CN108742840A - The scaling method and device of robot - Google Patents
The scaling method and device of robot Download PDFInfo
- Publication number
- CN108742840A CN108742840A CN201810317759.0A CN201810317759A CN108742840A CN 108742840 A CN108742840 A CN 108742840A CN 201810317759 A CN201810317759 A CN 201810317759A CN 108742840 A CN108742840 A CN 108742840A
- Authority
- CN
- China
- Prior art keywords
- parameter
- calibrated
- amount
- measured value
- particle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
- A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
- A61B34/30—Surgical robots
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
- A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
- A61B2034/2046—Tracking techniques
- A61B2034/2059—Mechanical position encoders
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
- A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
- A61B2034/2046—Tracking techniques
- A61B2034/2063—Acoustic tracking systems, e.g. using ultrasound
Abstract
The present invention provides a kind of scaling method and device of robot, and wherein scaling method includes：Determine the standard value with each relevant parameter of amount to be calibrated and each parameter, the measured value of the measured value and parameter of each amount to be calibrated when by predeterminated frequency acquisition robot operation；For any one parameter, the measured value of the parameter is demarcated according to the standard value of multidimensional particle cluster algorithm and the parameter；For any one amount to be calibrated, know other relevant amounts to be calibrated and each parameter according to constraint equation, according to after the relevant each parameter calibration of amount to be calibrated measured value and other amounts to be calibrated described in calibration are completed, obtain the calculated value of the amount to be calibrated, according to multidimensional particle cluster algorithm and the measured value of the amount to be calibrated, the calculated value of the amount to be calibrated is demarcated.The embodiment of the present invention has the advantageous effect that fast convergence rate, local extremum incidence is low, can effectively solve the problem that the calibration of multi-dimensional complicated Relation Parameters.
Description
Technical field
The present invention relates to robotic technology fields, more particularly, to the scaling method and device of robot.
Background technology
In deep vein puncture operation, exist operation demand is big, and doctor's first time puncture success rate is low, complication mostly with
And medical staff's training time it is long the problems such as.In recent years, many doctors take ultrasound guidance in order to solve these problems
Deep vein puncture, this piercing method can greatly improve one-time successful puncture rate, and it is concurrent to reduce operation to a certain extent
Disease.But still remain following problem：The first, the position after puncture needle enters in vivo can not be predicted before puncturing operation
It sets, so the accuracy rate and success rate that puncture rely heavily on the clinical experience of doctor, still needs to performing the operation
Doctor is trained for a long time.The second, puncture needle can be because of puncture angle, illumination reflection and puncture needle sometimes sometimes
It is non-coplanar with popping one's head in, often occur the case where can not seeing puncture needle in piercing process.In fact, because operation custom and
These problems existing for ultrasonic puncture, have substantial portion of doctor still to use the blind method worn in deep vein puncture.
To solve the above-mentioned problems, scientific research personnel develops the deep vein puncture Algorithms of Robots Navigation System of ultrasound guidance.
The precision of the navigation system dependent on entire robot system multiple mechanical parameters and control parameter precision, mechanical parameter and
Control parameter can also be referred to as parameter.
Wherein, the precision of mechanical parameter is subject to processing and assembles both sides and influences, and control parameter precision is by motor essence
The influence of degree, zero-bit proximity sensor equally accurate.If demarcated according to each parameter, high-precision measuring device pair is needed
Its precision is tested.Meanwhile since it is desired that the quantity of the parameter of calibration is relatively more, the workload of inspection is very big.Has skill
There is no the calibration for Ultrasound-guided Biopsy robotic and control parameter in art, is only directed to the movement of all-purpose robot
Scaling method is learned, this method is suitable for improving the running orbit precision of robot and general use high precision measuring instrument
The method that each mechanical parameter and control parameter are demarcated.These methods are all not applied for Needle-driven Robot navigation system
System, cannot automatic Calibration guiding puncture line in a short time, and the cost of the measuring instrument of the higher precision for detection is high
It is expensive.
Invention content
The present invention provides a kind of calibration side for the robot for overcoming the above problem or solving the above problems at least partly
Method and device.
According to an aspect of the present invention, a kind of scaling method of robot is provided, including：
It determines the standard value with each relevant parameter of amount to be calibrated and each parameter, machine is acquired by predeterminated frequency
The measured value of the measured value and parameter of each amount to be calibrated when people runs；
For any one parameter, according to the standard value of multidimensional particle cluster algorithm and the parameter to the measured value of the parameter
It is demarcated；
For any one amount to be calibrated, other relevant amounts to be calibrated and each parameter are known according to constraint equation, according to
With after the relevant each parameter calibration of amount to be calibrated measured value and other amounts to be calibrated described in calibration are completed, be somebody's turn to do
The calculated value of amount to be calibrated, according to multidimensional particle cluster algorithm and the measured value of the amount to be calibrated, the calculating to the amount to be calibrated
Value is demarcated.
Preferably, described for any one parameter, according to the standard value of multidimensional particle cluster algorithm and the parameter to this
The step of measured value of parameter is demarcated, including：
Using times of collection as dimension, established according to deviation of the parameter under all dimensions between measured value and standard value
Fitness function；
Using deviation of the parameter under each dimension between measured value and standard value as a particle, population is established,
Initialize the position and speed of each particle；
The iterative search that particle is carried out with multidimensional particle cluster algorithm, when iterative search to the fitness function minimum,
Obtain optimal solution；
The measured value at the parameter each moment is demarcated according to the optimal solution.
Preferably, it is described with multidimensional particle cluster algorithm carry out particle iterative search the step of, further include before：
According to preset deviation threshold, the larger particle of deviation between measured value and standard value is picked from the population
It removes.
Preferably, described according to multidimensional particle cluster algorithm and the measured value of the amount to be calibrated, to the meter of the amount to be calibrated
The step of calculation value is demarcated, including：
Any one amount to be calibrated is surveyed according to the amount to be calibrated under all dimensions using times of collection as dimension
Deviation between magnitude and calculated value establishes fitness function；
Using the deviation of the amount to be calibrated under each dimension between measured value and calculated value as a particle, particle is established
Group, initializes the position and speed of each particle；
The iterative search that particle is carried out with multidimensional particle cluster algorithm, when iterative search to the fitness function minimum,
Obtain optimal solution；
The calculated value at amount each moment to be calibrated is demarcated according to the optimal solution.
Preferably, the acceleration constant used in the multidimensional particle cluster algorithm is according to the total number of the parameter, acquisition
The number of number and the constraint equation obtains；
The inertia weight used in the multidimensional particle cluster algorithm is obtained according to the times of collection.
Other side according to the ... of the embodiment of the present invention also provides a kind of caliberating device of robot, including：
Acquisition module, for determining the standard value with each relevant parameter of amount to be calibrated and each parameter, by pre-
The measured value of the measured value and parameter of each amount to be calibrated when if frequency collection robot is run；
One layer of demarcating module is used for for any one parameter, according to multidimensional particle cluster algorithm and the standard of the parameter
Value demarcates the measured value of the parameter；
Two layers of demarcating module, for for any one amount to be calibrated, knowing that relevant other wait marking according to constraint equation
Quantitative and each parameter, according to after the relevant each parameter calibration of amount to be calibrated measured value and it is completed described in calibration
His amount to be calibrated, obtains the calculated value of the amount to be calibrated, right according to multidimensional particle cluster algorithm and the measured value of the amount to be calibrated
The calculated value of the amount to be calibrated is demarcated.
