Triaxial micro-displacement driving device for motion planning based on genetic algorithm

Abstract

The invention discloses a triaxial micro-displacement driving device for motion planning based on a genetic algorithm. The triaxial micro-displacement driving device is characterized by comprising an execution mechanism part, a control system part and a motion planning program part. The actuating mechanism part consists of a piezoelectric ceramic stack, a micro magnet, a micro slide rail and a shell, and can receive an electric signal sent by the control system and perform micro displacement output; the control system part is composed of a single-chip microcomputer and a drive circuit and can output an electric drive signal according to the result of a motion planning program. The motion planning program is a step number distribution program adopting a genetic algorithm, mathematical solution is carried out on the motion planning optimization model of the displacement platform, and motion planning can be carried out according to target displacement given by a user. According to the invention, the three-axial displacement output with micron-level precision can be realized by adopting a smaller actuating mechanism and a lower price, and the three-axial displacement output device has great significance in subjects such as material science and biological science.

Description

Technical Field

[0001] This invention discloses a micro-displacement drive device, and more particularly, relates to a three-axis micro-displacement drive device for motion planning based on a genetic algorithm. Background Art

[0002] With the increasing demand for precision displacement in the fields of materials science and life sciences and the miniaturization of precision devices, designing a displacement platform that can simultaneously have the characteristics of compact structure and support high-precision resolution has become an important demand in academia and engineering. Currently, the widely used displacement platforms generally choose to neutralize these two requirements, or design asymmetric structures, combining engineering needs to sacrifice one requirement for the other. At present, the constraint relationship between the miniaturization of the displacement platform and the output accuracy is still a topic to be studied.

[0003] Compared with other controllers, piezoelectric ceramics have significant advantages in ultra-precision positioning and control. The structure itself has no gaps, no mechanical friction, and has the advantages of easy control, high resolution, fast response speed, no influence of magnetic fields, and low cost. The inverse piezoelectric effect of piezoelectric ceramics is used to make the piezoelectric material generate corresponding mechanical stress when different voltages are applied to drive the micro-displacement to work. At the same time, the micro-displacement output length of the piezoelectric ceramic can be controlled by adjusting the voltage input to the piezoelectric ceramic stack.

[0004] The friction inertia piezoelectric ceramic stack actuator based on the stick-slip mechanism can achieve sub-nanometer resolution while ensuring a compact drive structure. However, the complex preload mechanism, precision machining and precision assembly requirements increase the cost of traditional friction inertia nanopositioners. Therefore, by utilizing the excellent characteristics of piezoelectric ceramics with small size and large output force and the stick-slip drive principle for structural design, it is possible to maintain a relatively low cost and a certain output accuracy while ensuring a compact drive structure.

[0005] When reducing and simplifying the design of the drive mechanism, the output accuracy of the displacement platform will inevitably be impaired. Therefore, a motion planning program is needed to distribute the motion of the output of the drive structure, and the accuracy of the displacement output is guaranteed by combining the program and the drive structure. Genetic algorithms can explore in a wide search space and are suitable for solving complex optimization problems and avoiding falling into local optimality. Genetic algorithms are applicable to various types of problems (such as continuous, discrete, and multi-objective optimization), and can handle uncertain and noisy data. Therefore, using genetic algorithms to mathematically model and solve the optimization problem of the number of steps of the displacement platform output can solve the problem of insufficient output accuracy of the displacement platform. Summary of the invention

[0006] The purpose of this application is to design a displacement drive platform based on a genetic algorithm with a small drive mechanism, high displacement accuracy and low price, so as to realize the small drive mechanism and high resolution application required in engineering.

[0007] The technical solution is as follows:

[0008] This project design includes three sub-parts: motion planning program part, control system part and actuator part.

[0009] The actuator of the displacement motion platform is a motion platform designed based on the stick-slip mechanism. It combines the stick-slip drive module with the displacement platform and outputs micron-level precision displacement according to the working mode and instructions. The actuator is designed with an interface to connect the control system, so that the electric drive signal transmitted from the control system drives the actuator.

[0010] The control system consists of a single-chip microcomputer and a control circuit. The single-chip microcomputer can be used to receive the control program and motion planning results of the host computer, and output the control signal to the drive circuit. The drive circuit solves the single-chip control signal and outputs the drive current to the actuator.

[0011] The motion planning program is a step allocation program that uses a genetic algorithm to mathematically solve the motion planning optimization model of the displacement platform, and can perform motion planning based on the target displacement given by the user. The number of steps for each gear can be planned and the control program can be burned into the microcontroller.

