[TE(x，1)，TE(x，2)，TE(x，3)，TE(x，4)]

there are multiple rows, i.e. there are only multiple tasks in a group:

TECO(1)＝[TE(x₁，1)，TE(x₁，2)，TE(x₁，3)，TE(x₁，4)

TE(x₂，1)，TE(x₂，2)，TE(x₂，3)，TE(x₂，4)

TE(x₃，1)，TE(x₃，2)，TE(x₃，3)，TE(x₃，4)

TE(x₄，1)，TE(x₄，2)，TE(x₄，3)，TE(x₄，4)

…………………]

in the formula, x₁、x₂、x₃、x₄The frame positions of the task area are all positive integers, 1<x、x₁、x₂、x₃、x₄≤SDE

The number of rows of each TECO (i) is DCO (i), and the TECO (i) represents the task of the ith row;

(62) calculating an objective function aim (i):

wherein j is 1-DCO (i), and S is a coefficient related to the spatial distribution density of the task; j is a control variable set for summation for selecting a row in the matrix teco (i);

(63) solving max [ AIM (i) ], max [ AIM (i) ]) as the optimal task combination, combining with TE (p) to be the planning result of the task, and expressing as follows by using a block matrix: TEOP [ te (p) yeco (i) ];

deleting the tasks in the selected combination TEOP from the executable task matrix TE, wherein the tasks left in the TE are unplanned tasks;

(64) and repeating the steps until all the matrixes in the TE matrix are deleted.

The implementation process of the step (6) is described in detail:

if the robot can perform 8 tasks at a time, TE (1) ═ 1,2, P1, a), TE · 2 ═ 2, is to complete two tasks in left workspace No. 1. TM-TE · 2 equals 6, which can be added with 6 tasks later. Then the possible combinations of this task are:

TECO(1)＝TE(2)＝[5，4，P5，A]

TECO(2)＝TE(2)+TE(5)＝[(5，4，P5，A)(8，1，P8，A)]

TECO(3)＝TE(2)+TE(5)+TE(6)＝[(5，4，P5，A)(8，1，P8，A)(12，2，P12，A)]

there are also many combinations that must be optimized as an objective function. Assuming that the selected optimal combination is TECO (3), the optimal task combination is as follows: [ (1, 2, P1, a) (5, 4, P5, a) (8, 1, P8, a) (12, 2, P12, a) ], deletion of this combination from TE, then:

TE＝[(5，8，P5，A)(6，3，P6，A)(14，1，P14，A)(16，1，P16，B)(16，8，P16，B)(16，1，P16，A)]

(7) the position where the robot should stop before completing a task combination is calculated, and some initial actions of the robot are completed at the position. Assuming that the control variable K1 is 1 and K2 is 1, the position P0 of the robot is initially calculated:

comparing the actual positions of the smallest task numbers in TE1[ K1], TE2[ K2 ]:

for TE1[ K1] <4, PTE1 ═ 1,

p1 ═ P1- Σ [ D1(1), … … … D1(3) ] (random position)

TE1 (K1) 4 ≦ 8, PTE1 ═ 4 (fixed position)

TE1 ≤ 9 [ K1] <11, PTE1 ═ 9 (random position)

P1＝P1-Σ[D1(9)，………D1(11)]

TE1 (K1) 12 ≦ M (PTE 1 ═ 12 (fixed position)

For TE2[ K ] <4, PTE2 ═ 1,

P2＝P2-Σ[D2(1)，………D2(3)]

TE2 (K2) 4 ≦ 8, PTE2 ═ 4 (fixed position)

TE2 ≤ 9 [ K2<11, PTE2 ═ 9 (random position)

P2＝P2-Σ[D2(1)，………D2(3)]

TE2[ K2] < N, PTE2 ═ 12 (fixed position)

Get the position P of the robot min (P1, P2)

Setting the position of the robot, as in this example 1 in TE1 as minimum, the robot crosses the left workspace 1, moves to the position aligned with the archive shelf No. 1 in the left workspace, and uses this archive shelf to calibrate its attitude.

(8) And (4) moving the initial position of the first task combination by the robot, and starting to execute the task according to the task matrix optimized in the step (6).

(9) And the robot returns to the starting position, reports the task state to the control center and stands by.