Preferably, one layer of demarcating module includes：
First fitness function establishes unit, is used for any one parameter, using times of collection as dimension, according to the ginseng
Deviation of the number under all dimensions between measured value and standard value establishes fitness function；
First population establishes unit, for being made with deviation of the parameter under each dimension between measured value and standard value
For a particle, population is established, initializes the position and speed of each particle；
First computing unit, the iterative search for carrying out particle with multidimensional particle cluster algorithm, when iterative search is to described
When fitness function minimum, optimal solution is obtained；
First calibration unit, for being demarcated to the measured value at the parameter each moment according to the optimal solution.
Preferably, first computing unit is additionally operable to：
Before the iterative search for carrying out particle with multidimensional particle cluster algorithm, according to preset deviation threshold, by measured value
The larger particle of deviation is rejected from the population between standard value.
Preferably, two layers of demarcating module includes：
Second fitness function establishes unit, is used for for any one amount to be calibrated, using times of collection as dimension, root
Fitness function is established according to the deviation of the amount to be calibrated under all dimensions between measured value and calculated value；
Second population establishes unit, for inclined between measured value and calculated value under each dimension with the amount to be calibrated
Difference is used as a particle, establishes population, initializes the position and speed of each particle；
Second computing unit, the iterative search for carrying out particle with multidimensional particle cluster algorithm, when iterative search is to described
When fitness function minimum, optimal solution is obtained；
Second calibration unit, for being demarcated to the calculated value at amount each moment to be calibrated according to the optimal solution.
Preferably, one layer of demarcating module and two layers of demarcating module execute the acceleration used when multidimensional particle cluster algorithm
Constant is obtained according to the number of the total number of the parameter, times of collection and the constraint equation；
One layer of demarcating module and two layers of demarcating module execute the inertia weight that is used when multidimensional particle cluster algorithm according to
The times of collection obtains.
The scaling method and device of robot proposed by the present invention, wherein scaling method specifically include：It determines and respectively waits marking
The standard value of quantitative relevant parameter and each parameter, the survey of each amount to be calibrated when by predeterminated frequency acquisition robot operation
The measured value of magnitude and parameter.For any one parameter, according to the standard value of multidimensional particle cluster algorithm and the parameter to this
The measured value of parameter is demarcated.For any one amount to be calibrated, other relevant amounts to be calibrated are known according to constraint equation
With each parameter, according to after the relevant each parameter calibration of amount to be calibrated measured value and other be completed described in calibration wait for
Scalar quantity obtains the calculated value of the amount to be calibrated, according to multidimensional particle cluster algorithm and the measured value of the amount to be calibrated, waits for this
Gauged calculated value is demarcated.It is machinery and control parameter that the parameter demarcated is needed in the embodiment of the present invention, is not fortune
It is dynamic to learn parameter, and the quantity of parameter and exists far more than the number of kinematics parameters between parameter in the embodiment of the present invention
Multi-dimensional complicated relationship, therefore the embodiment of the present invention is used and is first demarcated to all parameters, then treated based on calibrated parameter
The nesting type multidimensional particle cluster algorithm that scalar quantity is demarcated demarcates machinery and control parameter.The machine of the embodiment of the present invention
Device people's scaling method is capable of the scalar quantity of multiple and different error conditions of the multiple types of Fast Calibration, has fast convergence rate, office
Portion's extreme value incidence is low, can effectively solve the problem that the advantageous effect of multi-dimensional complicated Relation Parameters calibration.
Description of the drawings
Fig. 1 is the flow diagram according to the scaling method of the robot of the embodiment of the present invention；
Fig. 2 is to use the flow diagram that multidimensional particle cluster algorithm is demarcated to parameter according to the embodiment of the present invention；
Fig. 3 is to be illustrated using the flow that multidimensional particle cluster algorithm is demarcated according to the scalar quantity for the treatment of of the embodiment of the present invention
Figure；
Fig. 4 is the functional block diagram according to the calibration system of the robot of the embodiment of the present invention；
Fig. 5 is the functional block diagram according to the caliberating device of the robot of the embodiment of the present invention.
Specific implementation mode
With reference to the accompanying drawings and examples, the specific implementation mode of the present invention is described in further detail.Implement below
Example is not limited to the scope of the present invention for illustrating the present invention.
Fig. 1 shows the flow diagram of the scaling method of the robot of the embodiment of the present invention, the machine of the embodiment of the present invention
Device people scaling method first calculates all parameters by determining and each relevant parameter of amount to be calibrated using multidimensional particle cluster algorithm
The optimal compensation value, achievees the purpose that Fast Convergent, then by multidimensional particle cluster algorithm and the optimal compensation value of all parameters, according to
It is secondary that the optimal compensation value is sought to each amount to be calibrated, realize the calibration of all amounts to be calibrated.As shown, the scaling method is specific
Including：
S101, determination and each relevant parameter of amount to be calibrated and the standard value of each parameter, by predeterminated frequency harvester
The measured value of the measured value and parameter of each amount to be calibrated when device people runs.
It should be noted that amount crucial when amount to be calibrated is robot motion, for example, joint of robot angle data,
Sitting posture data etc., the embodiment of the present invention do not limit to specific parameter and number in amount to be calibrated.General a variety of amounts to be calibrated
Between also can exist influence each other.For each amount to be calibrated, there are some to influence the amount to be calibrated for robot body
Parameter (including mechanical parameter and control parameter), and the case where overlapping can also be had by influencing the parameter of each amount to be calibrated.This hair
Bright embodiment can determine the standard value of these parameters first, for example, hand height is lifted as an amount to be calibrated by robot, lift hand is high
Degree can be related to the length of some connecting rod, and this length is just used as a parameter, and more specifically, the length of connecting rod is as one
Kind mechanical parameter there is standard value, for example, 10cm in making machine people, but when robot starts to lift manual make, by
Small variation can occur for the length of the interaction between different component, connecting rod, add measurement accuracy the problem of so that
Each measured value all may be different, therefore the embodiment of the present invention can press preset frequency to this parameter in this course
(length of connecting rod) measures, and records measured value, and hand height is lifted as an amount to be calibrated by robot, also needs with pre-
If frequency is acquired, the measured value as the amount to be calibrated.
S102, for any one parameter, according to the standard value of multidimensional particle cluster algorithm and the parameter to the parameter
Measured value is demarcated.
It should be noted that it is machinery and control parameter that the parameter demarcated is needed in the embodiment of the present invention, it is not movement
Parameter is learned, the number of kinematics parameters is generally 6, and the quantity of parameter is far more than kinematics parameters in the embodiment of the present invention
Number, and there are multi-dimensional complicated relationship between parameter, therefore the embodiment of the present invention overcomes the calibration process needs of parameter
The multidimensional multi-target parameter optimizing problem of complex relationship can not demarcate multi-parameter using one-dimensional particle cluster algorithm, therefore
The embodiment of the present invention demarcates parameter using multidimensional particle cluster algorithm.
