CN104848856A - Transformer substation inspection robot track calculation method and device based on inter-wheel differential - Google Patents

Abstract

The invention discloses a transformer substation inspection robot track calculation method and device based on inter-wheel differential. A feedback link is added in the track calculation method, track data treated by virtue of Kalman filtering is fed back to a track calculation equation, the self-correction of the equation is realized, and the influence of interference on the track calculation can be effectively alleviated. The track calculation precision is relatively high in a local part, the circuit is simple, the parameter is simple to debug, and the algorithm is easy to realize. The transformer substation inspection robot track calculation method is combined with a magnetic track navigation technology, a magnetic track can be buried in a segmenting manner, the in-situ construction can be effectively reduced, and the cost is reduced; the transformer substation inspection robot track calculation method also can be combined with a GPS global navigation system, in a failure area of the GPS, the track information with high local precision can be outputted, and the weaknesses of the GPS global navigation system can be effectively compensated.

Description

A kind of Intelligent Mobile Robot reckoning method and apparatus based on differential between wheel

Technical field

The present invention is under the jurisdiction of localization for Mobile Robot and field of navigation technology, particularly relates to a kind of Intelligent Mobile Robot reckoning method and apparatus based on differential between wheel.

Background technology

Along with the application that Intelligent Mobile Robot is in recent years more deep, complicated substation inspection task to be run reliably, accurately in transformer station for crusing robot and is had higher requirement, robot body position and course angle information Perception are that robot realizes patrolling and examining the basis of a little accurately stopping, and the technology that current Intelligent Mobile Robot position and course angle information Perception adopt mainly contains following several:

1, magnetic orbital course guides the mode adding RFID label tag auxiliary positioning.Which is combined with the array of magnetic sensors that robot body carries by the magnetic orbital of pre-plugged in ground surface, the guiding of real-time course is carried out between advancing, control system by trying to achieve range ability to average wheel speed integration, then realizes robot location information perception by RFID label tag is auxiliary.Although which is stable, stop accurately, which needs complete station to carry out magnetic orbital to lay and intensively bury RFID assisted tag underground, and great in constructing amount, cost is high, maintenance inconvenience.

2, GPS adds the mode of digital compass.Which is carried out the acquisition of positional information in real time, carries out robot Course Acquisition by digital compass, thus realize the perception to robot body position and course angle information.Which effectively decreases the construction volume of substation field, improve the integration degree of system, but high-precision differential GPS GPS cost is high, serious field signal interference is there is in addition in transformer station, local differential GPS signal receives unstable, digital compass Course Acquisition is influenced, is unfavorable for local accurately location and Course Acquisition, stability in like fashion and less reliable.

3, odometer adds gyrostatic mode.Which carries out to average wheel speed the perception that integration realizes location information by control system, by realizing the calculating in course to the process of gyro data.Although which local positioning precision is high, accurately and reliably, which hardware circuit is complicated in positional information perception, and gyroscope is comparatively large by the impact of interference such as jolting, and the synchronism for data requires high.

So it is simple to need a kind of scheme badly, the reckoning method and apparatus that cost is low, carries out the reckoning that local accuracy of transformer station is higher, realizes robot location's information Perception.

Summary of the invention

The present invention is in order to solve the problem, and propose a kind of Intelligent Mobile Robot reckoning method and apparatus based on differential between wheel, the present invention effectively reduces site operation amount, reduces costs, and improves the precision of local tracks reckoning.

To achieve these goals, the present invention adopts following technical scheme:

A kind of Intelligent Mobile Robot dead reckoning tracer based on differential between wheel, comprise robot revolver tachogenerator signaling interface, robot is right takes turns tachogenerator signaling interface, signal conditioning circuit and controller, wherein, robot revolver tachogenerator signaling interface receives the tachogenerator signal of left side wheel, robot is right takes turns the tachogenerator signal that tachogenerator signaling interface receives right side wheels, robot revolver tachogenerator signaling interface, the sensor signal of collection is transferred to signal conditioning circuit by the right tachogenerator signaling interface of taking turns of robot, signal conditioning circuit by conditioning after Signal transmissions to controller.

Described controller is provided with communication interface, communicates with industrial computer or other control system, realizes data interaction and transmission in robot system.

