A method for probabilistic compensation of feed axis repeated positioning error based on digital map

Abstract

A method for compensating the probability of repeated positioning errors of a feed axis based on a digital map is characterized in that firstly, a digital map of repeated positioning errors of the position is drawn based on a large number of random errors, and then the maximum probability point of the digital map is solved in combination with a dynamic optimization algorithm, and then an error probability threshold is set to control the possibility of compensation errors, and the size and direction of the final compensation value are output. Finally, the maximum probability point is used as the compensation starting point, compensation is started, and the compensated error is detected. Error data that does not meet the requirements will enter the probability statistics link again, and the digital map is redrawn and updated. The present invention effectively solves the problem that repeated positioning errors are easily damaged by repeated disassembly in the past, and builds a solution channel between random errors and compensation values.

Description

Technical Field

[0001] The invention relates to the field of numerical control machining, in particular to a method for suppressing repeated positioning errors of a feed axis, specifically to a method for compensating repeated positioning errors of a feed axis based on a digital map. Background Art

[0002] Precision machine tools are widely used in military and civilian industries, and can process a variety of precision parts such as boxes and discs. The feed shaft is the core transmission component of the machine tool, and the repeatability error of the feed shaft is one of the important indicators for evaluating machine tools. Precision machine tools require high speed, high precision and high repeatability. However, the randomness of the repeatability error of the linear axis determines that it is impossible to build a deterministic compensation model similar to the positioning error. How to reduce the repeatability error of the feed shaft is an urgent need for precision machining.

[0003] Previous studies have found that the lack of a model for repeatable positioning errors has led to a single means of suppressing repeatable positioning errors. Usually, the possible influencing factors of repeatable positioning errors (such as preload and assembly accuracy, etc.) and the magnitude of the influence of each factor are determined by experimental methods, and then adjustments are made by disassembling and assembling parts based on the experimental results. Adjusting the assembly sequence can reduce the repeatable positioning error to a certain extent, but as the number of times the machine tool is used increases, the various parts of the machine tool will change, and the parts must be disassembled and assembled again to adapt to the new changes. This repeated assembly method is blind and cannot fully grasp the distribution law of errors, and repeated disassembly and assembly are very likely to damage parts. How to establish a compensation model that can be applied to the repeatable positioning error of the feed axis is a difficulty in current research. How to get rid of this method of suppressing repeatable positioning errors through repeated assembly is an urgent need for actual processing. Summary of the invention

[0004] The purpose of the present invention is to address the problems of existing methods for suppressing feed axis repetitive positioning errors, such as easy damage to parts, poor adaptability to actual working condition changes and high labor intensity. A method for compensating for the probability of feed axis repetitive positioning errors based on digital maps is invented, which constructs a feed axis repetitive positioning error probability compensation model from three aspects: digital map drawing, dynamic target search and compensation range control. On the basis of the digital map, combined with the dynamic optimization algorithm, the compensation points of each position of the feed axis are obtained. The size and direction of the compensation value are determined by setting the error probability threshold and the compensation range, and finally the error of the compensation error is extracted through error detection, and the original error is reapplied to the digital map drawing.

[0005] The technical solution of the present invention is:

[0006] A method for compensating the repeat positioning error probability of a feed axis based on a digital map, which draws a digital map by using the statistics of the repeat positioning error, searches for the highest point of the map in combination with a dynamic optimization algorithm, takes the coordinates of the highest point of the map as the compensation starting point, and finally solves the size and direction of the compensation value by controlling the probability of the compensation error, thus getting rid of the traditional method of suppressing the repeat positioning error by repeatedly disassembling parts, and is characterized in that it includes the following steps:

[0007] Step 1: Measure the repeated positioning error of a certain position of the feed axis m times, and obtain the error value of each measurement. The larger the m, the more helpful it is to the accuracy of the compensation model.