In multidimensional particle cluster algorithm, the potential solution of each optimization problem can be imagined as d dimension spaces search space
Point on as soon as, this point are referred to as particle (particle), and all particles are all suitable there are one being determined by object function
It should be worth (fitness value), there are one speed to determine the direction circled in the air and distance for each particle, and then particles are just followed
Current optimal particle is in Searching Resolution Space.In embodiments of the present invention, using times of collection as dimension, for each ginseng
It counts, there are one deviations for the measured value and standard value under each dimension, and therefore, the purpose using multidimensional particle cluster algorithm is exactly to obtain
Obtain an offset for making above-mentioned deviation minimum of the particle under all dimensions, such as survey of some parameter under 3 dimensions
Magnitude is respectively 9.99,9.99 and 10.01, and the standard value of the parameter is 10.00, finds out the measured value and standard value of every dimension
Difference be -0.01, -0.01 and 0.01, can be in the hope of being measured value of the parameter under 3 dimensions by multidimensional particle cluster algorithm
And the offset of actual value deviation minimum is 0.01, by the way that the offset compensates measured value, the survey after being compensated
Magnitude is 10.00,10.00 and 10.02, although still there is the measured value after 1 supplement to be not equal to actual value, most of parameter is
Optimized to equal than actual value, it is believed that optimization is completed.
S103, for any one amount to be calibrated, other relevant amounts to be calibrated and each parameter are known according to constraint equation,
According to after the relevant each parameter calibration of amount to be calibrated measured value and other amounts to be calibrated of calibration are completed, be somebody's turn to do
The calculated value of amount to be calibrated, according to multidimensional particle cluster algorithm and the measured value of the amount to be calibrated, the calculating to the amount to be calibrated
Value is demarcated.
For the angle of structure, for any one amount to be calibrated, the amount to be calibrated and a certain number of parameters are very
To also related to other amounts to be calibrated, therefore, pass through integrated structure and kinematics analysis, so that it may to obtain robot architecture's
Constraint equation, analytic solutions of the amount to be calibrated as constraint equation.In general, the number of the number of constraint equation and amount to be calibrated
Unanimously, and there are amount to be calibrated the solution based on another amount to be calibrated is needed the case where calculating, therefore, to implement in the present invention
When treating scalar quantity progress multidimensional particle cluster algorithm calibration in example, first to only being optimized with the relevant amount to be calibrated of parameter,
It after amount to be calibrated optimization (the demarcating), then pair is optimized with other relevant amounts to be calibrated of the amount to be calibrated, up to institute
Need scalar quantity calibration to be completed.
The robot calibration method of the embodiment of the present invention is capable of multiple and different error conditions of the multiple types of Fast Calibration
Scalar quantity, with fast convergence rate, local extremum incidence is low, can effectively solve the problem that the beneficial of multi-dimensional complicated Relation Parameters calibration
Effect.
On the basis of the various embodiments described above, for any one amount to be calibrated, according to constraint equation know it is relevant its
The step of his amount to be calibrated and each parameter, further include before：
According to the Machine Design of robot, the constraint equation for indicating restriction relation between each amount to be calibrated and parameter is established；
The number of constraint equation is consistent with the number of amount to be calibrated.
On the basis of the above embodiments, for any one parameter, according to multidimensional particle cluster algorithm and the parameter
The step of standard value demarcates the measured value of the parameter, referring to Fig. 2, as shown, the step includes：
S201, using times of collection as dimension, it is inclined between measured value and standard value under all dimensions according to the parameter
Difference establishes fitness function；
For example, fitness function fd(si)=Cd(si)-Sd(si), wherein siIndicate that i-th of parameter, d indicate dimension, Cd
(si) indicate parameter s under d dimensionsiMeasured value, Sd(si) indicate parameter s under d dimensionsiStandard value.It should be noted that measuring
Deviation between value and standard value can indicate incessantly with simple difference, can also use square, square of difference difference,
Absolute value of difference etc. form indicates that the embodiment of the present invention does not limit this specifically.
S202, using deviation of the parameter under each dimension between measured value and standard value as a particle, establish grain
Subgroup initializes the position and speed of each particle；
S203, the iterative search that particle is carried out with multidimensional particle cluster algorithm, when iterative search to fitness function minimum,
Obtain optimal solution；
The specific an iteration process of multidimensional particle cluster algorithm is described below：
The update of speed and position is carried out to each particle with multidimensional particle cluster algorithm, wherein particle rapidity formula is：
The location formula of particle is：
Wherein, Lvx, LxxFor Restriction Operators, it is specified that the threshold speed and position threshold of algorithm,For particle a
In dimension xda(t) the jth item of velocity vector on,For the velocity vector of following iteration number,It is particle a in dimension xda(t) the jth item of position vector on,For following iteration number
Position vector,It is particle a in dimension xda(t) the jth item of individual optimum position vector on,For the jth item of the population overall situation optimum position vector on dimension d, LvxIt is rate limitation range, LxxIt is
Position limits range.c1, c2For acceleration constant, w (t) is inertia weight, r1,j(t), r2,j(t) it is the random number of (0,1).
The corresponding fitness function of t+1 moment each particle is calculated, itself optimum position of a particle is obtained, and is obtained
The global optimum position of population；
To each particle, compare the corresponding fitness function in t moment itself optimum position and itself optimum position of t+1 moment
The fitness function of itself optimum position of t+1 moment is assigned to t moment itself most by corresponding fitness function if the former is big
The fitness function of best placement.
Compare the adaptation corresponding with the moment overall situation optimum positions t+1 of the corresponding fitness function in t moment overall situation optimum position
Function is spent, if the former is big, the fitness function of t+1 moment overall situations optimum position is assigned to the suitable of t moment overall situation optimum position
Response function.
Judge whether to reach iteration deployment, if reaching iterative steps, obtains all particles of current iteration itself optimum bits
It sets and otherwise global optimum position carries out next iteration using global optimum position as final prioritization scheme.
S204, the measured value at the parameter each moment is demarcated according to optimal solution.
Specifically, such as measured value of some parameter 3 moment is respectively 5.51,5.52 and 5.53, and optimal solution is
0.01, then calibrated measured value is 5.52,5.53 and 5.54.It should be noted that the measured value to parameter is demarcated
It does not need to be ranked up parameter, i.e., need not be demarcated in a fixed order.
It should be noted that the measured value of some parameters can have larger error, and this error can greatly influence
The convergence of multidimensional particle cluster algorithm, therefore before proceeding by multidimensional particle cluster algorithm, need to screen particle, lead to
It crosses and screens out the error beyond error range.Therefore, on the basis of the above embodiments, particle is carried out with multidimensional particle cluster algorithm
Iterative search the step of, further include before：It is according to preset deviation threshold, deviation between measured value and standard value is larger
Particle is rejected from population.
On the basis of the various embodiments described above, after being demarcated to all parameters, so that it may to utilize calibrated parameter
Each amount to be calibrated is demarcated, it should be noted that constraint equation embody derive each magnitude relation to be calibrated to lower and
On process, therefore when actually demarcating each amount to be calibrated, this sequence from bottom to top can be also followed, specifically, when depositing
In 3 amount a, b and c to be calibrated, the number of constraint equation is 3, it is assumed that amount a to be calibrated is calculated by parameter s1, s2 and s3
Go out, amount b to be calibrated is calculated by parameter s2, s4, s5 and amount a to be calibrated, and amount c to be calibrated is calculated by parameter s6, s7, a and b
Go out, then the embodiment of the present invention first demarcates a by multidimensional particle cluster algorithm, then b is demarcated again, finally c is marked
It is fixed.
