Method and system for optimizing processing efficiency based on mold polishing process

Abstract

This invention proposes a method and system for optimizing the processing efficiency of a mold polishing process, relating to the field of polishing process optimization technology. The method involves identifying local polishing units, assigning a unit identifier to each unit, determining the initial permissible processing state for each unit, and generating a local polishing unit state table based on the initial permissible processing states of all local polishing units. It then determines the actual area of ​​the current path on the local polishing unit, calculates the coverage ratio and path integrity of the actual area, obtains surface quality inspection results, identifies the local polishing units requiring retesting based on the path action record table and the current path segment identifier, calculates the local incremental proximity and local effective processing value of the local polishing units, and constructs a local effective processing state table. Finally, it calculates the subsequent path retention coefficient to determine control actions, and adjusts the candidate polishing paths according to the control actions to generate adjusted subsequent mold polishing paths and process parameters.

Description

Technical Field

[0001] This invention relates to the field of polishing process optimization technology, and more particularly to a method and system for optimizing the processing efficiency of mold polishing processes. Background Technology

[0002] Mold polishing is a critical post-processing step in the manufacturing of injection molds, die-casting molds, stamping dies, and precision forming molds. Its results directly affect the appearance quality of the finished product, demolding performance, cavity dimensional stability, and mold lifespan. Existing mold polishing processes can utilize 3D mold models, scanned point clouds, or preset process files to generate equidistant paths, intersecting paths, and area-covering paths. Processing can be performed by robots, CNC polishing equipment, or constant-force polishing mechanisms according to preset pressure, feed rate, and dwell time. Furthermore, roughness inspection, visual texture inspection, or morphology inspection can be used to determine whether local surfaces are close to the target quality.

[0003] While these techniques can improve path repeatability and processing consistency in scenarios with regular curved surfaces or stable process experience, they still have significant shortcomings in complex mold cavities. Mold surfaces typically include free-form surfaces, deep cavity sidewalls, rounded transitions, edges, parting lines, insert boundaries, and high-precision mating areas. The tool center trajectory passing through a certain position does not equate to that position receiving a stable, sufficient, and quality-improving polishing effect. In areas of abrupt curvature change, narrow groove edges, or deep cavity sidewalls, the tool contact range, actual contact pressure, and dwell time will vary due to tool posture, flexible deformation, and local surface changes. This can easily lead to situations such as trajectory coverage but insufficient contact, contact existing but limited quality improvement, and local areas being repeatedly rubbed without effective processing.

[0004] Existing path planning methods often rely on geometric coverage, fixed path iterations, preset processing times, or single-pass quality inspection results. They lack a process judgment mechanism that continuously correlates the initial quality of a local area, target quality, boundary sensitivity, the actual effect of the path, and post-processing quality changes. Therefore, they are prone to misinterpreting path passage as effective processing and continuing repeated coverage even when local areas are close to the target quality or at over-polishing sensitive boundaries. Since the purpose of mold polishing is to balance surface quality and shape accuracy, continuing to subject parting lines, edges, fillets, thin walls, and mating surfaces to high pressure or prolonged processing can lead to edge collapse, over-polishing, dimensional deviations, uneven local reflections, and rework. Thus, the low efficiency of existing mold polishing methods is not simply due to insufficient equipment speed, but rather to a lack of distinction between the actual effect of the path and effective processing, and a lack of control mechanisms to promptly remove fully processed or low-yield repetitive areas from subsequent paths, resulting in increased ineffective coverage, over-polishing, and rework time. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for optimizing the processing efficiency of mold polishing processes, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, this invention proposes a method for optimizing the processing efficiency of a mold polishing process, comprising: Acquire detection data of the mold surface to be polished, determine the effective range of the mold surface to be polished based on the detection data, divide the effective range into multiple local polishing units, assign a unit identifier to each local polishing unit, calculate the boundary sensitivity level of the local polishing unit, determine the initial processing allowable state of each local polishing unit according to the unit identifier and boundary sensitivity level of each local polishing unit, and generate the local polishing unit state table based on the initial processing allowable state of all local polishing units. Acquire path execution data, unify the path execution data to the mold coordinate system, map the path sampling points to the local polishing unit, determine the actual action area of ​​the current path on the local polishing unit, calculate the coverage ratio of the actual action area, calculate the local action reference coefficient according to the local polishing unit status table, calculate the path action integrity based on the coverage ratio and the local action reference coefficient, determine the path action type of the local polishing unit, and summarize the local polishing unit status table, local action reference coefficient, path action integrity and path action type to construct a path action record table; After obtaining the surface quality detection results of each local polishing unit after the current path segment is applied, the local polishing units that need to be retested are determined according to the local polishing units in the path application record table and the current path segment identifier. The local incremental proximity of the local polishing unit is calculated based on the surface quality detection results. The local effective processing value is calculated based on the local incremental proximity. A local effective processing status table is constructed based on the local incremental proximity and the local effective processing value. The candidate polishing path to be executed is read. For each local polishing unit covered by the candidate polishing path, the subsequent path retention coefficient is calculated based on the local effective processing state table. The control action of the candidate polishing path on the corresponding local polishing unit is determined according to the subsequent path retention coefficient. The candidate polishing path is adjusted according to the control action on each local polishing unit to generate the adjusted subsequent mold polishing path and process parameters. The processing efficiency is optimized based on the subsequent mold polishing path and process parameters.