[0008] Step 2: Count the frequency of each error value in the positive stroke and negative stroke respectively, correspond each error value to its occurrence frequency, and further obtain the error probability distribution of the positive stroke and negative stroke;

[0009] Step 3: Construct a digital map of the location. The x-axis represents the positive travel error value of the location, the y-axis represents the negative travel error value of the location, and the z-axis represents the probability of taking the error value. A negative value indicates that the actual movement does not reach the ideal point, a positive value indicates that the actual movement exceeds the ideal point, and 0 indicates that the actual movement just reaches the ideal point;

[0010] Step 4: Repeat steps 1 to 3 for n times to obtain a digital map of n points on the feed axis, where the larger the n, the more helpful it is for the integrity of the compensation of the repeated positioning error of the feed axis;

[0011] Step 5: Initialize the parameters of the sensitive particle swarm algorithm, particle swarm size N, particle dimension d, inertia weight W, learning factor c, and number of sensitive particles M;

[0012] Step 6: Construct the fitness function of the dynamic particle swarm algorithm as:

[0013] fitness(x,y)=h(x,y)

[0014] Where h(x,y) is the height of the digital map of the repeated positioning error;

[0015] Step 7: Call all the digital maps of positions under a certain working condition, use the dynamic particle swarm optimization algorithm to search for the highest point of each digital map, and set the coordinates of each highest point as the error compensation point of the corresponding position, and obtain n position compensation points on the feed axis under this working condition. The x coordinate represents the positive compensation point, and the y coordinate represents the negative compensation point.

[0016] Step 8: Set the error probability threshold and compensation range, and solve the compensation value size and direction. The compensation values ​​of n positions on the model together constitute the repeated positioning error probability compensation model under this working condition;

[0017] Step 9: Start compensation, compensate for the actual repeated positioning error, and record the error after compensation;

[0018] Step 10: Detect the compensated error, add the actual error that does not meet the requirements after compensation into the probability statistics link, and repeat steps 2 and 3;

[0019] Step 11: Build a digital map model library, aggregate the maps obtained for each location and each working condition in the model library, and re-add the digital map updated in step 10 to the model library to achieve dynamic update of the map.

[0020] The digital map construction needs to meet the following conditions, which ensure the integrity of the three-dimensional map and that the data outside the coordinate axis will not affect the search results:

[0021] p 1-3 <p1

[0022] p 1-3 <p3

[0023] Among them, p 1-3 represents the probability of error points outside the coordinate axis, and p1 and p3 represent the probability of error points on the coordinate axis.

[0024] The error probability distribution is a collection of all possible error values ​​of a positive stroke or a negative stroke at a certain position and their corresponding occurrence probabilities.

[0025] The error compensation point is an ideal error point after optimal search, and this point is used as a reference when calculating the compensation value. If this point is located at the origin of the coordinate system, the compensation command is not executed.

[0026] The error probability threshold is the probability of allowing error compensation errors, which is determined according to actual needs.

[0027] The compensation range described is the ideal error interval after compensation, which determines the repeated positioning error after compensation and is determined according to actual requirements.

[0028] The compensation value is a vector value, which includes magnitude and direction. The compensation value of a certain position is determined by jointly solving the error probability distribution of the position, the error compensation point, the error probability threshold and the compensation range.

[0029] The beneficial effects of the present invention are:

[0030] The present invention effectively solves the problem of parts damage caused by repeated disassembly to suppress repeated positioning errors, and builds a solution channel between random errors and compensation values. In order to obtain an accurate and comprehensive digital map, it is necessary to statistically analyze a large amount of repeated positioning error data and obtain the probability of each error value. In order to enable the model to adapt to changes in different working conditions, it is necessary to build a digital map model library, dynamically improve the digital map through real-time detection and feedback, and call different digital maps for different working conditions.

[0031] The present invention breaks away from the traditional method of suppressing repeated positioning errors by repeatedly disassembling parts, and has the advantages of high speed and high precision. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 Construct a flow chart for the repeatability error compensation model.

[0033] Figure 2 A process for building digital maps of repeatable positioning errors.

