TWI667559B - Automatic surface error compensation method and computer program product therefor - Google Patents

Abstract

An automatic surface error compensation method suitable for controlling according to a processing program The machine tool automates the machining of the workpiece by creating a plurality of imaginary planes on an ideal surface to obtain a plurality of curve features. These curve features are a collection of curves where each imaginary plane intersects the ideal surface. According to the curve feature, the layout calculation is performed to obtain the position information of the ideal measurement point. Then, the on-line surface measurement is directly performed on the workpiece on the machine tool to obtain the actual position information of the actual measurement point. Perform error calculations on actual measurement points and multiple ideal measurement points corresponding to the ideal surface to perform surface fitting and produce a finishing path. A computer program product having at least one code is provided, and the method can be completed when the electronic device loads and executes the code.

Description

Automatic surface error compensation method and computer program product thereof

The invention relates to an error compensation method and a computer program product thereof, and in particular to an automatic surface error compensation method and a computer program product suitable for performing the automatic surface error compensation method.

At present, when the workpiece is processed by the industrial industry and the size of the workpiece is found to be in error, the workpiece is generally subjected to wear compensation by a computer numerical control (CNC) controller. That is to say, according to the measurement results, the workpiece is gradually compensated, and it is usually necessary to perform two to three uniform compensations to obtain the optimum value. In addition, the conventional CNC controller compensation method is limited to the compensation of simple shapes and simple surfaces, and cannot compensate for complex free-form surfaces.

Another approach is to manually modify the NC file to adjust the machining path, or use an external program such as Computer-Aided Inspection. Planning; CAIP) software to perform the layout operation to obtain a more precise processing size. However, manually modifying the NC file is time consuming and difficult to complete. In addition, when using an external program, since the steps of machining, measuring, and compensating the workpiece are not performed on the same system, there are many unnecessary troubles in combination with industrial software automation, such as different parameter settings or coordinate systems. Conversion, positioning error, software compatibility, etc.

As surface applications become more widespread, the market's demand for continuous curved surface processing products is increasing, and the requirements for workpiece dimensional accuracy are also increasing. In addition, in response to the development of Industry 4.0, automated processing is an inevitable trend. Therefore, how to complete the steps of machining, measuring and compensating the workpiece on the same system to simplify the process and improve the dimensional accuracy of the workpiece has become a problem that researchers in this field are eager to solve.

The invention provides an automatic surface error compensation method and a computer program product, which can complete the automatic surface error compensation method, can effectively combine the computer-aided software development automation, and can improve the dimensional accuracy of the processed workpiece.

The automated surface error compensation method of the present invention is suitable for automatically processing workpieces according to a machining program controlled machine tool. The automated surface error compensation method includes creating a plurality of imaginary planes on an ideal surface to obtain a plurality of curve features, the curve features being a set of curves where each imaginary plane intersects the ideal surface. According to the curve feature, the layout calculation is performed to obtain the position information of the ideal measurement point. Then, the workpiece is directly measured on the surface of the machine to obtain the actual position information of the actual measurement point. Perform error calculation on the actual measurement points and the ideal measurement points on the ideal surface And performing mirror compensation to perform surface fitting and generating a finishing path, wherein performing the error calculation comprises projecting the actual measuring point onto a normal direction of the ideal surface, and then performing mirror compensation to obtain a compensation point .

In an embodiment of the invention, the performing the error calculation and performing the mirror compensation further comprises: obtaining the actual measurement point according to the normal vector measurement of the ideal measurement point, and connecting the theoretical point with the actual measurement point to obtain the connection. And the ideal measuring point is extended along the normal direction, and the angle between the connecting line and the extending line or the shortest distance from the ideal measuring point to the normal of the actual measuring point is obtained. If the angle or the shortest distance is not 0, it means actual. When measuring the point on the illegal vector, the actual measurement point is projected onto the ideal surface, and the normal direction of the actual measurement point intersects the ideal surface to obtain the projection point, and the mirror direction compensation is performed by the normal direction to obtain compensation. point.