[0007] In some embodiments, the detection data includes three-dimensional geometric data, polishing process targets, and initial surface quality detection results, which are obtained by contact roughness detection, white light interferometry detection, structured light morphology detection, or visual texture detection.

[0008] In some embodiments, when the initial surface quality test results exist in the form of multiple sampling points, each sampling point is mapped to its corresponding local polishing unit; When the initial surface quality detection result is a region image, the region image is first matched with the position of the three-dimensional curved surface, and then the initial quality state at the corresponding position is read.

[0009] In some embodiments, the coverage ratio is obtained by the surface overlap relationship between the tool contact range within the current path segment and the local polishing unit; When consecutive sampling times cover the same local polishing unit, the overlap of each sampling time is merged according to the temporal continuity to form a proportional coverage of the local polishing unit.

[0010] In some embodiments, the path action integrity and path action type are both recorded using the local polishing unit as an index.

[0011] In some embodiments, the control action of determining the candidate polishing path on the corresponding local polishing unit based on the subsequent path retention coefficient specifically includes: When the subsequent path retention coefficient is higher than the first threshold, the control action is to retain the path and execute according to the contact pressure and residence time set in the current process. When the subsequent path retention coefficient is between the first threshold and the second threshold, the control action is to shorten the dwell time or reduce the contact pressure. When the subsequent path retention coefficient is lower than the second threshold and the low-return repetition state is continuous low-return repetition or boundary repetition type inefficiency, the control action is local compression or detour. When the subsequent path retention coefficient is lower than the second threshold and the quality improvement status is inefficient due to insufficient effect, the control action is to include it in the re-action area and adjust the subsequent action conditions according to the type of insufficient effect.

[0012] In some embodiments, the subsequent mold polishing path and process parameters include a retained path segment, a compressed path segment, a bypass path segment, an adjusted contact pressure, an adjusted dwell time, and a re-action area.

[0013] In some embodiments, the sensitivity level is calculated and generated using a scaled proximity value, a scaled surface change value, a process sensitivity identifier, and a preset weight.

[0014] In some embodiments, the local effective processing status table is constructed using local polishing units as indexes. The local effective processing status table records local polishing units, local incremental proximity, local effective processing values, effective processing degree, quality improvement status, low-yield repetition status, covered segment sequence number, and current path segment identifier.