On the basis of the above embodiments, according to multidimensional particle cluster algorithm and the measured value of the amount to be calibrated, this is waited for
The step of gauged calculated value is demarcated, referring to Fig. 3, including：
S301, for any one amount to be calibrated, using times of collection as dimension, according to the amount to be calibrated in all dimensions
Lower deviation between measured value and calculated value establishes fitness function；
For example, fitness function fd(Ri)=Cd(Ri)-Sd(Ri), wherein RiIndicate that i-th of amount to be calibrated, d indicate dimension
Degree, Cd(Ri) indicate parameter R under d dimensionsiMeasured value, Sd(Ri) indicate parameter R under d dimensionsiCalculated value.It needs to illustrate
It is that the deviation between measured value and calculated value can be indicated incessantly with simple difference, square, square of difference can also be used
Difference, absolute value etc. the form of difference indicate that the embodiment of the present invention does not limit this specifically.
S302, using the deviation of the amount to be calibrated under each dimension between measured value and calculated value as a particle, build
Vertical population, initializes the position and speed of each particle；
S303, the iterative search that particle is carried out with multidimensional particle cluster algorithm, when iterative search to fitness function minimum,
Obtain optimal solution；
The specific an iteration process of multidimensional particle cluster algorithm is described below：
The update of speed and position is carried out to each particle with multidimensional particle cluster algorithm, wherein particle rapidity formula is：
The location formula of particle is：
Wherein, Lvx, LxxFor Restriction Operators, it is specified that the threshold speed and position threshold of algorithm,For particle a
In dimension xda(t) the jth item of velocity vector on,For the velocity vector of following iteration number,It is particle a in dimension xda(t) the jth item of position vector on,For following iteration number
Position vector,It is particle a in dimension xda(t) the jth item of individual optimum position vector on,For the jth item of the population overall situation optimum position vector on dimension d, LvxIt is rate limitation range, LxxIt is
Position limits range.c1, c2For acceleration constant, w (t) is inertia weight, r1,j(t), r2,j(t) it is the random number of (0,1).
The corresponding fitness function of t+1 moment each particle is calculated, itself optimum position of a particle is obtained, and is obtained
The global optimum position of population；
To each particle, compare the corresponding fitness function in t moment itself optimum position and itself optimum position of t+1 moment
The fitness function of itself optimum position of t+1 moment is assigned to t moment itself most by corresponding fitness function if the former is big
The fitness function of best placement.
Compare the adaptation corresponding with the moment overall situation optimum positions t+1 of the corresponding fitness function in t moment overall situation optimum position
Function is spent, if the former is big, the fitness function of t+1 moment overall situations optimum position is assigned to the suitable of t moment overall situation optimum position
Response function.
Judge whether to reach iteration deployment, if reaching iterative steps, obtains all particles of current iteration itself optimum bits
It sets and otherwise global optimum position carries out next iteration using global optimum position as final prioritization scheme.
S304, the calculated value at amount each moment to be calibrated is demarcated according to optimal solution.
It should be noted that the measured value of some parameters can have larger error, and this error can greatly influence
The convergence of multidimensional particle cluster algorithm, therefore before proceeding by multidimensional particle cluster algorithm, need to screen particle, lead to
It crosses and screens out the error beyond error range.Therefore, on the basis of the above embodiments, particle is carried out with multidimensional particle cluster algorithm
Iterative search the step of, further include before：It is according to preset deviation threshold, deviation between measured value and standard value is larger
Particle is rejected from population.
On the basis of the various embodiments described above, the acceleration constant that is used in the multidimensional particle cluster algorithm of the embodiment of the present invention
It is obtained according to the number of the total number of parameter, times of collection and constraint equation；The inertia power used in multidimensional particle cluster algorithm
Repeated root is obtained according to times of collection.
Wherein, c1, c2For acceleration constant, w (t) is inertia weight, and n indicates the number of constraint equation, j expression parameters
Total number, d indicate dimension.
It should be noted that in the prior art the acceleration constant of multidimensional particle cluster algorithm and inertia weight not with dimension
Correlation, this, which there is, restrains slower drawback, and the embodiment of the present invention builds dimension and acceleration constant and inertia weight respectively
Vertical contact, in the situation that other parameters are constant it can be seen from above-mentioned formula, dimension is bigger, then acceleration constant and inertia power
Weight is smaller.Due to acceleration constant and inertia weight and particle pace of change positive correlation, acceleration constant and inertia weight are smaller,
Particle pace of change is smaller, and convergence is also better.Parameter 0.2,0.4 and 0.5 in the verified embodiment of the present invention is especially suitable
Together in the Needle-driven Robot of small volume (length/width/height is in 30-50cm ranges).
Referring to Fig. 4, the embodiment of the present invention provides a kind of calibration system of robot, including robot body 403, connection
The robot controller 401 of robot body 403, the acquisition dress of connection robot controller 401 and robot body 403
It sets 404 and connects the caliberating device 402 of harvester 404.
Robot controller 401 communicates to connect robot body 403, for the embodiment of the present invention, robot control dress
Setting 401 can be integrated in robot body 403, such as the forms such as controller；And in another embodiment, robot control dress
It sets 401 to be alternatively separately from the device outside robot body 403, such as switch board etc.；Need to illustrate when, either controller
Or switch board all have can with the hardware system of runs software, including such as processor (MCU or CPU or other realizations),
Memory (ROM or RAM etc. or other realizations).
Caliberating device 402 communicates to connect harvester 404, and for the embodiment of the present invention, caliberating device 402 can be meter
Calculation machine, industrial personal computer, microcomputer or any equipment with data processing function can also be the information upload for connecting these equipment
Terminal；Caliberating device 402 can be separately provided, and can also be additionally installed at robot body 402, robot controller 401
Or on harvester 404.
Robot body 403 can be the mechanical arm of multiaxis, be that by automatically controlling, repeatable programming, more
Degree of freedom, freedom of motion abbreviation space right-angle relationship, multiduty operation machine.The behavior of its work is mainly by complete
At along the linear movement on X, Y and Z axis, robot, which can be any type, can be described by D-H representations, model but not only limit
In the robot modeling for the description of D-H identifiers, modeling.
Harvester 404 communicates to connect robot body 403 and robot controller 401, for acquiring robot sheet
The control parameter of the mechanical parameter and robot controller 401 of body 403, harvester 404 include laser tracker, magnetic force biography
Sensor etc. can also be the equipment that other support information terminal news.
In embodiments of the present invention, pass through wired or wireless ether Netcom between robot body 403 and harvester 404
Letter connection, and/or, it is communicated to connect by wired or wireless Ethernet between harvester 404 and caliberating device 402,
The agreement of communication includes but not limited to：Ethernet TCP/IP,EthernetCAT,Profibus-DP,CC-Link,
Any one in Modbus, RS22 and other standard agreements or custom protocol.