Claims (8)

1.A navigation positioning and task planning method of a robot for files is characterized by comprising the following steps:

(1) pretreatment: defining coordinate axes and initial parameters of a file rack area; the robot acquires a task instruction of a target file and defines an executable task matrix as TE;

(2) starting from an initial position, the robot penetrates through a file area, detects the number of left and/or right file racks, the width of a working area and whether the file racks are opened or not, and records coordinate parameters;

(3) judging whether the robot has passed through the file shelf with the largest number in the task;

(4) judging whether the task is an executable task or not according to the task matrix to be executed, if so, forming an executable task matrix and executing the step (5), otherwise, deleting the task;

(5) rearranging the task execution sequence according to the map of the archive region, and decomposing the tasks according to the number of tasks that the robot can process once to form a new task matrix;

(6) establishing an objective function, and optimizing a new task matrix;

(7) and (4) placing the robot at an initial position, and executing the task according to the task matrix optimized in the step (6).

2. The archival robot navigation positioning and mission planning method of claim 1, wherein the initial parameters of step (1) include: defining a width matrix and a width error of each file rack according to the width of the file rack; defining a position matrix of the file rack according to the distance between the initial position of the file rack and the coordinate axis; defining a map variable matrix according to the opening and closing state of the file rack; a distance matrix is defined according to the distance between adjacent file racks.

3. The method for navigation positioning and mission planning of an archival robot as recited in claim 1, wherein said step (2) comprises the steps of:

(20) initializing variables: the method comprises the following steps that m and n are counting variables of positive integers, the number of file shelves is 0, X1 and X2 are defined as the distances from a robot to a left wall and a right wall respectively, Y1 and Y2 are defined as the distances from the robot to a rear wall and a front wall respectively, and detection is carried out through sensors located on the periphery of the robot; the robot is located at the initial position, and the X1 value at the moment is recorded as X10;

(21) the robot starts from an initial position, updates X1 in real time, indicates that a left archive shelf is detected when X1 ≠ X10, records that the value of Y1 is Y10, PA (m) is Y2+ W1(m)/2, and simultaneously makes X10 ═ X1;

(22) the robot moves forward continuously, X1 is updated continuously, when X1 is not equal to X10, the robot is shown to exceed the left file rack, enters a working area between the left file racks, and the Y1 value at the moment is updated, so that Y11 is equal to Y10; the current Y1 value is then recorded again as Y10, i.e.: y10 ═ Y1, calculated Y11 ═ Y11-Y1;

(23) calculating and comparing according to the following formula, if the value of Y11-W1(m) | > (K1+5) × Δ W is Y11 ═ Y11-W1(m), m ═ m +1, PA (m) ═ PA (m-1) + [ W1(m-1) + W1(m) ]/2, and K1 ═ K1+1, continuing the circulation step (23), otherwise, indicating that the number of the left-side file frames is detected at this time, the m file frames are detected at present, making K1 ═ 0, and performing the step (24);

(24) from the matrix MA, counting m 1 from left to right, and detecting whether the next numerical value of the m 1 is 0; if not, because the left file rack is detected to be opened at the moment, inserting 0 to indicate the mark after the mth 1, and deleting the last 0 in the MA;

(25) the detection of the right file rack is consistent with the steps (22) to (24);

(26) the mobile robot intermittently moves forwards, when X1 is not equal to X10, the robot exceeds a working channel in the middle of the left file rack, the left file rack is detected again, the Y1 value at the moment is recorded, and the last Y10 is stored by Y11, namely Y11 is equal to Y10; the current Y1 value is recorded as Y10, i.e.: y10 ═ Y1: calculating the distance of the opening of the working area, and storing the distance into a matrix D1, wherein D1[ m ] (Y11-Y1);

when X2 is not equal to X20, the robot is indicated to exceed a working channel in the middle of the right file rack, and the right file rack is detected again; recording the Y1 value at this time, and storing the last Y20, i.e., Y21 — Y20, in Y21; the current Y1 value is recorded as Y20, i.e.: y20 ═ Y1: the distance the work area is open is calculated and stored in matrix D2, D2[ n ] ═ Y21-Y1.

4. The method for navigation positioning and mission planning of an archival robot as recited in claim 1, wherein said step (3) comprises:

(31) when M and N are simultaneously satisfied, indicating that the file region is detected, executing the step (4), otherwise, executing the step (31);

(32) and (6) judging whether Y1 is larger than a safety value, if so, returning to the step (2), otherwise, returning the robot to the initial position for standby.