[0015] To achieve the above objectives, another aspect of the present invention proposes a system for optimizing the processing efficiency of a mold polishing process, comprising: The status table construction module is used to acquire detection data of the mold surface to be polished, determine the effective range of the mold surface to be polished based on the detection data, divide the effective range into multiple local polishing units, assign a unit identifier to each local polishing unit, calculate the boundary sensitivity level of the local polishing unit, determine the initial processing allowable state of each local polishing unit according to the unit identifier and boundary sensitivity level of each local polishing unit, and generate the local polishing unit status table based on the initial processing allowable state of all local polishing units. The action record table construction module is used to acquire path execution data, unify the path execution data into the mold coordinate system, map the path sampling points to the local polishing unit, determine the actual action area of ​​the current path on the local polishing unit, calculate the coverage ratio of the actual action area, calculate the local action reference coefficient according to the local polishing unit status table, calculate the path action integrity based on the coverage ratio and the local action reference coefficient, determine the path action type of the local polishing unit, and summarize the local polishing unit status table, local action reference coefficient, path action integrity and path action type to construct a path action record table; The processing status table construction module is used to obtain the surface quality detection results of each local polishing unit after the current path segment is applied, determine the local polishing unit that needs to be retested based on the local polishing unit in the path application record table and the current path segment identifier, calculate the local incremental proximity of the local polishing unit based on the surface quality detection results, calculate the local effective processing value based on the local incremental proximity, and construct a local effective processing status table based on the local incremental proximity and the local effective processing value. An efficiency optimization module is used to read candidate polishing paths to be executed, calculate a subsequent path retention coefficient for each local polishing unit covered by the candidate polishing path based on the local effective processing state table, determine the control action of the candidate polishing path on the corresponding local polishing unit according to the subsequent path retention coefficient, adjust the candidate polishing path according to the control action on each local polishing unit, generate the adjusted subsequent mold polishing path and process parameters, and optimize the processing efficiency based on the subsequent mold polishing path and process parameters.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention first establishes initial surface quality, target surface quality, boundary sensitivity level, and initial processing allowable state for each local polishing unit based on the mold geometry model, polishing process target, and initial surface quality detection results. This ensures that ordinary smooth areas, rounded corner areas, areas near parting lines, and mating surface areas possess different local process attributes before polishing begins. Subsequently, when the equipment executes the polishing path, the tool contact range, actual contact pressure, and dwell time are mapped to the corresponding local polishing unit, forming a path action record. This distinguishes the differences between the tool center trajectory, tool contact range coverage, and the actual action on the local area. Then, combined with the path action... The surface quality changes generate locally effective processing states, determining whether the path action truly contributes to achieving the quality target and identifying different process states such as weakening quality improvement, low-yield repeated coverage, inefficient boundary rubbing, and insufficient action. Finally, based on the locally effective processing states, the subsequent mold polishing path and process parameters are actually controlled. This ensures that areas with effective processing value retain necessary path action, while areas that have been fully processed, have weakened quality improvement, or are repeatedly affected by boundary rubbing have reduced dwell time, lower pressure, compressed paths, detours, or no coverage. Areas that have not been effectively processed due to low pressure or short dwell time are adjusted for contact conditions before further processing. Through this continuous process chain, the present invention can transform the local action facts during path execution into subsequent control actions that the equipment can execute, reducing processing actions that pass through the path but have no significant quality benefit, reducing the risk of repeated coverage of fully processed areas and sensitive boundary areas, and reducing ineffective re-polishing, over-polishing, and rework time while ensuring mold surface quality and profile accuracy, thus improving the overall processing efficiency of the mold polishing process. Attached Figure Description

[0017] Figure 1 This is a flowchart of the method for optimizing the processing efficiency of the mold polishing process proposed in the embodiments of this application; Figure 2 This is a block diagram of a system for optimizing the processing efficiency of a mold polishing process, as proposed in an embodiment of this application. Figure 3 This is a schematic diagram illustrating the calculation results of the local effective processing value and subsequent path retention coefficient proposed in the embodiments of this application. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] refer to Figure 1As shown, one embodiment of this application proposes a method for optimizing the processing efficiency of a mold polishing process, including: S1: Obtain detection data of the mold surface to be polished, determine the effective range of the mold surface to be polished based on the detection data, divide the effective range into multiple local polishing units, assign a unit identifier to each local polishing unit, calculate the boundary sensitivity level of the local polishing unit, determine the initial processing allowable state of each local polishing unit according to the unit identifier and boundary sensitivity level of each local polishing unit, and generate the local polishing unit state table based on the initial processing allowable states of all local polishing units, specifically including: This step establishes a local state foundation that can be directly used by subsequent processes before the mold polishing begins. It organizes the mold surface to be polished into a set of local polishing units that can be aligned by equipment paths, contact actions, and surface quality checks. In actual engineering, mold surfaces typically have structures such as cavity bottoms, sidewalls, rounded transitions, areas near parting lines, insert boundaries, and slender ribs. If a full-surface polishing path is generated directly at fixed intervals, it's easy to mix ordinary smooth areas with sensitive boundary areas. For example, the same path might simultaneously cover a smooth area that still needs rough polishing and a rounded edge that is close to the target quality. If subsequent equipment processes both with the same pressure and dwell time, it can easily cause over-polishing of the edges. Therefore, before polishing begins, the mold surface to be polished is first divided into local polishing units with process significance. Each local polishing unit is then programmed with initial quality, target quality, boundary sensitivity level, and initial processing allowable state. This ensures that whether subsequent paths act on a certain area, whether that action results in effective processing, and whether that area should continue processing all point to the same local object.