[0034] Figure 3 Repeated positioning error measurement and digital mapping for feed axes.

[0035] Figure 4 is the probability distribution of the forward motion error at position 240mm.

[0036] Figure 5 It is the probability distribution of negative motion error at position 240mm.

[0037] Figure 6 This is the flow chart of the dynamic particle swarm algorithm based on sensitive particles.

[0038] Figure 7 Schematic diagram of dynamic optimization based on sensitive particles.

[0039] Figure 8 Compensates the measured position for magnitude and direction.

[0040] Fig. 9 is the compensation point position and error probability.

[0041] Fig.10 Modeling and application of probabilistic compensation models.

[0042] Fig.11 Error comparison before and after application of the compensation model.

[0043] Fig.12 To compensate for the repeated positioning errors at each position before and after. DETAILED DESCRIPTION

[0044] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0045] like Figure 1-12 shown.

[0046] A method for compensating the repeated positioning error of a lead screw of a SiC substrate back grinding machine based on a digital map, comprising:

[0047] 1. Repeated positioning error digital mapping.

[0048] Drawing a digital map of repeated positioning errors is the key to the probability compensation model. Through a large amount of repeated positioning error data, the frequency of a certain error occurring at a certain position is found, and the probability of the occurrence of this error is further calculated. The digital map of repeated positioning errors is drawn using the statistically obtained probability distribution of the positive and negative direction errors of the feed axis. The drawing method is as follows: Figure 2 shown.

[0049] The process of building a digital map of repeated positioning errors has the following steps:

[0050] The first step is to arrange the laser interferometer according to the measurement position;

[0051] The second step is to move the feed axis to further drive the reflector;

[0052] The third step is to measure the feed axis error using the reflector position;

[0053] Step 4: Repeat steps 2 and 3 n times, and record the measurement error value of each time;

[0054] Step 5: According to the recorded measurement error values, the error distribution frequency of each position in the positive stroke and negative stroke is counted;

[0055] Step 6: According to the error distribution frequency of each position, the error distribution probability of each position is further calculated;

[0056] Step 7: Construct the coordinate axes of the digital map of the specific location. The x-axis coordinate represents the positive travel error value of the location, the y-axis represents the negative travel error value of the location, and the z-axis represents the probability of taking the error value;

[0057] Step 8: Draw a digital map of the repeated positioning error at each location.

[0058] In order to obtain a complete digital map, it is necessary to draw three-dimensional points outside the coordinate axis, but this point has no practical significance. In order to ensure that the data points outside the coordinate axis do not affect the search results, such as Figure 2 As shown, the following conditions should be met:

[0059] p 1-3 <p1

[0060] p 1-3 <p3

[0061] Among them, p1-3 represents the probability of error points outside the coordinate axis, and p1 and p3 represent the probability of error points on the coordinate axis.

[0062] The x-axis and y-axis represented by the positive and negative strokes intersect at the origin, and the corresponding z-coordinate value takes the maximum value of the x-axis and y-axis at the intersection. In the following analysis, it will be mentioned that the origin does not correspond to compensation, and the maximum probability of zero error in positive and negative strokes at the origin can reduce the probability of erroneous compensation.

[0063] like Figure 3 As shown in Figures 3(a) and 3(b), the object of digital mapping in this embodiment is the feed axis of a horizontal CNC machining center, and digital maps of 11 positions are drawn. Among them, the digital maps of the positions 240 mm and 640 mm are shown in Figure 3(a). Figure 3 (c) As shown. Figure 3 In (c), the x-axis and y-axis have negative values ​​indicating that the actual motion does not reach the ideal point, positive values ​​indicating that the actual motion exceeds the ideal point, and 0 indicating that the actual motion reaches the ideal point exactly. Figure 3 The z-axis in (c) represents the probability of taking the error value. Comparing the two digital maps, it can be seen that the map morphology at different locations is quite different, which may be a single peak morphology or a multi-peak morphology. These situations also reflect that the probability distribution of repeated positioning errors at each location is different.