In an embodiment of the invention, the surface fitting is performed according to the compensation point, and the finishing path is the fitted curved surface.

In an embodiment of the invention, after performing the dot calculation, the workpiece is roughed and machined, and after the middle machining, the wire surface measurement is performed on the workpiece.

In an embodiment of the invention, after the on-line surface measurement is performed, the actual measurement point position is added to the CAD model, and the error calculation, the mirror compensation and the surface fitting are performed to form a finishing path, and the automatic surface error compensation method is performed. It also includes replacing the ideal surface with a fitted surface to remodel after the surface fit.

In an embodiment of the invention, after remodeling is performed, the processing in the copy is performed to maintain the same processing conditions and recalculated to produce a finishing path.

In an embodiment of the invention, performing the dot calculus according to the curve feature includes determining a curvature change of the curve to determine the dot density.

In an embodiment of the invention, after performing the layout calculation, the processing program generates a measurement path according to the result of the layout calculation to control the machine tool to perform on-line surface measurement on the workpiece.

In an embodiment of the invention, after the machine tool performs on-line measurement on the workpiece, the measurement report is automatically output from the controller end to the local end, so that the processing program automatically reads the information of the actual measurement point, Perform error calculations.

The computer program product of the present invention has at least one code, and the automatic surface error compensation method can be completed when the electronic device loads and executes the code.

Based on the above, the automatic surface error compensation method of the present invention can perform the steps of processing, measuring and compensating the workpiece by a single computer program, and the traditional technology must rely on different software to perform the above steps, thereby causing time-consuming parameter setting or coordinates. Problems such as positioning errors occur on the conversion. In addition, the automated surface error compensation method of the present invention can effectively integrate and manage the system in response to the development of Industry 4.0.

The above described features and advantages of the invention will be apparent from the following description.

10‧‧‧Workpiece

10a‧‧‧Ideal surface

100, 200, 300‧‧‧ method

110, 120, 130, 140, 201~217, 301~314‧‧‧ steps

400‧‧‧ system

402‧‧‧Processor

403‧‧‧Internet interface

404‧‧‧Storage media

406, 407‧‧‧ code

408‧‧‧Measurement information

Part A, B, C‧‧‧

CL‧‧‧ connection

EL‧‧‧ extension line

CP‧‧‧ Compensation point

HN‧‧‧ ideal measuring point normal vector

HP‧‧‧ ideal measurement point

HP’‧‧‧ projection point

HS‧‧‧ ideal surface

MN‧‧‧ normal direction

MP‧‧‧ actual measurement point

MS‧‧‧Processed workpiece surface

P‧‧‧imaginary plane

P ₀ ‧‧‧ ideal point

SF‧‧‧ curve features

1 is a flow chart of an automated surface error compensation method in accordance with an embodiment of the present invention.

2 is a schematic diagram of a CAD model of imaginary plane and curve features established on a workpiece in accordance with an embodiment of the present invention.

3 is a schematic diagram of a CAD model of a point surface curved surface ideally measured on a workpiece in accordance with an embodiment of the present invention.

4A is a schematic diagram of the normal concept of an ideal measurement point and an actual measurement point.

4B is a schematic diagram of the normal concept of an ideal measurement point and an actual measurement point in accordance with an embodiment of the present invention.

5 is a flow chart of an automated surface error compensation method in accordance with another embodiment of the present invention.

6 is a flow chart of an automated surface error compensation method in accordance with another embodiment of the present invention.

7 is a block diagram of a computer program product in accordance with an embodiment of the present invention.