The caliberating device of the robot of the embodiment of the present invention is first used more by determining and each relevant parameter of amount to be calibrated
Dimension particle cluster algorithm calculates the optimal compensation value of all parameters, achievees the purpose that Fast Convergent, then pass through multidimensional particle cluster algorithm
And the optimal compensation value of all parameters, the optimal compensation value is sought to each amount to be calibrated successively, realizes all amounts to be calibrated
Calibration.Fig. 5 is the functional block diagram of the caliberating device of the robot of the embodiment of the present invention, as shown, caliberating device includes：
Acquisition module 501, for determining the standard value with each relevant parameter of amount to be calibrated and each parameter, by default
The measured value of the measured value and parameter of each amount to be calibrated when frequency collection robot is run；
It should be noted that amount crucial when amount to be calibrated is robot motion, for example, joint of robot angle data,
Sitting posture data etc., the embodiment of the present invention do not limit to specific parameter and number in amount to be calibrated.General a variety of amounts to be calibrated
Between also can exist influence each other.For each amount to be calibrated, there are some to influence the amount to be calibrated for robot body
Parameter (including mechanical parameter and control parameter), and the case where overlapping can also be had by influencing the parameter of each amount to be calibrated.This hair
Bright embodiment can determine the standard value of these parameters, such as robot lift hand height as an amount to be calibrated first, and lift hand is high
Degree can be related to the length of some connecting rod, and this length is just used as a parameter, and more specifically, the length of connecting rod is as one
Kind mechanical parameter there is standard value, for example, 10cm in making machine people, but when robot starts to lift manual make, by
Small variation can occur for the length of the interaction between different component, connecting rod, add measurement accuracy the problem of so that
Each measured value all may be different, therefore the embodiment of the present invention can press preset frequency to this parameter in this course
(length of connecting rod) measures, and records measured value.Similarly, robot lift hand height is equally needed as an amount to be calibrated
It to be acquired with predeterminated frequency, the measured value as the amount to be calibrated.
One layer of demarcating module 502 is used for for any one parameter, according to multidimensional particle cluster algorithm and the mark of the parameter
Quasi- value demarcates the measured value of the parameter；
It should be noted that it is machinery and control parameter that the parameter demarcated is needed in the embodiment of the present invention, it is not movement
Parameter is learned, the number of kinematics parameters is generally 6, and the quantity of parameter is far more than kinematics parameters in the embodiment of the present invention
Number, and there are multi-dimensional complicated relationship between parameter, therefore the embodiment of the present invention overcomes the calibration process needs of parameter
The multidimensional multi-target parameter optimizing problem of complex relationship can not demarcate multi-parameter using one-dimensional particle cluster algorithm, therefore
The embodiment of the present invention demarcates parameter using multidimensional particle cluster algorithm.
In multidimensional particle cluster algorithm, the potential solution of each optimization problem can be imagined as d dimension spaces search space
Point on as soon as, this point are referred to as particle (particle), and all particles are all suitable there are one being determined by object function
It should be worth (fitness value), there are one speed to determine the direction circled in the air and distance for each particle, and then particles are just followed
Current optimal particle is in Searching Resolution Space.In embodiments of the present invention, using times of collection as dimension, for each ginseng
It counts, there are one deviations for the measured value and standard value under each dimension, and therefore, the purpose using multidimensional particle cluster algorithm is exactly to obtain
Obtain a supplement value for making above-mentioned deviation minimum of the particle under all dimensions, such as survey of some parameter under 3 dimensions
Magnitude is respectively 9.99,9.99 and 10.01, and the standard value of the parameter is 10.00, finds out the measured value and standard value of every dimension
Difference be -0.01, -0.01 and 0.01, can be in the hope of being measured value of the parameter under 3 dimensions by multidimensional particle cluster algorithm
And the offset of actual value deviation minimum is 0.01, by the way that the offset compensates measured value, the survey after being compensated
Magnitude is 10.00,10.00 and 10.02, although still there is the measured value after 1 supplement to be not equal to actual value, most of parameter is
Optimized to equal than actual value, it is believed that optimization is completed.
Two layers of demarcating module 503, for for any one amount to be calibrated, knowing that relevant other wait for according to constraint equation
Scalar quantity and each parameter, according to after the relevant each parameter calibration of amount to be calibrated measured value and be completed calibration other
Amount to be calibrated obtains the calculated value of the amount to be calibrated, according to multidimensional particle cluster algorithm and the measured value of the amount to be calibrated, to this
The calculated value of amount to be calibrated is demarcated.
For the angle of structure, for any one amount to be calibrated, the amount to be calibrated and a certain number of parameters are very
To also related to other amounts to be calibrated, therefore, pass through integrated structure and kinematics analysis, so that it may to obtain robot architecture's
Constraint equation, analytic solutions of the amount to be calibrated as constraint equation.In general, the number of the number of constraint equation and amount to be calibrated
Unanimously, and there are amount to be calibrated the solution based on another amount to be calibrated is needed the case where calculating, therefore, to implement in the present invention
When treating scalar quantity progress multidimensional particle cluster algorithm calibration in example, first to only being optimized with the relevant amount to be calibrated of parameter,
It after amount to be calibrated optimization (the demarcating), then pair is optimized with other relevant amounts to be calibrated of the amount to be calibrated, up to institute
Need scalar quantity calibration to be completed.
The Robot calibration device of the embodiment of the present invention is capable of multiple and different error conditions of the multiple types of Fast Calibration
Scalar quantity, with fast convergence rate, local extremum incidence is low, can effectively solve the problem that the beneficial of multi-dimensional complicated Relation Parameters calibration
Effect.
On the basis of the above embodiments, one layer of demarcating module includes：
First fitness function establishes unit, is used for any one parameter, using times of collection as dimension, according to the ginseng
Deviation of the number under all dimensions between measured value and standard value establishes fitness function；
For example, fitness function fd(si)=Cd(si)-Sd(si), wherein siIndicate that i-th of parameter, d indicate dimension, Cd
(si) indicate parameter s under d dimensionsiMeasured value, Sd(si) indicate parameter s under d dimensionsiStandard value.It should be noted that measuring
Deviation between value and standard value can indicate incessantly with simple difference, can also use square, square of difference difference,
Absolute value of difference etc. form indicates that the embodiment of the present invention does not limit this specifically.
First population establishes unit, for being made with deviation of the parameter under each dimension between measured value and standard value
For a particle, population is established, initializes the position and speed of each particle；
First computing unit, the iterative search for carrying out particle with multidimensional particle cluster algorithm, when iterative search to adaptation
When spending function minimum, optimal solution is obtained；
The specific an iteration process of multidimensional particle cluster algorithm is described below：
Carry out the update of speed and position, wherein grain to each particle using multidimensional particle cluster algorithm with the first computing unit
Sub- speed formula is：
The location formula of particle is：
Wherein, Lvx, LxxFor Restriction Operators, it is specified that the threshold speed and position threshold of algorithm,For particle a
In dimension xda(t) the jth item of velocity vector on,For the velocity vector of following iteration number,It is particle a in dimension xda(t) the jth item of position vector on,For following iteration number
Position vector,It is particle a in dimension xda(t) the jth item of individual optimum position vector on,For the jth item of the population overall situation optimum position vector on dimension d, LvxIt is rate limitation range, LxxIt is
Position limits range.c1, c2For acceleration constant, w (t) is inertia weight, r1,j(t), r2,j(t) it is the random number of (0,1).