[0020] The data used in this step comes from data already available or obtained through testing before the mold polishing process begins. This includes the 3D geometric data of the mold surface to be polished, the polishing process objectives, and the initial surface quality inspection results. The 3D geometric data of the mold surface to be polished can come from the mold CAD model, exported from the mold design file, including the polished surface, parting line, insert boundary, fillet boundary, and mating surface position; or it can come from the physical scan data before polishing, acquired by 3D scanning equipment, structured light measurement equipment, or online probes on machine tools. If both CAD model and physical scan data exist, the CAD model provides the design boundaries and process areas, while the physical scan data provides the actual surface position. The two are registered through mold reference holes, positioning surfaces, clamping references, or preset reference points to ensure that the design coordinates correspond to the physical coordinates. The polishing process objectives come from process documents or production task documents, such as a certain area requiring a mirror finish, a certain area requiring the elimination of tool marks, or a certain area being a mating surface requiring restricted strong polishing. The initial surface quality test results are obtained from the pre-polishing test, which can be obtained through contact roughness detection, white light interferometry detection, structured light morphology detection or visual texture detection. If the detection device outputs multiple sampling point results, the sampling points are mapped to the corresponding local polishing units. If the detection result is a regional image, the image region is first mapped to the position of the three-dimensional curved surface, and then the initial quality state of the corresponding position is read.

[0021] When determining the effective range of the surface to be polished, for molds already marked as surfaces to be polished in the CAD model, the surface to be polished is read; for mold data in the form of scanned point clouds, the range to be polished is determined based on the processing area selected by the equipment, the mold cavity boundary, or the matching result with the CAD model. When determining the range, the cavity forming surface, the area corresponding to the product's appearance surface, the sidewall area where tool marks need to be removed, and the rounded corner transition area are included in the processing objects. For example, for an injection mold cavity, the area to be polished may include the cavity bottom surface, sidewalls, and rounded corner transition area corresponding to the product's appearance surface; for a die-casting mold, the area to be polished may include the cavity forming surface and the area near the local gate. Subsequently, the surface to be polished is divided into local polishing units. The division is handled in conjunction with the tool contact range, surface variations, and functional boundaries: in locations with relatively gentle surfaces, far from parting lines and edges, local polishing units can cover relatively continuous surface patches; at locations such as rounded corners, edges, parting lines, deep cavity corners, and insert mating boundaries, the range of local polishing units is correspondingly reduced, allowing these areas to be individually identified and restricted in subsequent control. For example, a relatively flat area on the bottom surface of the mold cavity can be classified as a normal curved surface element, while a narrow area near the parting line, even if adjacent to the bottom surface, can be classified as a boundary element. In this way, when the subsequent path passes through this area, the system can reduce the dwell time or pressure on the boundary elements, while maintaining normal processing on the normal curved surface elements.

[0022] After the division is completed, each local polishing unit is assigned a unique unit identifier, denoted as . This identifier is linked to the mold surface coordinates, and subsequent path action records and effective machining states are all attributed to the same local area through this identifier. This represents a local surface region that can be covered by the tool's contact area, rather than a single detection point; even if the number of detection points or path sampling points is large, subsequent control still uses stable local polishing units as the object. Next, the initial surface quality obtained before polishing is written into the corresponding local polishing unit, denoted as... When multiple detection results exist within a single local polishing unit, the detection value that best represents the main processing state of that area can be used. For example, if a region has a few isolated detection noise points, but most detection points show obvious tool marks, the initial surface quality of that unit reflects the main tool mark state. If a region has obvious local defects located in areas inevitably covered by the tool, the detection result corresponding to the defect is written into the unit's state. Then, the polishing process target is written into the corresponding local polishing unit, denoted as... Different local polishing units can have different objectives. For example, the target surface quality of the corresponding area on the product's exterior surface is higher, while the target for the non-exterior internal surface is relatively lower, and the mating surface area emphasizes shape preservation. If the process document gives a uniform objective for the entire surface to be polished, then the ordinary local polishing unit can use the same objective. If the process document gives different levels for different areas, then the corresponding objective should be read according to the mold area where the local polishing unit is located.