1 is a flow chart of an automatic surface error compensation method according to an embodiment of the present invention, and FIG. 2 is a schematic diagram of a curved CAD model of an imaginary plane and a curved feature established on a workpiece according to an embodiment of the present invention, FIG. 3 is a schematic diagram of FIG. FIG. 4A is a schematic diagram of a normal point measurement point and a normal measurement point normal direction concept according to an embodiment of the present invention, and FIG. 4B is an implementation according to the present invention. FIG. A schematic diagram of the normal measurement point of an ideal measurement point and an actual measurement point in an example. Referring to FIG. 1 to FIG. 4B, the automatic curved surface error compensation method 100 is adapted to control a machine tool according to a machining program (for example, a five-axis machining machine or other suitable CNC machine tool). The workpiece 10 is automatically measured and processed, and the processing method may include a machine tool and a machining system that is electrically connected to the machine tool and used to control the operation of the machine tool. The host computer can also store position information, processing methods, measurement information, etc. about the workpiece to be processed. The automated surface error compensation method 100 includes at least steps 110 to 140. The steps are performed, for example, in a single computer program. The implementation may be a computer program for processing the workpiece, and executed by a computer host. Thereafter, the power tool is controlled to perform the functions of the embodiments of the present invention, and the various steps of the automated curved surface error compensation method 100 are detailed below.

At step 110, a plurality of imaginary planes P are created on an ideal curved surface 10a of the workpiece 10 to obtain a plurality of curved features SF, as shown in FIG. 2A. The curve feature SF is, for example, a set of curves in which each of these imaginary planes P intersects the ideal curved surface 10a. In some embodiments of complex workpieces, the ideal curved surface 10a can be a complex freeform surface. For example, the ideal curved surface 10a may include a plurality of portions (A, B, C as shown in FIG. 2). Part A is, for example, a free-form surface in which the UV curve of the quadratic curve changes constantly, and part B is, for example, a free-form surface swept by a quadratic curve (concave) and a quadratic curve (convex), and part C is, for example, scanned by two quadratic curves. A free-form surface with two guiding curves and a constant UV direction. The imaginary plane P can be established to divide the ideal curved surface 10a according to the direction in which the machine tool cuts the workpiece 10.

Next, in step 120, the layout calculation is performed according to the curve feature SF to obtain the position information of the ideal measurement point. For example, a curve intersecting the imaginary plane P with the ideal curved surface 10a is subjected to a UV direction (for example, in the same direction as the cutting direction), and the points on these curved surfaces are also referred to as curve features SF. Discriminating the curvature of the ideal surface 10a The step of the line feature SF includes curve fitting the ideal measurement points and calculating the tangent slope of the ideal measurement points and the fitted curve (or the curvature of the curve feature SF). When the slope between two adjacent points changes too much or the slope changes from positive to negative, it is determined that the block between the two points is the main block that may cause inaccurate processing and poor fitting effect, and These major blocks increase the number of distribution points. On the other hand, in a block with a small or constant change in slope between two points, the distribution point can be loose, so that the measurement time is not too long, thereby adjusting the number of dots.

Next, in step 130, the on-line surface measurement is performed on the workpiece 10 without removing the workpiece 10, so as to obtain actual position information of the plurality of actual measurement points, and the position information of the actual measurement point is established. The CAD model is shown in Figure 3. Since the measurement is performed without disassembling the workpiece 10, the problem of positioning error caused by mounting the workpiece 10 after disassembly can be avoided. Then, in step 140, error calculation is performed on the actual measurement points and the plurality of ideal measurement points corresponding to the ideal curved surface and mirror compensation is performed to perform surface fitting and generate a finishing path, wherein the error calculation is performed and mirrored The shot compensation can include projecting the actual measurement point onto the normal direction of the ideal surface, and then performing mirror compensation to obtain the compensation point. Surface fitting is performed based on these compensation points and a finishing path is generated.

In general, the path of the finished tool path can be calculated by the principle of mirroring. The principle of mirroring is usually based on the assumption that the processed surface is approximately the same plane as the ideally machined surface, and the measurement direction will be closer to the ideal measurement point by the normal direction. Therefore, the actual measurement point and the ideal measurement are generally determined. The point is on the same normal line, but it is not. The reason is that measurement may be due to measurement error, machine processing error and processing. The complexity of the back surface affects the relationship between the fitted surface and the ideal surface, not just the simple offset, that is, the actual measurement point is not in the normal direction of the ideal measurement point. Specifically, referring to FIG. 4A, if the surface MS of the processed workpiece is a complex free-form surface, the actual measurement point MP is not an ideal measurement point on the ideal surface HS of the CAD model when performing on-line surface measurement. A compensation point CP at the normal direction HN of the HP that intersects the surface MS of the processed workpiece.