The corresponding fitness function of t+1 moment each particle is calculated, itself optimum position of a particle is obtained, and is obtained
The global optimum position of population；
To each particle, compare the corresponding fitness function in t moment itself optimum position and itself optimum position of t+1 moment
The fitness function of itself optimum position of t+1 moment is assigned to t moment itself most by corresponding fitness function if the former is big
The fitness function of best placement.
Compare the adaptation corresponding with the moment overall situation optimum positions t+1 of the corresponding fitness function in t moment overall situation optimum position
Function is spent, if the former is big, the fitness function of t+1 moment overall situations optimum position is assigned to the suitable of t moment overall situation optimum position
Response function.
Judge whether to reach iteration deployment, if reaching iterative steps, obtains all particles of current iteration itself optimum bits
It sets and otherwise global optimum position carries out next iteration using global optimum position as final prioritization scheme.
First calibration unit, for being demarcated to the measured value at the parameter each moment according to optimal solution.
Specifically, such as measured value of some parameter 3 moment is respectively 5.51,5.52 and 5.53, and optimal solution is
0.01, then calibrated measured value is 5.52,5.53 and 5.54.It should be noted that the measured value to parameter is demarcated
It does not need to be ranked up parameter, i.e., need not be demarcated in a fixed order.
It should be noted that the measured value of some parameters can have larger error, and this error can greatly influence
The convergence of multidimensional particle cluster algorithm, therefore before proceeding by multidimensional particle cluster algorithm, need to screen particle, lead to
It crosses and screens out the error beyond error range.Therefore, on the basis of the above embodiments, the first computing unit is additionally operable to：With
It, will be between measured value and standard value according to preset deviation threshold before multidimensional particle cluster algorithm carries out the iterative search of particle
The larger particle of deviation is rejected from the population.
On the basis of the above embodiments, two layers of demarcating module include：
Second fitness function establishes unit, is used for for any one amount to be calibrated, using times of collection as dimension, root
Fitness function is established according to the deviation of the amount to be calibrated under all dimensions between measured value and calculated value；
Second population establishes unit, for inclined between measured value and calculated value under each dimension with the amount to be calibrated
Difference is used as a particle, establishes population, initializes the position and speed of each particle；
Second computing unit, the iterative search for carrying out particle with multidimensional particle cluster algorithm, when iterative search to adaptation
When spending function minimum, optimal solution is obtained；
The second computing unit is described with the specific an iteration process of multidimensional particle cluster algorithm below：
The update of speed and position is carried out to each particle with multidimensional particle cluster algorithm, wherein particle rapidity formula is：
The location formula of particle is：
Wherein, Lvx, LxxFor Restriction Operators, it is specified that the threshold speed and position threshold of algorithm,For particle a
In dimension xda(t) the jth item of velocity vector on,For the velocity vector of following iteration number,It is particle a in dimension xda(t) the jth item of position vector on,For following iteration number
Position vector,It is particle a in dimension xda(t) the jth item of individual optimum position vector on,For the jth item of the population overall situation optimum position vector on dimension d, LvxIt is rate limitation range, LxxIt is
Position limits range.c1, c2For acceleration constant, w (t) is inertia weight, r1,j(t), r2,j(t) it is the random number of (0,1).
The corresponding fitness function of t+1 moment each particle is calculated, itself optimum position of a particle is obtained, and is obtained
The global optimum position of population；
To each particle, compare the corresponding fitness function in t moment itself optimum position and itself optimum position of t+1 moment
The fitness function of itself optimum position of t+1 moment is assigned to t moment itself most by corresponding fitness function if the former is big
The fitness function of best placement.
Compare the adaptation corresponding with the moment overall situation optimum positions t+1 of the corresponding fitness function in t moment overall situation optimum position
Function is spent, if the former is big, the fitness function of t+1 moment overall situations optimum position is assigned to the suitable of t moment overall situation optimum position
Response function.
Judge whether to reach iteration deployment, if reaching iterative steps, obtains all particles of current iteration itself optimum bits
It sets and otherwise global optimum position carries out next iteration using global optimum position as final prioritization scheme.
Second calibration unit, for being demarcated to the calculated value at amount each moment to be calibrated according to optimal solution.
It should be noted that the measured value of some parameters can have larger error, and this error can greatly influence
The convergence of multidimensional particle cluster algorithm, therefore before proceeding by multidimensional particle cluster algorithm, need to screen particle, lead to
It crosses and screens out the error beyond error range.Therefore, on the basis of the above embodiments, the second computing unit is additionally operable to：With
It, will be between measured value and standard value according to preset deviation threshold before multidimensional particle cluster algorithm carries out the iterative search of particle
The larger particle of deviation is rejected from the population.
On the basis of the various embodiments described above, when one layer of demarcating module and two layers of demarcating module execute multidimensional particle cluster algorithm
The acceleration constant used is obtained according to the number of the total number of parameter, times of collection and constraint equation；
One layer of demarcating module and two layers of demarcating module execute the inertia weight used when multidimensional particle cluster algorithm according to acquisition
Number obtains.
Illustrate the scaling method of the embodiment of the present invention with a specific embodiment below.In the present embodiment, robot mark
It includes Needle-driven Robot, control device, harvester and caliberating device to determine system.
Control device includes first motor, the second motor and third motor, before 3 motors control Needle-driven Robot respectively
Into-retreat, it faces upward-bows and inserting needle-withdraw of the needle.The rotation of these motors can feed back code value to harvester, and the code value fed back passes through
Harvester calculates, and can obtain motor position information value of feedback, such as the step number of motor rotation.
Harvester, including puncture needle pressure sensor, two magnetometric sensors, magnetic field generator and for obtain wear
The device of the mechanical parameter and control parameter of robot is pierced, puncture needle pressure sensor is for acquiring puncture needle to test model
Pressure, two magnetometric sensors are as detecting instrument, wherein the first magnetometric sensor is fixed in puncture needle, the second magnetic force sensing
Device is fixed on one end of ultrasonic probe, to be fixed on the image starting point of ultrasonic probe as standard, Needle-driven Robot coordinate system
Dead-center position is the position of the second magnetometric sensor.Magnetic field generator is mounted below test model, which is available
In ultrasonic puncture manikin or can be used for the vascular pattern of ultrasonic puncture.
Using the second magnetometric sensor of dead-center position as reference coordinate, puncture angle is measured by the first magnetometric sensor
ACi, puncture entry point location BCiAnd paracentesis depth position CCi.Wherein ACi, BCi, CCiIndicate that the i-th moment punctured angle respectively
Degree, the measured value of entrance and depth.
Using the second magnetometric sensor of dead-center position as reference coordinate, caliberating device is calculated by servo motor position feedback
Go out puncture angle AS to be calibratedi, puncture entry point location coordinate BSiAnd paracentesis depth position coordinates CSi, ASi, BSi, CSi
The i-th moment puncture angle, the calculated value of entrance and depth are indicated respectively.
Puncture angle ASiIt is by parameter S1、S2、S3、S4、S5And No. 2 motor zero control compensating parameter S6(due to motor
It is needed when returning dead-center position using close to switch, so dead-center position is there are certain error, and Zero magnitude control supplementary parameter
Exactly it is used to adjust dead-center position), No. 2 motor position feedback value S7It is calculated.