[0012] The actuator part is composed of a piezoelectric ceramic stack, a micro magnet, a micro slide rail and a housing. It can receive the electrical signal from the control system and output a small displacement. The displacement output step can be adjusted according to the input drive voltage signal. It is characterized in that the three-axis displacement devices are independent of each other and can work separately in an independent state. Each axial device is fixed by screws, and the three-axis movements do not interfere with each other.

[0013] The micro slide rail is characterized in that it comprises a slider, a slide rail and a ball bearing, wherein both the slider and the slide rail are provided with screw holes for fixing, and the slider and the slide rail are in contact and slide with each other through the ball bearing.

[0014] The housing is characterized in that: for each axial displacement device, the housing is divided into two parts, one part is fixed to the slider of the micro slide rail, and the other part is fixed to the slide rail of the micro slide rail. The slider part can produce an adsorption effect with the magnet.

[0015] The micro magnet is characterized in that the micro magnet is adsorbed on the slider by magnetic force and has no friction contact with the shell.

[0016] The piezoelectric ceramic stack is characterized in that two ends are fixed to the housing and the magnet respectively, and the power supply circuit extends outside the housing and is connected to the drive circuit.

[0017] The stick-slip inertial piezoelectric actuator is driven by the principle of friction inertia, and realizes movement by accumulating displacement under the action of continuous driving signals. Its movement process mainly uses the piezoelectric stack as the driving source. The piezoelectric stack is continuously driven under the action of the continuous sawtooth wave driving signal, and the moving unit of the stick-slip inertial piezoelectric actuator accumulates displacement, and the displacement direction of the moving unit can also be changed. The stick-slip inertial piezoelectric actuator uses an asymmetric sawtooth wave as the driving signal.

[0018] The control system is composed of a single chip microcomputer and a drive circuit, and can output an electric drive signal based on the result of the motion planning program. The drive signal is generated by the single chip microcomputer and converted and amplified by the drive circuit.

[0019] The single chip microcomputer is characterized in that the program it runs can output an asymmetric sawtooth wave with corresponding steps corresponding to the voltage reserved gear according to the result of the motion planning program, so as to realize the stick-slip inertial piezoelectric drive under the voltage corresponding to different gears.

[0020] The driving circuit is characterized by having a digital-to-analog conversion function and a power amplification function. It can receive eight-bit binary information sent by the single-chip microcomputer and convert it into analog signals respectively. Subsequently, the power of the signal is amplified to a suitable size for driving the piezoelectric ceramic stack and transmitted to the driving device.

[0021] After completing the actuator part and the control system part, it is necessary to connect and conduct a power-on test to obtain the voltage characteristic curve of the device under the user's operating conditions. In the subsequent optimization process, it is necessary to know the specific voltage characteristics under the user's operating conditions to provide the step size parameters in the optimization mathematical model.

[0022] Reserve voltage levels, each of which corresponds to the driving voltage of the displacement platform. For the target displacement, the output steps of the two voltage levels are planned to enable the slide to meet the user's displacement output requirements. The influence of boundary effects is minimized by balancing the steps, thereby improving output accuracy.

[0023] The motion planning program is a step allocation program using a genetic algorithm to mathematically solve the motion planning optimization model of the displacement platform. On the basis of driving the displacement motion platform, combined with the target displacement requirements, the gear selection of the driving voltage is planned, and the genetic algorithm is used to optimize the target problem.

[0024] The optimization model is characterized by: optimizing the motion planning problem of the motion platform by mathematical modeling. For the average step lengths corresponding to different driving voltages, the steps are distributed. It is necessary to optimize the output displacement of the driving platform under the premise of considering the average distribution of the steps, so that the difference between the output displacement of the driving platform and the target distance given by the user is minimized.

[0025] The genetic algorithm is characterized in that it solves the optimization problem of the number of output steps that the driving device should correspond to in each gear position. The genetic algorithm includes an initialization function, an iteration function and a selection function.

[0026] Compared with the prior art, the beneficial results of the present invention are:

[0027] This design proposes a small-scale displacement drive platform, which transforms the displacement drive problem handled by the platform into an optimization problem of obtaining variable solutions under constraints while satisfying the minimum objective function under preset conditions. The genetic algorithm is used for optimization and solution, so that it can achieve a small-volume drive structure and have high-precision resolution while ensuring low price. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a block diagram of the overall structure of the device applied for in the present invention.

[0029] Figure 2 This is a flowchart of the genetic algorithm used in the present invention.