Based on a reckoning method for said system, comprise the following steps:

(1) according to different running environment setting basic parameter matrix element, dynamic parameter matrix rule list is built;

(2) controller reads the revolver of t and right wheel speed, in conjunction with the index logic of dynamic parameter matrix rule list, builds reckoning model;

(3) controller is according to the flight path information of reckoning normatron device people, utilizes Kalman filter to carry out filtering to this flight path information, and controller, by homeward for filtered flight path information feed back mark prediction model, carries out reckoning.

In described step (1), concrete grammar is: structure dynamic parameter matrix rule list, and the basic parameter matrix element in this rule list is,

$\mathrm{Factor}\_m=\left[\begin{array}{c}\mathrm{para}11,\mathrm{paral}2\\ \mathrm{para}21,\mathrm{para}22\end{array}\right]---\left(1\right)$

Parameter p ara11 in above-mentioned matrix, para12 are the linear coefficients that left and right wheel speed is expert in journey reckoning, and para21, para22 are the linear coefficient of left and right wheel speed in course angle calculates.

By adjusting above-mentioned linear coefficient, mathematics prediction model more accurately can be set up according to conditions such as robotic's structure, machine error, test environments.Debug phase is for different running environment, the multiple basic parameter matrix element of different Design of Mechanical Structures, tested by ecotopia, selected standard one-way distance such as 10m carries out the adjustment of para11, para12 linear coefficient, selected standard crank degree such as 90 ° carries out para21, para22 linear coefficient, if physical construction is symmetrical, machine error is symmetrical, para11, para12 and para21, para22 are absolute value equivalent parameters pair.Same reason, can carry out corresponding parameter to adjustment according to different running environment types.By these parameters to being organized into parameter matrix element, and these basic parameter matrix elements being filled up in rule list according to certain types index order, in this rule list, at least having a life type basic parameter matrix element.

In described step (2), reckoning model equation is as follows:

$\left[\begin{array}{c}\mathrm{\Δd}\\ \mathrm{\Δw}\end{array}\right]=\mathrm{Factor}\_m*\left[\begin{array}{c}\mathrm{SpeedL}\\ \mathrm{SpeedR}\end{array}\right]$

W(t)＝W(t-1)+Δw (2)

X(t)+＝Δd*cos((W(t-1)+Δw)/2.0)

Y(t)+＝Δd*cos((W(t-1)+Δw)/2.0)

In above formula, SpeedL, SpeedR are the distance that sheet left and right wheels respectively passes by of fixing time, Δ d, Δ w are timeslice robot body motion Distance geometry angle, [X (t), Y (t), W (t)] be respectively t flight path X-coordinate, Y-coordinate, overall course angle.

The concrete grammar of described step (3) is:

By reckoning model equation by calculating t revolver and right wheel speed and the basic parameter matrix element that indexes out, export t robot flight path information [X (t), Y (t), W (t)]; By Kalman filter, filtering is carried out to the robot flight path information that reckoning model equation exports again, export the flight path information [X (t) ', Y (t) ', W (t) '] of the higher robot t of Reliability ratio; By the flight path information [X (t) ' of the robot t that Kalman filter exports, Y (t) ', W (t) '] upload to industrial computer or other control system, this information is fed back to reckoning model equation by feedback element simultaneously, by ranking operation to [X (t), Y (t), W (t)] carry out upgrading the formula of ranking operation and be:

[X(t),Y(t),W(t)]＝(1-r)*[X(t),Y(t),W(t)]+r*[X(t)',Y(t)',W(t)']

In above formula, r is weights coefficient, by rule of thumb setting in debug process, if test environment is more satisfactory, the trust exponent for reckoning model is high can be turned down r value, if test environment interference is comparatively large, then need r to be heightened some to reduce the impact of interference on system.

In described step (3), reckoning model data equation accepts the flight path information of external control system feedback such as in conjunction with absolute fix modes such as laser navigation technology, the output data that this feedback combination model carries feedback element feedback realize the correction of two feedbacks to reckoning model equation flight path information together, and two feedback weight updating formula is:

[X (t), Y (t), W (t)]=(1-o-r) * [X (t), Y (t), W (t)]+r* [X (t) ', Y (t) ', W (t) ']+o* [X (t) ", Y (t) ", W (t) "] r in above formula, o is weights coefficient, also be by rule of thumb setting in debug process, if test environment is more satisfactory, trust exponent for reckoning model is high can by r, o value is turned down, if test environment interference is larger, then need r to be heightened some to reduce the impact of environmental interference on system, if external control system feedback positioning precision is more reliable and stable, the absolute fix modes such as such as laser navigation technology, then o coefficient should be heightened some to increase the weights of external feedback in correction.