Puncture entry point location coordinate BSiIt is by parameter S8、S9、S10、S13、ASiAnd No. 1 motor zero control compensation ginseng
Number S12, No. 1 motor position feedback value S11It is calculated.
Paracentesis depth position coordinates CSiIt is by parameter S9、S10、S14, and ASi, BSi, No. 3 motor zero control compensation ginsengs
Number S15It is calculated.
S1-S16Type include mechanical dimension, mechanism assembly initial supplement, angle of assembling, zero compensation, motor position model
It encloses, instrument size and fitting dimension etc., the type of the error caused by these parameters is broadly divided into mismachining tolerance, assembly misses
Difference, control error, tactful error etc..
Three amounts to be calibrated --- puncture angle ASi, puncture entry point location coordinate BSiWith paracentesis depth position coordinates CSi
Between interact, while influencing the machinery of these three amounts to be calibrated and control parameter and also have coincidence.
Build restriction on the parameters relation equation Fn{Sj},SjThe collection of the machinery, control and the measurement parameter that are related to for value to be calibrated
It closes, j is the number of parameter, and n is the quantity of calibration value.Constraint equation is as follows in the present embodiment：
Calculate measured value ACi, BCi, and CCiWith calculated value ASi, BSi, and CSiBetween error.
Due to calculating the error of gained it sometimes appear that very big error, this error can very big influence multidimensional particles
Group's convergence, it is therefore desirable to error be screened, the error beyond error range is weeded out by screening：
Above-mentioned scaling method through the embodiment of the present invention, first get parms S1The optimal compensation value at 1-100 moment, then
Obtain S2In the optimal compensation value at 1-100 moment, until obtaining S16In the optimal compensation value at 1-100 moment.By calculating,
Measured value of the embodiment of the present invention just to all parameters at each moment is compensated.
For tri- amounts to be calibrated of AS, BS and CS, by constraint equation it is found that BS needs to obtain based on AS, and
CS needs to obtain based on AS and BS again, therefore with reference to the scaling method of the above embodiment of the present invention, first to AS in institute
Sometimes the calculated value inscribed compensates, and is then compensated again to the BS calculated values inscribed when all, finally sometimes to CS institutes
The calculated value inscribed compensates.
The apparatus embodiments described above are merely exemplary, wherein can be as the unit that separating component illustrates
Or may not be and be physically separated, the component shown as unit may or may not be physical unit, i.e.,
A place can be located at, or may be distributed over multiple network units.It can select according to the actual needs therein
Some or all of module achieves the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying creative labor
In the case of dynamic, you can to understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It is realized by the mode of software plus required general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be expressed in the form of software products in other words, should
Computer software product can store in a computer-readable storage medium, such as ROM/RAM, magnetic disc, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
The method of certain parts of example or embodiment.
Finally it should be noted that：The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, it will be understood by those of ordinary skill in the art that：It still may be used
With technical scheme described in the above embodiments is modified or equivalent replacement of some of the technical features；
And these modifications or replacements, various embodiments of the present invention technical solution that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (10)
1. a kind of scaling method of robot, which is characterized in that including：
The standard value with each relevant parameter of amount to be calibrated and each parameter is determined, by predeterminated frequency acquisition robot fortune
The measured value of the measured value and parameter of each amount to be calibrated when row；
For any one parameter, the measured value of the parameter is carried out according to the standard value of multidimensional particle cluster algorithm and the parameter
Calibration；
For any one amount to be calibrated, other relevant amounts to be calibrated and each parameter are known according to constraint equation, according to this
Measured value after the relevant each parameter calibration of amount to be calibrated and other described amounts to be calibrated that calibration is completed, obtain this and wait marking
Quantitative calculated value, according to multidimensional particle cluster algorithm and the measured value of the amount to be calibrated, to the calculated value of the amount to be calibrated into
Rower is fixed.
2. scaling method according to claim 1, which is characterized in that it is described for any one parameter, according to multidimensional grain
The step of standard value of swarm optimization and the parameter demarcates the measured value of the parameter, including：
Using times of collection as dimension, is established and adapted to according to deviation of the parameter under all dimensions between measured value and standard value
Spend function；
Using deviation of the parameter under each dimension between measured value and standard value as a particle, population is established, initially
Change the position and speed of each particle；
The iterative search that particle is carried out with multidimensional particle cluster algorithm is obtained when iterative search to the fitness function minimum
Optimal solution；
The measured value at the parameter each moment is demarcated according to the optimal solution.
3. scaling method according to claim 2, which is characterized in that described to carry out changing for particle with multidimensional particle cluster algorithm
The step of generation search, further include before：
According to preset deviation threshold, the larger particle of deviation between measured value and standard value is rejected from the population.
4. scaling method according to claim 1, which is characterized in that it is described according to multidimensional particle cluster algorithm and this wait marking
Quantitative measured value, the step of calibration to the calculated value of the amount to be calibrated, including：
For any one amount to be calibrated, using times of collection as dimension, according to the amount to be calibrated under all dimensions measured value
Deviation between calculated value establishes fitness function；
Using the deviation of the amount to be calibrated under each dimension between measured value and calculated value as a particle, population is established,
Initialize the position and speed of each particle；
The iterative search that particle is carried out with multidimensional particle cluster algorithm is obtained when iterative search to the fitness function minimum
Optimal solution；
The calculated value at amount each moment to be calibrated is demarcated according to the optimal solution.
5. scaling method according to claim 1, which is characterized in that the acceleration used in the multidimensional particle cluster algorithm
Constant is obtained according to the number of the total number of the parameter, times of collection and the constraint equation；
The inertia weight used in the multidimensional particle cluster algorithm is obtained according to the times of collection.
6. a kind of caliberating device of robot, which is characterized in that including：
Acquisition module, for determining the standard value with each relevant parameter of amount to be calibrated and each parameter, by default frequency
Rate acquires the measured value of each amount to be calibrated and parameter when robot is run；
One layer of demarcating module is used for for any one parameter, according to multidimensional particle cluster algorithm and the standard value pair of the parameter
The measured value of the parameter is demarcated；
Two layers of demarcating module, for for any one amount to be calibrated, knowing other relevant amounts to be calibrated according to constraint equation
With each parameter, according to after the relevant each parameter calibration of amount to be calibrated measured value and other be completed described in calibration wait for
Scalar quantity obtains the calculated value of the amount to be calibrated, according to multidimensional particle cluster algorithm and the measured value of the amount to be calibrated, waits for this
Gauged calculated value is demarcated.
7. caliberating device according to claim 6, which is characterized in that one layer of demarcating module include：
First fitness function establishes unit, for being existed according to the parameter using times of collection as dimension to any one parameter
Deviation under all dimensions between measured value and standard value establishes fitness function；
First population establishes unit, for using deviation of the parameter under each dimension between measured value and standard value as one
A particle, establishes population, initializes the position and speed of each particle；
First computing unit, the iterative search for carrying out particle with multidimensional particle cluster algorithm, when iterative search to the adaptation
When spending function minimum, optimal solution is obtained；
First calibration unit, for being demarcated to the measured value at the parameter each moment according to the optimal solution.
8. caliberating device according to claim 6, which is characterized in that first computing unit is additionally operable to：
Before the iterative search for carrying out particle with multidimensional particle cluster algorithm, according to preset deviation threshold, by measured value and mark
The larger particle of deviation is rejected from the population between quasi- value.