[0030] Figure 3 It is a structural diagram of the actuator in the implementation case of the present invention.

[0031] Figure 4 Schematic diagram of the driving circuit used in the implementation case of the present invention.

[0032] Figure 5 It is a voltage characteristic curve diagram in an implementation case of the present invention.

[0033] Figure 6 It is the error image obtained by the motion planning algorithm in the implementation case of the present invention. DETAILED DESCRIPTION

[0034] The technical solution of the present application is further described below in conjunction with the accompanying drawings. The technical solution of the present application provides a design solution for a three-axis displacement platform that can output micron-level displacement while ensuring a compact driving structure and low price. Specifically, a practical application of this project will be demonstrated below, that is, making a set of equipment with all the features in the invention content.

[0035] First, the actuator part is manufactured. The actuator part needs to have the structure and characteristics described in the invention content. Figure 3We give a design structure diagram of an actuator in a practical situation. It should be noted that the shell of this structure should not be made of magnetic materials, otherwise it will interfere with the stick-slip drive process.

[0036] Next, the driving circuit in the control system described in the invention is designed and printed. The circuit design needs to have three sets of identical circuit structures, which are used to receive and convert the three-dimensional signals of the displacement platform. The structure needs to include:

[0037] Digital-to-analog conversion module, which converts the digital signal of the microcontroller into an analog signal;

[0038] The power amplifier module amplifies the analog signal to a power level that can drive the piezoelectric ceramic stack.

[0039] Here, if Figure 4 We give the design scheme of the driving circuit in this implementation scheme.

[0040] Next, write the MCU program, which needs to be able to generate asymmetric sawtooth waves of specified size and number.

[0041] Next, connect the output of the microcontroller to the corresponding input of the drive circuit, and connect the output of the drive circuit to the drive mechanism of the corresponding dimension. Power on the drive circuit and perform a power-on test on the device.

[0042] In order to use genetic algorithms for motion planning, we need to obtain the voltage characteristic curve of the drive mechanism. It should be clear that this design adopts open-loop control, so voltage characteristic tests are required before use under different working conditions. In this implementation, a micron-level grating ruler is used for measurement. The specific method is to test how many steps are required to move 10 microns under a certain voltage, and then the average step length under that voltage can be obtained. Figure 5 We have given the voltage characteristic curve under this implementation scheme.

[0043] In the design and testing, we found that when the upper and lower limits of the driving voltage do not exceed 20V, the voltage characteristics of this design present a relatively obvious linear relationship. Therefore, in most cases, a small number of voltage characteristic samples can be taken for measurement and linear fitting to obtain the voltage characteristic curve.

[0044] After obtaining the voltage characteristic curve, the voltage gear can be reserved. In this implementation process, two voltage gears are selected for reservation. The method of gear reservation is to specify the corresponding sawtooth wave peak value in the single-chip computer program. In the drive circuit, the output signal of the single-chip computer is solved and amplified. Therefore, the waveform of the single-chip computer signal and the final output waveform of the drive circuit have a mathematically amplified corresponding relationship. In this implementation process, the corresponding relationship is linear.

[0045] Next, mathematical modeling is performed. The core problem of this algorithm is to plan how many steps of output are required for each voltage reserve gear corresponding to the set of equipment. Therefore, the corresponding number of steps for each gear should be set as the quantity to be determined. At the same time, in order to avoid boundary effects to the greatest extent, a balance operator should be introduced. In this embodiment, the final optimization modeling is as follows:

[0046] minf(x1,x2)=(n1x1+n2x2-T) 2 +C(x1-x2) 2

[0047] stx1,x2∈N * ,C∈(0,1)

[0048] in:

[0049] [x1,x2]: The number of steps corresponding to gear 1 and 2.

[0050] [n1,n2]: Step length corresponding to gear 1 and 2.

[0051] C: Parameters to be fitted. T: Target displacement.

[0052] It can be seen that the final solution target is x1, x2 and parameter C. Next, the genetic algorithm is used to solve the problem. The flowchart of the genetic algorithm is as follows Figure 2 shown.

[0053] First, define the initial value of the genetic algorithm. In this implementation, directly use:

[0054]

[0055] C=0.5

[0056] In this state, the initial value is assigned to all initial individuals. The reason for stipulating the initial value in this way is that this is the optimal solution without considering the fitting constraint, so rounding it off and assigning it to the algorithm initial value can improve the algorithm's search rate.