[0064] The digital map at 240 mm was extracted for further analysis. Figure 4 and 5 As shown in the figure, there are obvious differences in the probability distribution of positive stroke error and negative stroke error at the same position, and the positions of maximum probability are also different. This shows that it is necessary to place the positive stroke and negative stroke repeated positioning errors on different coordinate axes, which ensures that the probability distribution of errors in the two directions will not affect each other.

[0065] The map appearance is determined by the error probability of each position. Draw a digital map of the repeated positioning error of each position of the feed axis to form a map database for subsequent calls.

[0066] 2. Construct a probability compensation model for repeated positioning errors.

[0067] In actual processing, the position of the feed axis is constantly changing, and the digital maps of different positions are also different. This requires the search environment to be constantly changing, and the search target value is also constantly changing. It can be seen that this is a dynamic optimization problem. This embodiment solves this problem by using a dynamic particle swarm optimization algorithm based on sensitive particles. The process of the dynamic particle swarm algorithm based on sensitive particles is as follows: Figure 6 shown.

[0068] The fitness function of the dynamic particle swarm algorithm is:

[0069] fitness(x,y)=h(x,y)

[0070] Among them, h(x,y) is the height of the digital map of repeated positioning error.

[0071] Combined with the digital map obtained in the previous part, the search process of the dynamic optimization algorithm is described in detail, such as Figure 7 shown. Figure 7 The change curve in the figure describes the sum of the fitness of sensitive particles at each position of the positive stroke of the feed axis. In each iteration step, the x-axis and y-axis coordinates of the sensitive particles remain unchanged, and the sum of the z-axis coordinates of all sensitive particles (i.e., the fitness sum) is used to determine whether the map has changed. If the map has not changed, the search process is skipped, and the optimal value is the last search result. If the map has changed, the ordinary particles are reinitialized, and the optimal value is obtained according to the standard particle swarm search process. All particles approach the particles with high fitness until all particles reach the optimal position.

[0072] The digital map of repeated positioning errors at 11 positions of the feed axis is used in combination with the dynamic particle swarm algorithm to obtain the compensation points (i.e., the maximum probability points or optimal value points) of the positive and negative strokes of each position of the feed axis, such as Fig. 9 (a) shown.

[0073] Before compensation, the size of the compensation value needs to be set, and the size of the compensation value is determined by the compensation target. The compensation target of this embodiment is that the repeated positioning error after compensation is between ±1μm. Since this embodiment provides a probabilistic compensation model, there is a probability of compensation error: the compensation direction is opposite or the error after compensation exceeds the compensation target. Therefore, it is necessary to control the probability of compensation error. The model of this embodiment controls the error probability to be less than 20%. Combined with Figure 4 , Figure 5 and Fig. 9 (a), the specific process of determining the compensation value is as follows. The measurement data of each position can obtain the error probability distribution of positive and negative strokes. In order to concisely and clearly explain the compensation method, only some position error probability distributions are listed.

[0074] Position 0mm: The positive and negative compensation points are 0μm, no compensation.

[0075] Position 80mm: The positive and negative compensation points are 0μm, no compensation.

[0076] Position 160mm: positive compensation point -1μm, negative compensation point 1μm. Positive compensation 1μm, error rate 4%. Negative compensation 1μm, error rate 48%, so no compensation in the negative direction. Difference probability distribution is shown in Table 1 and Table 2.

[0077] Table 1 Probability distribution of positive repeat positioning error at position 160 mm

[0078]

[0079] Table 2 Probability distribution of negative repeat positioning error at position 160 mm

[0080]

[0081] Position 240mm: Positive compensation point -1μm, positive direction compensation 1μm, error rate 0. Negative compensation point -1μm, negative direction compensation 1μm, error rate 18%.