Beneficial effect of the present invention is:

(1) present invention achieves the parameter matrix element of dynamic conditioning reckoning model equation thus achieve the foundation of dynamic reckoning model, solve single parameter matrix to the unconformable problem of complicated running environment, ensure that the accuracy of robot reckoning model, dynamic reckoning model effectively raises adaptability;

(2) add feedback element in reckoning method of the present invention, by the track data after Kalman filtering process is fed back to reckoning equation, achieve the self-correcting of equation, effectively reduce the impact of interference on reckoning;

(3) the present invention has the higher reckoning precision in local, and circuit is simplified, and parameter testing is simple, and algorithm realization is easy;

(4) technology of magnetic orbital navigation of the present invention combines, and can carry out segmentation magnetic orbital and bury underground, effectively reduce site operation amount, reduce cost;

(5) the present invention can also combine with GPS Global Navigation System, and in the region that GPS lost efficacy, this invention can export the higher flight path information of local accuracy, effectively makes up the deficiency of GPS Global Navigation System.

Accompanying drawing explanation

Fig. 1 is structural representation of the present invention;

Fig. 2 is reckoning method step schematic diagram of the present invention.

Wherein, 1, dead reckoning tracer circuit board, 2, robot revolver tachogenerator signaling interface, 3, robot is right takes turns tachogenerator signaling interface, 4, signal conditioning circuit, 5, controller, 6, communication interface.

Embodiment:

Below in conjunction with accompanying drawing and embodiment, the invention will be further described.

As shown in Figure 1, based on the Intelligent Mobile Robot reckoning transposition of differential between wheel, it comprises dead reckoning tracer circuit board 1, there is robot revolver tachogenerator signaling interface 2 on the board, robot is right takes turns tachogenerator signaling interface 3, the output signal of tachogenerator signaling interface 2 and 3 is connected to signal conditioning circuit 4, and signal conditioning circuit 4 is connected with controller 5, and dead reckoning tracer is connected with industrial computer or other control system by communication interface 6.

Robot revolver tachogenerator signaling interface 2 can have multiple carry out homonymy take turns more tachogenerator signal access;

Robot is right takes turns tachogenerator signaling interface 3 can have and multiple carry out the signal access that homonymy takes turns tachogenerator more;

Robot signal conditioning circuit 4 nurses one's health revolver tachogenerator signaling interface 2 and right tachogenerator signaling interface 3 signal of taking turns, and the output of signal conditioning circuit is connected on controller 5 as the input of controller 5;

Controller 5 carries out reckoning according to the reckoning method step in accompanying drawing 2;

Communication interface 6 is interfaces that dead reckoning tracer communicates with industrial computer or other control system, mainly in robot system, realizes data interaction and controlling functions;

As shown in Figure 2, reckoning method step is:

Based on the Intelligent Mobile Robot reckoning method and apparatus of differential between wheel, need at least to combine there is kinetic control system Intelligent Mobile Robot on use.

[1] first design dynamic parameter matrix rule list, the basic parameter matrix element in this rule list is,

$\mathrm{Factor}\_m=\left[\begin{array}{c}\mathrm{para}11,\mathrm{paral}2\\ \mathrm{para}21,\mathrm{para}22\end{array}\right]---\left(1\right)$

Parameter p ara11 in above-mentioned matrix, para12 are the linear coefficients that left and right wheel speed is expert in journey reckoning.Parameter p ara21 in above-mentioned matrix, para22 are the linear coefficient of left and right wheel speed in course angle calculates., by adjusting above-mentioned linear coefficient, can set up according to conditions such as robotic's structure, machine error, test environments and following mathematics prediction model accurately.Debug phase is for different running environment, the multiple basic parameter matrix element of different Design of Mechanical Structures, tested by ecotopia, selected standard one-way distance such as 10m carries out the adjustment of para11, para12 linear coefficient, selected standard crank degree such as 90 ° carries out para21, para22 linear coefficient, if physical construction is symmetrical, machine error is symmetrical, para11, para12 and para21, para22 are absolute value equivalent parameters pair.Same reason, can carry out corresponding parameter to adjustment according to different running environment types.By these parameters to being organized into parameter matrix element, and these basic parameter matrix elements being filled up in rule list according to certain types index order, in this rule list, at least having a life type basic parameter matrix element.