9. caliberating device according to claim 6, which is characterized in that two layers of demarcating module include：
Second fitness function establishes unit, is used for for any one amount to be calibrated, using times of collection as dimension, according to this
Deviation of the amount to be calibrated under all dimensions between measured value and calculated value establishes fitness function；
Second population establishes unit, for being made with deviation of the amount to be calibrated under each dimension between measured value and calculated value
For a particle, population is established, initializes the position and speed of each particle；
Second computing unit, the iterative search for carrying out particle with multidimensional particle cluster algorithm, when iterative search to the adaptation
When spending function minimum, optimal solution is obtained；
Second calibration unit, for being demarcated to the calculated value at amount each moment to be calibrated according to the optimal solution.
10. caliberating device according to claim 6, which is characterized in that one layer of demarcating module and two layers of demarcating module
The acceleration constant used when multidimensional particle cluster algorithm is executed according to the total number of the parameter, times of collection and the constraint
The number of equation obtains；
One layer of demarcating module and two layers of demarcating module execute the inertia weight used when multidimensional particle cluster algorithm according to
Times of collection obtains.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810317759.0A CN108742840B (en) | 2018-04-10 | 2018-04-10 | Calibration method and device for robot |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810317759.0A CN108742840B (en) | 2018-04-10 | 2018-04-10 | Calibration method and device for robot |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108742840A true CN108742840A (en) | 2018-11-06 |
| CN108742840B CN108742840B (en) | 2020-07-17 |
Family
ID=63981320
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810317759.0A Expired - Fee Related CN108742840B (en) | 2018-04-10 | 2018-04-10 | Calibration method and device for robot |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108742840B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113676387A (en) * | 2021-08-11 | 2021-11-19 | 追觅创新科技(苏州)有限公司 | Zero calibration method and device for multi-legged robot, storage medium and electronic device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102129249A (en) * | 2011-01-10 | 2011-07-20 | 中国矿业大学 | Method for planning global path of robot under risk source environment |
| CN104331083A (en) * | 2014-11-21 | 2015-02-04 | 大连大学 | Method for optimizing wide-angle attitude control parameters of spacecraft |
| CN104517297A (en) * | 2013-09-28 | 2015-04-15 | 沈阳新松机器人自动化股份有限公司 | Robot calibrate method based on particle swarm optimization |
| CN105159096A (en) * | 2015-10-10 | 2015-12-16 | 北京邮电大学 | Redundancy space manipulator joint torque optimization method based on particle swarm algorithm |
| US20160374541A1 (en) * | 2014-10-24 | 2016-12-29 | Auris Surgical Robotics, Inc. | Automated endoscope calibration |
| CN106447022A (en) * | 2016-08-03 | 2017-02-22 | 长沙理工大学 | Multi-objective particle swarm optimization based on globally optimal solution selected according to regions and individual optimal solution selected according to proximity |
| CN106647744A (en) * | 2016-11-14 | 2017-05-10 | 深圳大学 | Robot path planning method and device |
| CN107685343A (en) * | 2017-08-28 | 2018-02-13 | 北京邮电大学 | A kind of Mechanical transmission test parameter calibration configuration optimization method |
-
2018
- 2018-04-10 CN CN201810317759.0A patent/CN108742840B/en not_active Expired - Fee Related
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102129249A (en) * | 2011-01-10 | 2011-07-20 | 中国矿业大学 | Method for planning global path of robot under risk source environment |
| CN104517297A (en) * | 2013-09-28 | 2015-04-15 | 沈阳新松机器人自动化股份有限公司 | Robot calibrate method based on particle swarm optimization |
| US20160374541A1 (en) * | 2014-10-24 | 2016-12-29 | Auris Surgical Robotics, Inc. | Automated endoscope calibration |
| CN104331083A (en) * | 2014-11-21 | 2015-02-04 | 大连大学 | Method for optimizing wide-angle attitude control parameters of spacecraft |
| CN105159096A (en) * | 2015-10-10 | 2015-12-16 | 北京邮电大学 | Redundancy space manipulator joint torque optimization method based on particle swarm algorithm |
| CN106447022A (en) * | 2016-08-03 | 2017-02-22 | 长沙理工大学 | Multi-objective particle swarm optimization based on globally optimal solution selected according to regions and individual optimal solution selected according to proximity |
| CN106647744A (en) * | 2016-11-14 | 2017-05-10 | 深圳大学 | Robot path planning method and device |
| CN107685343A (en) * | 2017-08-28 | 2018-02-13 | 北京邮电大学 | A kind of Mechanical transmission test parameter calibration configuration optimization method |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113676387A (en) * | 2021-08-11 | 2021-11-19 | 追觅创新科技(苏州)有限公司 | Zero calibration method and device for multi-legged robot, storage medium and electronic device |
| CN113676387B (en) * | 2021-08-11 | 2023-02-24 | 北京小米移动软件有限公司 | Zero calibration method and device for multi-legged robot, storage medium and electronic device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108742840B (en) | 2020-07-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111590581B (en) | Positioning compensation method and device for robot | |
| CN105945948B (en) | A kind of online quick calibrating methods of TCP applied to industrial robot and device | |
| CN104057290B (en) | A kind of robotic asssembly method and system of view-based access control model and force-feedback control | |
| Orange et al. | A transfer entropy analysis of leader-follower interactions in flying bats | |
| EP2910895B1 (en) | Coordinate measuring machine and method for calculating correction matrix by coordinate measuring machine | |
| CN106514651A (en) | Measurement system and calibration method | |
| CN110088797A (en) | Industrial equipment image recognition processor and controller | |
| CN105259786B (en) | The inertial parameter discrimination method and device of target to be identified | |
| JP3634275B2 (en) | Position measuring device | |
| CN105643641A (en) | Force sensor calibration device and method and force control robot | |
| CN107718050A (en) | SCARA robots brachium and Zero calibration method, device, medium and computer equipment | |
| US20230042756A1 (en) | Autonomous mobile grabbing method for mechanical arm based on visual-haptic fusion under complex illumination condition | |
| CN105865341B (en) | Industrial robot spatial pose repetitive positioning accuracy measuring device and method | |
| CN111506199B (en) | Kinect-based high-precision unmarked whole-body motion tracking system | |
| CN107076552B (en) | Method and coordinate measuring machine for single-point scanning of workpieces | |
| CN114398049A (en) | Self-adaptive dynamic updating method for digital twin model of discrete manufacturing workshop | |
| CN113176529B (en) | Magnetic gradiometer correction method, magnetic gradiometer correction device and storage medium | |
| CN112884842A (en) | Synchronous calibration method for zero position of double rudder wheels and external parameters of acquisition device and movable platform | |
| CN107471257A (en) | Robot geometric calibration method based on single stay encoder | |
| CN112223283A (en) | Robot arm, robot arm control method, processing apparatus, and medium | |
| CN108742840A (en) | The scaling method and device of robot | |
| CN106558070B (en) | A kind of method and system of vision tracking under the robot based on Delta | |
| CN108917595A (en) | Glass on-line measuring device based on machine vision | |
| CN110900608B (en) | Robot kinematics calibration method based on optimal measurement configuration selection | |
| CN111257954A (en) | Vehicle-mounted array type detection method and system based on feature inversion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200717 |