The error calculation method proposed by the invention is that the actual measurement point is projected onto the normal direction of the ideal surface, and then the mirror compensation is performed to obtain the compensation point. Please refer to FIG. 4B for the ideal concept of the modified ideal measurement point and the actual measurement point as the normal vector. First, the actual measurement point MP is obtained according to the ideal measurement point normal vector HN, and the two points (theoretical point P ₀ and The actual measurement point MP) is connected to obtain the two-point connection CL, and an ideal measurement point HP is established along the normal direction to form an extension line EL, and the angle between the two lines (the connection line CL and the extension line EL) is also obtained. It is the shortest distance from the ideal measurement point HP to the actual measurement point MP normal. If it does not coincide, it indicates the point on the actual measurement point MP illegal vector HN, so the actual measurement point MP is projected onto the ideal surface of the CAD model. On the HS, the normal direction MN of the actual measurement point MP intersects with the ideal curved surface HS to obtain the projection point HP', and the exact compensation point is obtained by mirror compensation by the normal direction MN. Therefore, for the workpiece model of the complex free-form surface, the finishing path generated by the above-mentioned modified mirror compensation method can reduce the error caused by the measurement, thereby improving the accuracy of the workpiece size.

The above steps 110 to 140 can be performed by a single processing program, thereby avoiding the separate execution of the workpiece by different systems and programs in the conventional method. Problems such as parameter setting conversion, positioning error, and software compatibility caused by processing, measurement, and compensation. In addition, the automated surface error compensation method 100 can process workpieces with complex free-form surfaces, which can improve the accuracy of the dimensions compared with the conventional method.

FIG. 5 is a flow chart of an automated surface error compensation method according to another embodiment of the present invention. FIG. 5 is a flow chart of the automatic curved surface error compensation method, and FIG. 5 is a flowchart showing a more complete technical framework of an automatic curved surface error compensation method. . Referring to FIG. 5, the automated surface error compensation method 200 includes at least steps 201 to 217.

In step 201 and step 202, the workpiece blank is supplied to the machine tool, and the rough and medium processing methods are automatically generated by the processing program. At step 203, the machining program automatically performs a spot calculation for the freeform surface. For example, the method described in step 110 and step 120 of FIG. 1 can be used to perform the layout calculation, the imaginary plane is established according to the cutting direction to divide the surface, and then the curve feature intersecting the imaginary plane and the curved surface is performed. Imagine the layout and determine the density of the dots by judging the curvature of the curve. After performing the layout calculation, the machining program generates a measurement path based on the result of the layout calculation, and generates an NC program through the post-measurement processing until the stage can be completed on the same software program. For example, in step 204, the machining program generates a measurement method based on the result of the layout calculation. Subsequently, in steps 205 and 206, the machining program controls the machine tool for roughing and machining. In the roughing stage, roughing can be performed using a tool with a larger tool diameter to save machining time. The same tool must be used during the in-process and subsequent finishing stages.

Next, for example, in step 207, after the in-process machining, the machining program controls the machine tool to perform an in-line surface measurement on the workpiece that has been machined in accordance with the result of the machining. At step 208, the report with the actual measurement point location information is output to a specific address of the host computer. In step 209, the plugin automatically reads the report to import the actual measurement point data into the CAD model, thereby calculating the deviation value of the surface after machining. For example, the measurement program can be executed on the workpiece by the machining program control tool machine without removing the workpiece to avoid repositioning errors. After the online measurement of the workpiece is completed, the actual measurement point position information is automatically stored, and then transmitted to a specific address of the local end through the network, and the position information of the actual measurement point is automatically read through the processing program, and processed through the processing. The program builds the actual measurement points directly on the CAD model.