[0082] Position 320mm: positive compensation point -3μm, if the positive direction is compensated directly by 3μm, the error rate is 40%, and the error rate of compensating 2μm is 20%, so the error rate of compensating 1μm is 4%. According to the probability threshold of excess tolerance, the positive compensation 1μm command is executed. The negative compensation point is 0μm, so no compensation is made in the negative direction. The distribution of positive errors is shown in Table 3.

[0083] Table 3 Probability distribution of positive repeat positioning error at position 320 mm

[0084]

[0085] Position 400mm: Positive compensation point -2μm, positive direction compensation 2μm, error rate 20%. Negative compensation point -1μm, negative direction compensation 1μm, error rate 4%.

[0086] Position 480mm: Positive compensation point -1μm, positive direction compensation 1μm, error rate 16%. Negative compensation point -1μm, negative direction compensation 1μm error rate 44%, so no compensation in the negative direction.

[0087] Position 560mm: Positive compensation point -1μm, positive direction compensation 1μm, error rate 18%. Negative compensation point -1μm, negative direction compensation 1μm, error rate 18%.

[0088] Position 640mm: Positive compensation point -3μm, the error rate of positive compensation 3μm is 40%, the error rate of compensation 2μm is 20%, so the compensation is 2μm. Negative compensation point -1μm, negative compensation 1μm, the error rate is 14%.

[0089] Position 720mm: positive compensation point -1μm, positive direction compensation 1μm, error rate 4%. Negative compensation point -2μm, negative direction compensation 2μm, error rate 16%.

[0090] Position 800mm: Positive compensation point -2μm, positive compensation 2μm error rate 24%, compensation 1μm error rate 5%, so compensation 1μm. Negative compensation point -3μm, negative compensation 3μm error rate 48%, compensation 2μm error rate 28%, compensation 1μm error rate 12%, so compensation 1μm.

[0091] Through the above analysis, the compensation direction and compensation value of each position are as follows Figure 8 The probability of compensation deviation at each position is statistically analyzed as shown in Fig. 9 (b) as shown.

[0092] 3. Application of probability compensation model.

[0093] Through the above analysis, a probability compensation model under specific working conditions is obtained, which includes the size and direction of the compensation value. However, in actual application, the working conditions are constantly changing, that is, the error probability distribution of each position is constantly changing, which requires a digital map model library containing all working conditions. In addition, different processing requirements have different requirements for the range of repeated positioning errors, which requires the model to be able to adapt to the adjustment of the error probability threshold and error range. Combined with the previous analysis, a method for self-learning modeling of a repeated positioning error probability compensation model is provided, such as Fig.10 As shown, the so-called “self-learning” means that new repeated positioning errors can participate in digital map drawing and update the digital map model library in real time.

[0094] In order to verify the actual application effect of the model, the probability compensation model obtained above is tested by another five sets of repeated positioning error data. The first set of data used for error detection is shown in Table 4. The first set of test data is shown in Table 4.

[0095] Table 4 The first set of model detection data

[0096]

[0097] Will Figure 8 The compensation data shown is applied to the error data in the table, and the compensation process and results are shown below.

[0098] Positive stroke error compensation process:

[0099] Position 0mm: actual error 0μm, compensation 0μm;

[0100] Position 80mm: actual error 0μm, compensation 0μm;

[0101] Position 160mm: actual error -2μm, positive direction compensation 1μm, error reduction 1μm, error after compensation -1μm;

[0102] Position 240mm: actual error -1μm, positive direction compensation 1μm, error reduction 1μm, error after compensation 0μm;

[0103] Position 320mm: actual error -3μm, positive direction compensation 1μm, error reduction 1μm, error after compensation -2μm;

[0104] Position 400mm: actual error -3μm, positive direction compensation 2μm, error reduction 2μm, error after compensation -1μm;

[0105] Position 480mm: actual error -1μm, positive direction compensation 2μm, error reduction 1μm, error after compensation 0μm;

[0106] Position 560mm: actual error -2μm, positive direction compensation 1μm, error reduction 1μm, error after compensation -1μm;