[0023] To ensure that the dimensions of subsequent formulas are consistent, and All values ​​are proportionalized in the same direction, with a range of values ​​of [value range missing]. to A larger value indicates that the surface is closer to the corresponding process target. If the testing equipment directly outputs raw physical quantities such as roughness value, texture defect area, tool mark depth, or morphological deviation, these values ​​should first be converted into the aforementioned proportional quality values ​​according to the acceptable upper limit, target value, and allowable deviation in the process document. For example, in scenarios where lower roughness equates to better quality, the original roughness value is not directly substituted into the formula; instead, it is first converted into a "the larger the better" quality ratio value before being written into the formula. and When a certain polishing unit lacks effective detection points, the system prioritizes using the detection results of adjacent similar surface units and the process target of that unit to form a retesting mark, and then performs supplementary testing on that unit before or after the execution of subsequent paths, rather than directly using missing data to participate in the effective processing judgment.

[0024] Then, the boundary sensitivity level of each local polishing unit was determined. This level characterizes the degree of over-polishing risk that the unit may experience during subsequent polishing processes due to repeated coverage or excessive processing. The boundary sensitivity level is determined by the proximity of the unit to the functional boundary, the degree of change in the surface on which the unit is located, and the sensitivity markers for that area in the process documents. Functional boundaries can be read from the CAD model or determined by registering the scanned point cloud with the design model, including parting lines, edges, rounded corner boundaries, insert boundaries, and mating surface boundaries; the degree of surface change is obtained from the geometric changes of the local polishing unit and its adjacent areas, such as changes in surface normal, curvature, or the angle between adjacent surfaces; process sensitivity markers are read from the process documents or design annotations, such as "light polishing near the parting line," "no heavy pressure on mating surfaces," and "rounded corners to maintain shape," which can all be converted into sensitivity markers for that local polishing unit. The boundary sensitivity level can be determined by the following formula: ; in, For local polishing unit The boundary sensitivity level is used as one of the states generated in this step; For local polishing unit The scaled approximation value to the nearest functional boundary is calculated from the mold CAD model or the registered 3D scan data. The closer to the functional boundary, the more accurate the calculation. The larger; For local polishing unit The proportional surface variation value is calculated from the geometric changes of the element and the adjacent surface region; For process-sensitive areas, the identification value is read from the mold design documents or polishing process documents. A higher identification value is used when the area is a sensitive area, and a lower identification value is used when the area is not a sensitive area. , , The preset weights are set by the mold manufacturer's process rules or historical process experience to reflect the influence of the proximity of functional boundaries, the degree of surface variation, and process sensitivity indicators on the boundary sensitivity level. The components involved in this calculation... , , All values ​​are first converted to proportional values ​​before being used in the calculation. This formula pre-constructs factors that may cause over-polishing in engineering into local conditions. For example, if a unit is very close to the parting line, even if the current surface still needs to be treated, prolonged high-pressure coverage should be avoided in subsequent stages; if a unit is located at the transition of a deep cavity fillet, tool contact spots are prone to concentrate, and its sensitivity level should be increased; if a unit is marked as a critical appearance surface or mating surface in the process document, even if it is geometrically relatively smooth, it should be handled with care in subsequent steps.

[0025] In practical implementation, , and All are limited to to Within the dimensionless range, , and The weights are non-negative and stored in the same set of process parameter tables. If a local polishing unit has no adjacent parting lines, edges, insert boundaries, or mating surface boundaries, its nearest functional boundary proximity value is treated as far from the boundary. If the surface change detection result is affected by point cloud noise, smoothing or taking stable statistical values ​​is performed first within the local unit and adjacent units before obtaining the final result. After this processing, all terms in the formula are proportional quantities of the same scale. The output is also limited to to It can be directly used as the sensitivity level in S2 and S4.

[0026] Finally, according to , and Determine the initial processing allowable state This state indicates the extent to which this local polishing unit is allowed to participate in the initial path execution before polishing begins. If and The gap is obvious, and Lower, then Set to normal processing status; if and The difference is small, or If it is higher, then Set to restricted processing state; if the process document explicitly states that this area must not be directly covered by ordinary polishing tools, then Set to a state where normal coverage is prohibited. For example, if there are still obvious tool marks on a certain area of ​​the cavity bottom surface, far from the parting line and fillet, then this local polishing unit can be set to normal processing state; if there is a certain area close to the parting line, although the surface still has slight texture, but continued high-intensity polishing may cause edge collapse, then this local polishing unit is set to restricted processing state; if a certain area is a high-precision mating surface, and the process document requires that normal path coverage should not be performed, then this local polishing unit is set to a state where normal coverage is prohibited. This forms a local polishing unit state table, which is based on local polishing units. As the basic record object, each record includes , , , and ;in, It is obtained by dividing the surface to be polished into local sections and binding them to the position of the mold surface. The result is obtained by mapping the surface quality inspection results before polishing to the corresponding local polishing unit. This is obtained by mapping the target of the polishing process to the corresponding local polishing unit. It is determined by the proximity of functional boundaries, the degree of surface variation, and process sensitivity indicators. The initial surface quality, target surface quality, and boundary sensitivity level are jointly determined. This local polishing unit state table serves as the basis for path action mapping in subsequent steps, mapping the tool contact range, contact pressure, and dwell time to the table. This forms the corresponding path action record.