[2] controller reads the revolver of the t that the unit inspection that tests the speed arrives and right wheel speed.

[3] design the types index logic for dynamic parameter matrix rule list in the controller, or slave computer or other control system control the basic parameter matrix element of controller index respective type by communication interface 6.

[4] by t revolver and right wheel speed, and reckoning model equation is input to from the basic parameter matrix element that dynamic parameter matrix rule list indexes out.

[5] reckoning model equation is as follows:

$\left[\begin{array}{c}\mathrm{\Δd}\\ \mathrm{\Δw}\end{array}\right]=\mathrm{Factor}\_m*\left[\begin{array}{c}\mathrm{SpeedL}\\ \mathrm{SpeedR}\end{array}\right]$

W(t)＝W(t-1)+Δw (2)

X(t)+＝Δd*cos((W(t-1)+Δw)/2.0)

Y(t)+＝Δd*cos((W(t-1)+Δw)/2.0)

[6] by reckoning model equation by calculating t revolver and right wheel speed and the basic parameter matrix element that indexes out, export t robot flight path information [X (t), Y (t), W (t)].

[7] by Kalman filter, filtering is carried out to the robot flight path information that reckoning model equation exports again, export the flight path information [X (t) ' of the higher robot t of Reliability ratio, Y (t) ', W (t) '].

Flight path information [the X (t) ' of the robot t [8] Kalman filter exported, Y (t) ', W (t) '] upload to industrial computer or other control system, this information is fed back to reckoning model equation by feedback element simultaneously, again by ranking operation to [X (t), Y (t), W (t)] upgrade, the formula of ranking operation is:

[X(t),Y(t),W(t)]＝(1-r)*[X(t),Y(t),W(t)]+r*[X(t)',Y(t)',W(t)']

In above formula, r is weights coefficient, by rule of thumb setting in debug process, if test environment is more satisfactory, the trust exponent for reckoning model is high can be turned down r value, if test environment interference is comparatively large, then need r to be heightened some to reduce the impact of interference on system.

[9] reckoning model data equation accept external control system feedback flight path information [X (t) "; Y (t) ", W (t) "]; such as in conjunction with absolute fix modes such as laser navigation technology; the output data that this feedback integrating step [8] model carries feedback element feedback realize the correction of two feedbacks to reckoning model equation flight path information together, and two feedback weight updating formula is:

[X (t), Y (t), W (t)]=(1-o-r) * [X (t), Y (t), W (t)]+r* [X (t) ', Y (t) ', W (t) ']+o* [X (t) ", Y (t) ", W (t) "] r in above formula, o is weights coefficient, also be by rule of thumb setting in debug process, if test environment is more satisfactory, trust exponent for reckoning model is high can by r, o value is turned down, if test environment interference is larger, then need r to be heightened some to reduce the impact of environmental interference on system, if external control system feedback positioning precision follows the reliable and stable of family, the absolute fix modes such as such as laser navigation technology, then o coefficient should be heightened some to increase the weights of external feedback in correction.

[10] repeat above-mentioned steps [2] ~ [8], this dead reckoning tracer is by flight path information higher for continuously output accuracy.

By reference to the accompanying drawings the specific embodiment of the present invention is described although above-mentioned; but not limiting the scope of the invention; one of ordinary skill in the art should be understood that; on the basis of technical scheme of the present invention, those skilled in the art do not need to pay various amendment or distortion that creative work can make still within protection scope of the present invention.

Claims (8)

1. the Intelligent Mobile Robot dead reckoning tracer based on differential between wheel, it is characterized in that: comprise robot revolver tachogenerator signaling interface, robot is right takes turns tachogenerator signaling interface, signal conditioning circuit and controller, wherein, robot revolver tachogenerator signaling interface receives the tachogenerator signal of left side wheel, robot is right takes turns the tachogenerator signal that tachogenerator signaling interface receives right side wheels, robot revolver tachogenerator signaling interface, the sensor signal of collection is transferred to signal conditioning circuit by the right tachogenerator signaling interface of taking turns of robot, signal conditioning circuit by conditioning after Signal transmissions to controller.