[0107] Position 640mm: actual error -3μm, positive direction compensation 2μm, error reduction 2μm, error after compensation -1μm;

[0108] Position 720mm: actual error -3μm, positive direction compensation 1μm, reduced error 1μm, error after compensation -2μm, exceeding the maximum preset error;

[0109] Position 800mm: actual error -2μm, positive direction compensation 1μm, reduced error 1μm, error after compensation -1μm. Negative stroke error compensation process:

[0110] Position 0mm: actual error 0μm, compensation 0μm;

[0111] Position 80mm: actual error 0μm, compensation 0μm;

[0112] Position 160mm: actual error 1μm, compensation 0μm;

[0113] Position 240mm: actual error 1μm, negative direction compensation 1μm, additional error 1μm, error after compensation 2μm, exceeding the maximum preset error;

[0114] Position 320mm: actual error 0μm, compensation 0μm;

[0115] Position 400mm: actual error -1μm, compensation in negative direction 1μm, error reduction 1μm, error after compensation 0μm;

[0116] Position 480mm: actual error 0μm, compensation 0μm;

[0117] Position 560mm: actual error 0μm, negative direction compensation 1μm, additional error 1μm, error after compensation 1μm;

[0118] Position 640mm: actual error -1μm, compensation in negative direction 1μm, error reduction 1μm, error after compensation 0μm;

[0119] Position 720mm: actual error -2μm, negative direction compensation 2μm, error reduction 2μm, error after compensation 0μm;

[0120] Position 800mm: actual error -2μm, compensation in negative direction 1μm, error reduction 1μm, error after compensation -1μm;

[0121] Based on the above analysis, a set of positive and negative strokes has a total of 22 measurement data, and the compensation command was started 15 times. The original data produced a total error of 28μm, and the error was reduced by 16μm after compensation, accounting for 57% of the error reduction. The error increased by 2μm after compensation, accounting for 7% of the error increase. Among them, the errors increased after compensation were 240mm and 560mm at the negative stroke position, and the positive stroke position exceeded the maximum preset error after compensation at 720mm and 240mm at the negative stroke position. Fig. 9 (b) It can be seen that the negative stroke position 240mm and the position 560mm are both positions with high probability of compensation out-of-tolerance.

[0122] In order to briefly describe the compensation process, only the first set of test data and compensation process are shown, and the other four sets of test data and compensation processes are omitted, but the methods involved are the same. The remaining four sets of measurement data are analyzed according to the above method, and the compensation results are as follows Fig.11 shown.

[0123] from Fig.11 It can be seen that the error fluctuation is significantly reduced after compensation. The remaining four groups of experiments have a total of 88 data sets, and the compensation command is started 60 times. The original data has a total error of 121μm. After compensation, the error is reduced by 114μm, accounting for 94% of the error reduction, and the error is increased by 4μm, accounting for 3%. The error of 3μm remains unchanged but changes direction. After compensation, a total of 6 groups of data exceed the maximum preset error, and the excess rate accounts for 7%.

[0124] According to the definition of calculating linear axis repeatability error in international standard ISO230-2, the repeatability error before and after compensation is calculated by using the above five sets of measurement data and compensated data. The calculation process is shown below.

[0125] The feed axis moves back and forth five times in a certain direction, and the jth movement reaches P i The error at X i,j for:

[0126] X i,j =Pi,j -P i

[0127] Among them, P i，j is the actual position reached by the feed axis, P i The target position of the feed axis.

[0128] Feed axis in P i The average position deviation X ave_i for:

[0129]

[0130] Position P i Reverse clearance B i for:

[0131] B i =X ave_i ↑-X ave_i ↓

[0132] The symbol ↑ represents positive movement, and ↓ represents negative movement.