[0027] S2: Obtain path execution data, unify the path execution data to the mold coordinate system, map the path sampling points to the local polishing unit, determine the actual action area of ​​the current path on the local polishing unit, calculate the coverage ratio of the actual action area, calculate the local action reference coefficient according to the local polishing unit status table, calculate the path action integrity based on the coverage ratio and the local action reference coefficient, determine the path action type of the local polishing unit, and summarize the local polishing unit status table, local action reference coefficient, path action integrity, and path action type to construct a path action record table, specifically including: This step builds upon the local polishing unit state table created in the previous step, transforming the continuous mold polishing path into one that targets each local polishing unit. The path effect record table. The output of the previous step... , , , and All of them are used in this step, among which Spatial objects used to determine the role of a path. and Used to determine the initial processing requirements of this unit relative to the target surface quality. Used to characterize the sensitivity of the unit to parting lines, edges, fillets, insert boundaries, or mating surfaces. This is used to constrain the intensity of the current path's effect on the unit. The path's effect in mold polishing exhibits significant local differences. While the same tool path's effect on the cavity bottom surface can typically be recorded as normal machining, near fillets or parting lines, even a mere brush of the tool edge can influence whether subsequent re-covering should continue. Therefore, this step decomposes the equipment's execution path into each... The facts of coverage, contact, and dwell on the path are combined with the local state of the previous step to form a path action record that can be directly used in the next step.

[0028] The path execution data is jointly provided by the polishing equipment's motion controller, force control module, and process program. During path execution, the motion controller outputs the tool center trajectory, path sampling time, current path segment identifier, and end-effector posture. The force control module outputs the actual contact pressure corresponding to each sampling time. The process program provides the current tool type, tool contact range parameters, and pressure and dwell settings for the current process stage. For robotic polishing equipment, the tool center trajectory and end-effector posture can be read by the robot controller, and the actual contact pressure can be read by the force control flange or constant force actuator. For CNC polishing equipment, the tool center trajectory can be read from the interpolation trajectory, and the contact pressure can be read by the pressure control module. The read path sampling points are aligned with the mold coordinate system and S1. The surface coordinates are matched so that the tool's action area at each sampling moment can fall on the local polishing unit.

[0029] The aforementioned path execution data is unified to the same time reference and the same mold coordinate system before entering the path action record. The tool center trajectory and end-effector posture retain physical units such as length and angle; the actual contact pressure retains the pressure or normal force units from the equipment force control system; and the coverage duration retains the time unit. However, the data participating in the path execution are not included in the overall data. , , , , and During calculations, all results are first converted to scaled-down values ​​relative to the current tool, current path segment, and current process settings. This avoids directly multiplying different units such as millimeters, Newtons, and seconds, while preserving the traceability of each original physical quantity within the equipment control program.

[0030] After the path sampling points are mapped onto the mold surface, the actual area of ​​action of the current tool on the curved surface is determined based on the tool center position, tool posture, and tool contact range. If the tool contact range covers a certain local polishing unit... In some areas, then... Establish coverage records; if the contact area at the same sampling time simultaneously covers multiple... Then, records are written to these local polishing units respectively. For example, when the flexible polishing wheel runs along the edge of the cavity bottom surface, the tool center may be located within the bottom surface unit, but the edge of the contact patch will cover the rounded corner unit. This step records the path action of both the bottom surface unit and the rounded corner unit. When the oilstone runs along the side wall of the narrow groove, if the end covers the unit near the edge, that edge unit also forms a path action record. Coverage ratio Based on the tool contact range within the current path segment and the local polishing unit The surface overlap relationship is obtained, and the continuous sampling time covers the same At that time, the overlap of each sampling moment is merged according to temporal continuity to form the path segment pair. The coverage ratio results.