After the machine tool performs on-line measurement on the workpiece, the error calculation can be performed by the machining program. For example, in step 210, the normal error calculation between the actual measurement point and the ideal measurement point is performed. At this stage, the error of the dimension processing (for example, machine error, machining error, etc.) can be known based on the measured position information of the point. Next, in step 211, mirror compensation is performed by the plugin. For example, the method of correcting the mirror compensation described in the step 140 of FIG. 1 and the embodiment of FIG. 4B is performed, and details are not described herein again.

Next, in step 212, the processing program performs surface fitting according to the compensation points generated after the mirror compensation calculation. For example, a non-uniform rational B-Splines (NURBS) algorithm can be used for surface fitting. At step 213, after the surface is fitted, the machining program replaces the CAD model with the fitted surface to remodel. At step 214, after performing the remodeling, The machining path of the machining process is recalculated to produce a finishing method.

Generally speaking, at this stage, the industry uses other programs (such as MATLAB) to perform surface fitting to generate a new fitted model, and then import this model into Computer Aided Design (CAD) software or Computer Aided Manufacturing (CAM) software performs secondary compensation processing, resulting in labor-intensive and time-consuming use of different software programs, which can cause inconvenience to users in both operation and management. In the embodiment of the present invention, the surface fitting can be directly performed on the CAD/CAM software by using the plug-in program, and after the surface fitting is completed, the original model can be directly modeled as a compensation surface to generate a finishing method. . Therefore, it is possible to perform the finishing without changing the tool and without disassembling the workpiece, and the machining path of the finishing is determined according to the fitted compensation surface. That is to say, the automatic surface error compensation method of the present invention can perform secondary development on the existing software, and the steps of processing, measuring and compensating the workpiece are completely integrated on a single software body, thereby realizing the purpose of automatic production. .

After the finishing is completed, the finished product can be selectively measured by a three-dimensional measurement, for example, by a Coordinate Measuring Machine (CMM) to determine whether the finished product is in the finished product. Within the specification tolerances, as in steps 216 and 217. If the judgment result falls within the specification tolerance, the finished product is a qualified finished product, and if the judgment result does not fall within the specification tolerance, the finished product is a defective product.

6 is a flow chart of an automated surface error compensation method in accordance with another embodiment of the present invention. 1 and 5, the flowchart of FIG. 6 is for example The invention discloses an automated surface error compensation method applied to a Manufacturing Execution System (MES), thereby realizing the smart manufacturing requirements of Industry 4.0.

Referring to FIG. 6, the automated surface error compensation method 300 applied to the MES may include steps 301 to 314. Specifically, the MES can be used to perform the task. First, in steps 301 and 302, the CAD model is automatically imported into the CAD/CAM software, and the coarse and medium processing methods are automatically generated. Next, in steps 303 and 304, the curve features can be automatically determined by the methods described in steps 110 and 120 of FIG. 1, and the layout calculation is performed by an internal program, and the details of the layout calculation are not described again. Next, after performing the dot calculation, in steps 305 and 306, the measurement method in the middle processing is automatically generated, and the program is automatically uploaded. The actual measurement can be performed according to this measurement method. After the rough and medium processing is completed, the machine can automatically change the probe to perform on-line measurement, as in step 307.

Then, in steps 308 and 309, the system can automatically output the measurement report and automatically read the report data. For example, the auto-generated measurement report can be stored in the controller through the subroutine of RENISHAW's measurement system, and the report is automatically output to the software local end, and then the corresponding folder is automatically read through the processing program. Report data for the location.

Next, in step 310, according to the read report data, the actual measurement point position information is mirror-compensated through an internal processing program. In step 311, the surface fitting is performed according to the compensation point obtained by the mirror compensation. The method of mirror compensation and surface fitting is the same as step 140 in the embodiment of FIG. 1 or step 210 in the embodiment of FIG. And step 211, which will not be repeated here. Then, in step 312 and step 313, according to the result of the surface fitting, the model can be automatically re-modeled, and then the processing method in the middle processing is automatically performed and the recalculation is performed to obtain the finishing method. For example, by using an internal plug-in program with a synchronous modeling method, the fitted surface automatically replaces the original CAD model, automatically re-modeling, and automatically recalculates the middle machining method, and sets the residual material to zero, forming The finishing method, the finishing path is the fitted surface. At step 314, automatic finishing compensation is performed and the task is completed. The method of the invention can be automated on the software end, so the entire process of workpiece surface error compensation can be automated by the MES system.