[0057] Next, the corresponding crossover genetic algorithm and mutation algorithm are specified for each generation. This is used to adjust the frequency of data changes between offspring and parent individuals in each iteration. It should be noted that both inheritance and mutation need to meet the constraints corresponding to the optimization problem.

[0058] Next, determine the fitness of each individual in the algorithm to determine the individual elimination conditions. At the same time, the fitness is also used to determine the activity of each population in each iteration. After each iteration, a function is established based on the required optimal solution conditions to monitor the activity of the population. In this optimization problem, the absolute value of the difference between the target drive displacement and the actual drive displacement is used as the activity judgment condition.

[0059] Finally, the program is run for multiple iterations, and the solution corresponding to the dominant population is selected after the end. This algorithm can be used to solve the number of steps while fitting the average operator. Figure 6 The motion planning residual image obtained in this embodiment is displayed, which more intuitively shows the result of motion planning.

[0060] After the above steps are completed, a set of three-axis micro-displacement drive devices including a motion planning program based on a genetic algorithm, a control system and an actuator is obtained.

Claims

1. A three-axis micro-displacement drive device based on genetic algorithm for motion planning, Its features include an actuator part, a control system part and a motion planning program part. The actuator part is composed of a piezoelectric ceramic stack, a micro magnet, a micro slide rail and a housing. It can receive the electrical signal from the control system and output a small displacement. The displacement output step can be adjusted according to the input drive voltage signal size. The control system part is composed of a single chip microcomputer and a drive circuit, and can output an electric drive signal based on the result of the motion planning program. The drive signal is generated by the microcontroller and converted and amplified by the drive circuit. The motion planning program is a step allocation program using a genetic algorithm, which mathematically solves the motion planning optimization model of the displacement platform and can perform motion planning according to the target displacement given by the user.

2. The actuator part according to claim 1, characterized in that: The three-axis displacement devices are independent of each other and can work independently. Each axial device is fixed by screws, and the three-axis motion does not interfere with each other.

3. The micro slide rail according to claim 1, characterized in that: The utility model comprises a slider, a slide rail and a ball bearing. The slider and the slide rail are both provided with screw holes for fixing, and the slider and the slide rail are contacted and slid by the ball bearing.

4. The housing according to claim 1, characterized in that: For each axial displacement device, the housing is divided into two parts, one part is fixed to the slider of the micro slide rail, and the other part is fixed to the slide rail of the micro slide rail. The slider part can produce an adsorption effect with the magnet.

5. The micro magnet according to claim 1, characterized in that: The micro magnets are magnetically attached to the slider without any frictional contact with the housing.

6. The piezoelectric ceramic stack according to claim 1, characterized in that: The two ends are fixed to the housing and the magnet respectively. The power supply circuit extends outside the housing and is connected to the drive circuit.

7. The driving circuit according to claim 1, characterized in that: It has digital-to-analog conversion and power amplification functions. It can receive the eight-bit binary information sent by the microcontroller and convert it into analog signals respectively. Then, the signal is amplified to a suitable size to drive the piezoelectric ceramic stack and transmitted to the driving device.

8. The single chip microcomputer according to claim 1, characterized in that: It can output sawtooth waves of corresponding size, period and quantity according to the requirements. The output voltage size corresponds to the gear position.

9. The motion planning program according to claim 1, characterized in that: The genetic algorithm is used to mathematically solve the motion planning optimization model of the displacement platform, and the motion of the drive device is planned according to the user's target displacement output. The planning results will be input into the single-chip microcomputer.

10. The optimization model according to claim 9, characterized in that: The motion planning problem of the motion platform is optimized and mathematically modeled. For the average step length corresponding to different driving voltages, the number of steps is distributed. It is necessary to optimize the output displacement of the driving platform under the premise of considering the average distribution of the number of steps, so that the difference between it and the target distance given by the user is minimized.

11. The genetic algorithm according to claim 9, characterized in that: The optimization problem of the output steps of the drive device in each gear is solved. The genetic algorithm includes initialization function, iteration function and optimization function.

12. The initialization function according to claim 10, characterized in that: The heritability, mutation rate and initial population of the iterative function are defined.

13. The iterative function according to claim 10, characterized in that: It includes genetic operation function, mutation operation function and population activity selection function. The iterative function needs to be run multiple times in a loop, and the activity of each population needs to be calculated after each iteration to determine the number of offspring generated in the next iteration.

14. The optimization function according to claim 10, characterized in that: The solution corresponding to the population with the largest number after the iteration is completed is selected as the solution to the optimization problem.

Images