[0031] Before generating the path action record, the local action reference coefficient is determined based on the local polishing unit state table from the previous step. This coefficient originates from the engineering weighted scoring method, which combines multiple proportionalized factors affecting the permissible degree of path effect in a multiplicative manner, such that a decrease in any key factor directly lowers the baseline effect of that unit. This approach is derived from the classic multi-factor normalization evaluation method: first using... This indicates the processing requirements of the local polishing unit relative to the target surface quality, and then... This represents the constraint of boundary sensitivity on the path effect, and then uses... This represents the process constraints on the intensity of the effect of the initial processing allowable state on the path, and finally sets the minimum recording baseline. This allows areas that are close to the target but still covered by the path to be recorded. Therefore: ; in, For local polishing unit The local action reference coefficient is used to subsequently determine whether the current path coverage exceeds the reference action range of the local unit in the current stage; Write to S1 The initial surface quality is obtained by mapping the surface quality test results before polishing; Write to S1 The target surface quality is obtained by target mapping in the polishing process; The boundary sensitivity level obtained from S1; To be from the initial processing allowed state The state coefficients obtained from the transformation, when Take the higher value when it is in normal processing condition, when To limit the processing state to take the intermediate value, when To prevent the use of lower values ​​during normal coverage, this conversion relationship is configured by the equipment process parameter table or the company's polishing rules; The minimum record baseline is configured by the process parameter table. , , , and All calculations use proportionalized values, so that Maintain the same scale as each input.

[0032] To avoid situations where local action benchmarks become unexecutable or unclear, and All set to to The scaling parameters within the range, and Do not take This is used to ensure that local cells that are close to the target quality but are still covered by the path can be recorded. It is used only as a baseline threshold in the calculation of path integrity and does not directly replace the actual contact pressure, residence time, or coverage area; the actual physical control quantity is still determined by subsequent steps. Convert to pressure setting, feed rate, dwell command, or path circumvention.

[0033] The derivation logic of this formula corresponds to the actual process relationship of mold polishing: when and When the difference is large, the region has a high processing demand; when When the level is high, the area is close to sensitive boundaries, increasing the risk of duplicate coverage; when When processing is restricted or normal coverage is prohibited, the path action reference for that area needs to be lowered. A multiplicative combination of these three factors allows highly sensitive boundaries or restricted conditions to directly suppress the path action reference of the element, thus significantly recording coverage exceeding the reference in subsequent path action integrity calculations. Taking a locally polished element as an example, if the initial surface quality ratio of the element... Target surface quality ratio Boundary sensitivity level Initial processing allowed status correspond Minimum Record Benchmark ,but This indicates that a relatively sufficient path effect can be formed in a generally flat area at the current stage; if another unit also exists... The quality difference is small, but it's close to the fractal line. ,and correspond ,but This indicates that even if there is still a need for processing, the smaller coverage of the sensitive boundary unit should be recorded as a key fact.

[0034] Subsequently, for each element covered by the current path's contact range... Calculate the completeness of the path action This value is derived from the classic product-based comprehensive evaluation method, where coverage, contact execution, and sensitive over-benchmark coverage jointly determine the path action recording intensity. Coverage ratio Indicates the area of ​​effect of the tool and The degree of overlap; the sufficiency of contact execution This indicates whether the pressure and dwell time during the coverage period meet the current process settings; the sensitive over-reference item is used to characterize the risk of boundary re-coverage unique to mold polishing. First, by... The base path action intensity is obtained, and then the intensity exceeding the local action benchmark is determined. Extract the covered portion and use boundary sensitivity level By zooming in, we can obtain the completeness of the path action in the mold polishing scenario: ; in, For local polishing unit Determine the completeness of the path's effect under the current path and write it to the path effect record table; For the current path, apply local polishing unit The coverage scaling result is determined by the tool's contact range and The surface overlap relationship is calculated; For the current path in The adequacy of contact execution is determined by the degree to which the actual contact pressure during the coverage period meets the process setting and the degree to which the coverage duration meets the process setting. Specifically, the lower of the two values ​​can be obtained by reading the pressure achievement and coverage duration from the equipment controller. The boundary sensitivity level obtained from S1; The local action reference coefficient is obtained from the previous equation. , , and All calculations use proportionalized values, so that It can be used as the relative intensity of action in subsequent effective processing judgments. In this formula... The fundamental facts governing the simultaneous occurrence of path coverage and contact execution. For additional recording weights exceeding the local action benchmark in the corresponding sensitive area, the outer layer value restriction is used to keep the recording results within a range that can be directly used in subsequent proportional calculations.