At present, the industry can not achieve the integration of CAD/CAM software automation. Generally, the model and processing method are established by manual method, and the compensation of the automatic controller can only be applied to the two-dimensional compensation, and it is difficult to automatically compensate the three-dimensional free-form surface. It must be achieved through the use of various software. However, if the steps of performing CAD/CAM modeling, measuring layout, and surface fitting are all performed using different software, it will cause many problems in combination with industrial software automation, and any software in these steps. Failure to combine with other software can cause the automated surface error compensation method to fail. The invention further provides a computer program product which can solve the above problems.

7 is a block diagram of a computer program product in accordance with an embodiment of the present invention. The computer program product of the present invention has at least one code, and after the electronic device loads and executes the code, the above automatic surface error compensation method can be performed. This code can be a plugin or other suitable application. For example, using Visual Studio C# with an application planning interface (such as the NXOpen API) The code may be written or may be implemented by other suitable programming language or other software development tools, and the invention is not limited thereto. For example, a five-axis machine from Siemens can be used to machine the surface model, and Siemens NX CAD/CAM software is used with Visual Studio C# to perform software secondary development on NX CAD/CAM to perform, for example, Figure 1. Figure 5 or Figure 6 is an automated surface error compensation method. In other embodiments, automated surface error compensation may be performed using other branded processing machines and applying appropriate CAD/CAM software in conjunction with a computer program having the described code of the present invention.

Specifically, referring to FIG. 7 , the electronic system 400 can be a general purpose computing device, which can include a processor 402 and a storage medium 404 . For example, processor 402 can be a hardware processor, such as a central processing unit (CPU), an application specific integrated circuit (ASIC), or other suitable processing unit. The processor 402 can be coupled to the storage medium 404 or other input/output (I/O) interfaces (such as a keyboard, mouse or touch screen, etc.) or the network interface 403 by means of a bus. The network interface 403 can be connected to the network such that the processor 402 and the storage medium 404 can be connected to external components via a network to enable the system 400 to communicate with the networks of other systems. The storage medium 404 can be a computer readable storage medium such as a random access memory (RAM), a hard disk, a compact disc, or other suitable storage medium. Storage medium 404 can be used to encode (store) computer code 406, 407. That is, the code is one or more sets of machine instructions that are executable, so in Figure 7, the code 407 is indicated by a dashed line. This instruction is for example Software, applications, plugins, or other, and this directive can be implemented using any suitable programming language, application programming interface (API), or other software development tools. For example, the code 406, 407 can perform the operations described above by the processor 402, or cause the processor 402 to generate an external machine readable command to perform the operations in the methods described above. The method of FIG. 1, FIG. 5 or FIG. 6 may be performed by processor 402 in a single system 400. The input and output of the code 406, 407 signals are processed by a single processor 402, that is, by configuring the processor 402 to execute the code 406, 407 encoded in the storage medium 404 to enable the system 400 to execute The above operation of the automatic curved surface error compensation method of the present invention does not require additional calling of other code to perform the operations in the above method, so that the steps in the above method can be completed in a single program operation. The system 400 can also receive and/or transmit the measurement information 408, for example, transmit the measured point information to the processor 402, or store the measurement information 408 in the storage medium 404 and use the code to read Take measurement information 408. With the computer program product of the present invention, after the electronic device loads and executes the code, for example, automatic placement and subsequent measurement, mirror compensation and surface fitting can be performed on a single program (such as NX CAD/CAM), and then, The CAD model is automatically re-modeled to form a compensation processing path, etc., without using various software to perform these modeling, measuring layout, surface fitting and other steps, thereby effectively achieving the automation end of the software end.