5. The method for navigation positioning and mission planning of an archival robot as recited in claim 1, wherein said step (4) comprises:

(41) screening out executable tasks on the left side

When DA [ TA1(m) ] < L (m 1-SDA) indicates that the width between file shelves is not enough, the task cannot be executed, the task is deleted from TA1, and TN1(m) is deleted from TN1, so that SDA is SDA-1;

(42) sifting out right-hand executable tasks

When DB [ TB1(n) ] < L (n 1 to SDB) indicates that the width between the archive shelves is not sufficient, the task cannot be executed, the task is deleted from TB1, TN2(n) is deleted from TN2, and SDB is made SDB-1;

(43) forming an executable task matrix TE:

each row of the TE matrix corresponds to one task information:

wherein pre-SDA behavior:

TE[m，1]＝TA1(m)，TE[m，2]＝TN1(m)，

TE[m，3]＝PA{TA1(m)+W1[TA1(m)]/2+DA[TA1(m)]/2；

TE[m，4]＝0；

m＝1～SDA；

next there is the SDB line:

TE[SDA+n，1]＝TB1(n)，TE[SDA+n，2]＝TN2(n)，

TE[SDA+n，3]＝PA{TB1(n)+W1[TB1(n)]/2+DB[TA1(n)]/2；

TE[SDA+n，4]＝1；

n＝1～SDB；

each row of TE may be denoted as TE (p), and the elements of p rows and k columns in TE may be denoted as TE (p, k).

6. The method for navigation positioning and mission planning of an archival robot as recited in claim 1, wherein said step (5) comprises:

(51) TE (m, 3) represents the position of a working area capable of executing tasks, the rows of the TE matrix are reordered according to the value of the TE (m, 3), wherein the value is smaller before the value is larger after the value is smaller;

(52) defining the number of tasks which can be processed by the robot at one time as TM, defining the task quantity which exceeds TM as a large task, defining the task quantity which is equal to TM as a full task, and defining the task quantity which is less than TM as a small task;

(53) the large task is divided into a plurality of full tasks and a small task.

7. The method for navigation positioning and mission planning of an archival robot as recited in claim 1, wherein said step (6) comprises:

(61) taking several task combinations TECO (i) in the tasks of the task matrix, and the combination condition can be expressed as: summing column 2 of the TECO matrix, no more than TM-TE (p, 2); wherein TM is the number of tasks that the robot can process at a time, and TE (p, 2) is the element of p rows and 2 columns in TE;

when TE (P, 2) ═ TM, TECO is an empty matrix;

TE (P, 2) < TM, TECO is a set of matrix sequences of unequal number of rows, denoted as TECO (i), of several forms:

there is only one row, i.e. there is only one task in the combination:

TECO(1)＝[TE(x，1)，TE(x，2)，TE(x，3)，TE(x，4)]

there are multiple rows, i.e. there are only multiple tasks in a group:

TECO(1)＝[TE(x₁，1)，TE(x₁，2)，TE(x₁，3)，TE(x₁，4)TE(x₂，1)，TE(x₂，2)，TE(x₂，3)，TE(x₂，4)TE(x₃，1)，TE(x₃，2)，TE(x₃，3)，TE(x₃，4)TE(x₄，1)，TE(x₄，2)，TE(x₄，3)，TE(x₄，4)…………………]

in the formula, x₁、x₂、x₃、x₄The frame positions of the tasks are all positive integers, 1<x、x₁、x₂、x₃、x₄≤SDE；

The number of rows of each TECO (i) is DCO (i), and the TECO (i) represents the task of the ith row;

(62) calculating an objective function aim (i):

AIM(i)＝∑{[s+TE(p,3)-TECO(i)(j·3)×TECO(i)(j·2)]}×{∑[TECO(i)(j·2)]}/8

wherein j is 1-DCO (i), and S is a coefficient related to the spatial distribution density of the task; j is a control variable set for summation for selecting a row in the matrix teco (i);

(63) solving max [ AIM (i) ], max [ AIM (i) ]) as the optimal task combination, combining with TE (p) to be the planning result of the task, and expressing as follows by using a block matrix: TEOP [ te (p) yeco (i) ];

deleting the tasks in the selected combination TEOP from the executable task matrix TE, wherein the tasks left in the TE are unplanned tasks;

(64) and repeating the steps until all the matrixes in the TE matrix are deleted.

8. A method for archival navigation positioning and mission planning of a robot as recited in claim 1, wherein said step (7) is followed by step (8): and the robot returns to the initial position, reports the task state to the control center and stands by.

Images