2. a kind of based on the Intelligent Mobile Robot dead reckoning tracer of differential between wheel as claimed in claim 1, it is characterized in that: described controller is provided with communication interface, communicate with industrial computer or other control system, in robot system, realize data interaction and transmission.

3., based on a reckoning method for system according to claim 1, it is characterized in that: comprise the following steps:

(1) according to different running environment setting basic parameter matrix element, dynamic parameter matrix rule list is built;

(2) controller reads the revolver of t and right wheel speed, in conjunction with the index logic of dynamic parameter matrix rule list, builds reckoning model;

(3) controller is according to the flight path information of reckoning normatron device people, utilizes Kalman filter to carry out filtering to this flight path information, and controller, by homeward for filtered flight path information feed back mark prediction model, carries out reckoning.

4. reckoning method as claimed in claim 3, is characterized in that: in described step (1), concrete grammar is: structure dynamic parameter matrix rule list, and the basic parameter matrix element in this rule list is,

$\mathrm{Factor}\_m=\left[\begin{array}{c}\mathrm{para}11,\mathrm{para}12\\ \mathrm{para}21,\mathrm{para}22\end{array}\right]---\left(1\right)$

Parameter p ara11 in above-mentioned matrix, para12 are the linear coefficients that left and right wheel speed is expert in journey reckoning, and para21, para22 are the linear coefficient of left and right wheel speed in course angle calculates.

5. reckoning method as claimed in claim 3, it is characterized in that: in described step (2), by the linear coefficient of the basic parameter matrix in regulation rule table, then set up mathematics prediction model in conjunction with the condition of robotic's structure, machine error and test environment.

6. reckoning method as claimed in claim 3, is characterized in that: in described step (2), reckoning model equation is as follows:

$\left[\begin{array}{c}\mathrm{\Δd}\\ \mathrm{\Δw}\end{array}\right]=\mathrm{Factor}\_m*\left[\begin{array}{c}\mathrm{SpeedL}\\ \mathrm{SpeedR}\end{array}\right]$

W(t)＝W(t-1)+Δw (2)

X(t)+＝Δd*cos((W(t-1)+Δw)/2.0)

Y(t)+＝Δd*cos((W(t-1)+Δw)/2.0)

In above formula, SpeedL, SpeedR are the distance that sheet left and right wheels respectively passes by of fixing time, Δ d, Δ w are timeslice robot body motion Distance geometry angle, [X (t), Y (t), W (t)] be respectively t flight path X-coordinate, Y-coordinate, overall course angle.

7. reckoning method as claimed in claim 3, is characterized in that: the concrete grammar of described step (3) is:

By reckoning model equation by calculating t revolver and right wheel speed and the basic parameter matrix element that indexes out, export t robot flight path information [X (t), Y (t), W (t)]; By Kalman filter, filtering is carried out to the robot flight path information that reckoning model equation exports again, the flight path information [X (t) ', Y (t) ', W (t) '] of output device people t; By the flight path information [X (t) ' of the robot t that Kalman filter exports, Y (t) ', W (t) '] upload to industrial computer or other control system, this information is fed back to reckoning model equation by feedback element simultaneously, by ranking operation to [X (t), Y (t), W (t)] carry out upgrading the formula of ranking operation and be:

[X(t),Y(t),W(t)]＝(1-r)*[X(t),Y(t),W(t)]+r*[X(t)',Y(t)',W(t)']

In above formula, r is weights coefficient, by rule of thumb setting in debug process, if test environment is more satisfactory, the trust exponent for reckoning model is high can be turned down r value, if test environment interference is comparatively large, then need r to be heightened some to reduce the impact of interference on system.

8. reckoning method as claimed in claim 3, it is characterized in that: in described step (3), reckoning model data equation accept external control system feedback flight path information [X (t) "; Y (t) ", W (t) "]; utilize absolute fix mode, the output data that this feedback combination model carries feedback element feedback realize the correction of two feedbacks to reckoning model equation flight path information together, and two feedback weight updating formula is:

[X (t), Y (t), W (t)]=(1-o-r) * [X (t), Y (t), W (t)]+r* [X (t) ', Y (t) ', W (t) ']+o* [X (t) "; Y (t) ", W (t) "] r, o are weights coefficient in above formula, by rule of thumb setting in debug process.