In summary, the automatic curved surface error compensation method of the present invention can perform the processing of the workpiece by a single system, and performs the steps of measuring and correcting the error according to the processing result, thereby not only improving the accuracy of the workpiece size, but also improving the accuracy. Traditional technique During the operation, it is necessary to rely on different software to perform the above steps, which leads to problems such as time-consuming positioning errors in parameter setting or coordinate conversion, and thus effectively achieve the automation goal of the software end. Furthermore, the present invention proposes an ideal concept of correcting the connection between the ideal measurement point and the actual measurement point as a normal vector, by first projecting the actual measurement point onto the normal direction of the ideal surface, and then obtaining the mirror compensation. The compensation point is more suitable for complex free-form surfaces than the general mirror principle. In addition, the automated surface error compensation method of the present invention can effectively integrate and manage the system in response to the development of Industry 4.0.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the present invention, and any one of ordinary skill in the art can make some changes and refinements without departing from the spirit and scope of the present invention. The scope of the invention is defined by the scope of the appended claims.

Claims (10)

An automatic surface error compensation method, which is suitable for automatically processing a workpiece according to a machining program control tool machine, the method comprising: establishing a plurality of imaginary planes on an ideal surface to obtain a plurality of curve features, wherein the ideal surface is free a curved surface, wherein the curved feature is a set of curves intersecting each of the imaginary planes with the ideal curved surface; performing a point calculus according to the curved feature to obtain position information of an ideal measuring point; Performing an on-line surface measurement to obtain actual position information of a plurality of actual measurement points; and performing error calculation and mirroring on the actual measurement points and the plurality of the ideal measurement points corresponding to the ideal curved surface Compensating for surface fitting and generating a finishing path, wherein performing the error calculation and performing mirror compensation comprises projecting the actual measurement point onto a normal direction of the ideal surface, and then performing the mirroring Compensation to obtain a compensation point. The automatic surface error compensation method of claim 1, wherein performing the error calculation and performing mirror compensation further comprises: obtaining the actual measurement point according to a normal vector measurement of the ideal measurement point, Connecting a theoretical point to the actual measurement point to obtain a connection, and establishing an extension line along the normal direction of the ideal measurement point, and determining an angle between the connection line and the extension line or the ideal amount Measuring the shortest distance from the point to the normal of the actual measurement point. If the angle or the shortest distance is not 0, indicating that the actual measurement point is not the point on the normal vector, then Projecting the actual measurement point onto the ideal curved surface, and intersecting the ideal surface by the normal direction of the actual measurement point to obtain a projection point, wherein the normal direction of the actual measurement point is Mirror compensation is performed to obtain the compensation point. The automatic surface error compensation method according to claim 1, wherein the surface fitting is performed according to the compensation point, and the finishing path is the fitted curved surface. The automatic curved surface error compensation method according to claim 1, wherein after performing the layout calculation, performing roughing and medium machining on the workpiece, and executing the middle machining, executing the workpiece The on-line surface measurement. The automated surface error compensation method of claim 4, wherein after performing the on-line surface measurement, generating a model of the workpiece, the method further comprising: after performing the surface fitting, The fitted surface replaces the ideal surface for remodeling. The automated surface error compensation method of claim 5, wherein after performing the remodeling, copying the machining path of the middle machining and recalculating to generate the finishing path. The automated surface error compensation method of claim 1, wherein performing the layout calculation according to the curve feature comprises: determining a curvature change of the curve to determine a distribution density. The automatic surface error compensation method of claim 1, wherein after performing the layout calculation, the processing program generates a measurement path according to the result of the layout calculation to control the machine tool pair The workpiece performs the on-line surface measurement. The automatic curved surface error compensation method according to claim 1, wherein the processing program automatically reads the actual measurement point after the machine tool performs the online surface measurement on the workpiece. The position information is described to perform the error calculation. A computer program product having at least one code, and when the electronic device loads and executes the code, the automatic surface error compensation method described in claim 1 can be completed.