[0133] Further obtain the position P i The one-way standard uncertainty S i for:

[0134]

[0135]

[0136] Position P i One-way repeat positioning error R i for:

[0137] R i ↑=4S i ↑

[0138] R i ↓=4S i ↓

[0139] Finally, we get the position P i The bidirectional repeat positioning error R at is:

[0140] R = max {R i ↑,R i ↓,2S i ↑+2S i ↓+|B i |}

[0141] The repeated positioning error of each point of the feed axis before compensation is obtained as follows: Fig.12 shown.

[0142] pass Fig.12It can be seen that the overall repeatability error after compensation is lower than that before compensation. The compensation effects at different positions are also significantly different. The compensation effects at positions 160mm, 480mm, and 640mm are the most obvious, reducing the repeatability error by more than 1μm. The main reason for the differences in compensation at different positions is that the error digital maps obtained at different positions are different, and the control of the maximum preset error and the probability of exceeding the tolerance result in different compensation amplitudes and directions at different positions.

[0143] The present invention does not involve parts that are the same as the prior art or can be implemented by using the prior art.

Claims

1. A method for compensating the repeated positioning error probability of a feed axis based on a digital map, which draws a digital map by statistically analyzing the repeated positioning error, searches for the highest point of the map in combination with a dynamic optimization algorithm, takes the coordinates of the highest point of the map as the starting point of compensation, and finally solves the size and direction of the compensation value by controlling the probability of compensation error, characterized in that It includes the following steps: Step 1: Measure the repeated positioning error of a certain position of the feed axis m times and obtain the error value of each measurement; Step 2: Count the frequency of each error value of the positive stroke and the negative stroke respectively, and further obtain the error probability distribution of the positive stroke and the negative stroke; Step 3: Construct a digital map of the location; wherein the x-axis coordinate represents the positive travel error value of the location, the y-axis represents the negative travel error value of the location, and the z-axis represents the probability of taking the error value; Step 4: Repeat steps 1 to 3 for n times to obtain a digital map of n points on the feed axis; Step 5: Initialize the parameters of the sensitive particle swarm algorithm, particle swarm size N, particle dimension d, inertia weight W, learning factor c, and number of sensitive particles M; Step 6: Construct the fitness function of the dynamic particle swarm algorithm as: fitness(x,y)=h(x,y) Where h(x,y) is the height of the digital map of the repeated positioning error; Step 7: Call all the digital maps of the locations, use the dynamic particle swarm optimization algorithm to search for the highest point of each digital map, and set the coordinates of each highest point as the error compensation point of the corresponding location; Step 8: Set the error probability threshold and compensation range, solve the compensation value size and direction, and build the error probability compensation model; Step 9: Start the compensation command to compensate for the actual repeated positioning error; Step 10: Detect the compensated error, add the actual error that does not meet the requirements after compensation into the probability statistics link, and repeat steps 2 and 3; Step 11: Build a digital map model library, aggregate the maps obtained for each location and each working condition in the model library, and re-add the digital map updated in step 10 to the model library.

2. The method according to claim 1, characterized in that The digital map construction needs to meet the following conditions: p 1-3 <p1 p 1-3 <p3 Among them, p 1-3 represents the probability of error points outside the coordinate axis, and p1 and p3 represent the probability of error points on the coordinate axis.

3. The method according to claim 1, characterized in that The error probability distribution is a collection of all possible error values ​​of a positive stroke or a negative stroke at a certain position and their corresponding occurrence probabilities.

4. The method according to claim 1, characterized in that: The error compensation point is an ideal error point after optimal search, and this point is used as a reference when calculating the compensation value. If this point is located at the origin of the coordinate system, the compensation command is not executed.

5. The method according to claim 1, characterized in that The error probability threshold is the probability of allowing error compensation errors, which is determined according to actual needs.

6. The method according to claim 1, characterized in that The compensation range described is the ideal error interval after compensation, which determines the repeated positioning error after compensation and is determined according to actual requirements.

7. The method according to claim 1, characterized in that The compensation value is a vector value, which includes magnitude and direction. The compensation value of a certain position is determined by jointly solving the error probability distribution of the position, the error compensation point, the error probability threshold and the compensation range.

Images