[0035] The source is the tool's contact range and The specific selection of the surface overlap area, overlap length, or sampling point coverage ratio should be consistent with the path data type that the device can output; The sources are the pressure achievement ratio and the residence achievement ratio, both of which are derived from actual values ​​relative to the current process setting. If the current path is relative to a certain... If there is only geometric projection but no effective contact pressure is formed, then It can exist and Lower The effect decreases accordingly; if the current path forms a small area but covers multiple times consecutively at sensitive boundaries, then... Recording the covered segments ensures that the fact of the action is preserved in subsequent steps. This clearly distinguishes between the three different levels of "track passage," "tool contact," and "formation of effective processing," reducing the risk of insufficient disclosure by mistaking ordinary path coverage for effective processing.

[0036] The calculation process of this formula can be directly embedded into the device path execution program. Taking the aforementioned ordinary flat area as an example, if the tool's contact range covers this area... Proportional results The adequacy of contact implementation is obtained by combining pressure and residence during the coverage period. Boundary sensitivity level Local action reference ,but This is recorded as a relatively sufficient path effect. For the aforementioned sensitive units near the fractal line, if... , , , ,but It can be seen that although the coverage ratio of this sensitive unit is smaller than that of the ordinary area, its boundary sensitivity level is high and its coverage exceeds the lower local action benchmark, so this path action is still recorded as an action fact that requires further attention. This result is consistent with the actual situation of mold polishing: large coverage areas are not required near rounded corners, parting lines, and edges to generate risks, and multiple small coverages can also cause boundary shape changes or local over-polishing.

[0037] In formation Simultaneously, this step writes the path action type for each local polishing unit. The path action type is based on... , , , and Direct determination. If To achieve the main coverage area of ​​the tool in the current path and If the current process setting range is reached, it is recorded as a complete function; if Achieve main coverage but If the pressure is insufficient to achieve the desired level, resulting in a lower pressure, it is recorded as a low-pressure effect; if Achieve main coverage but If the low score is due to insufficient coverage duration, it is recorded as a short dwell time effect; if Lower but If it is higher, it is recorded as boundary rubbing; if If the corresponding processing state is restricted or normal coverage is prohibited, and the current path still covers the cell, it is recorded as a restricted area action. Path action type and Write them together into the record table so that subsequent steps can distinguish between different causes such as "sufficient coverage but insufficient pressure," "sufficient pressure but insufficient dwell time," and "sensitive boundaries being rubbed by tool edges." For example, two units... They may be similar, but one is a complete function of a normal area, while the other is a boundary grazing near the fractal line. Different processing logics should be used when judging the effective processing status and controlling the path.

[0038] For cases where consecutive paths repeatedly cover the same local polishing unit, this step writes the coverage segment according to the path sampling order and merges adjacent segments under the same tool state. If the current path continuously covers... Then accumulate the coverage duration and update. , and If the path leaves If the robot then returns to cover the same cell, and the tool state, path segment identifiers, and pressure settings remain consistent throughout, the two segments are considered as the same continuous polishing action and are merged and recorded. If, during this period, the tool is lifted, the path segment identifier changes, the pressure setting changes, or the posture changes significantly, different coverage segments are generated. This processing method can preserve the reciprocating sweeping phenomenon commonly seen in mold polishing. For example, when the robot performs intersecting paths on the bottom surface of the cavity, rounded corner types... The tool may repeatedly sweep across multiple adjacent paths, with each sweep lasting only a short time. However, the path action record table will retain multiple boundary rubbing or restricted area actions, thus providing a basis for identifying the accumulation of repeated actions in the next step.

[0039] The path action record table generated in this step uses local polishing units. For indexing, each record includes , , The path's function type, the number of the covered segment, and the identifier of the current path segment. Among these, The local polishing unit status table directly from S1; Depend on , , and This is obtained as a local action reference used to express the current stage of the unit; The current path coverage ratio, contact execution sufficiency, and sensitive over-reference coverage are all considered. The path action type is determined by coverage, pressure, dwell time, boundary sensitivity level, and initial processing allowable state. The coverage segment number is generated by the path sampling order and the continuous coverage relationship of the same local unit. The current path segment identifier is read from the path segment number in the polishing equipment control program. Thus, the local polishing unit state table from the previous step is transformed into a path action record table oriented towards path execution facts, and the next step can be performed within the same... The surface quality changes before and after the combined path action are further used to generate a locally effective processing state.

[0040] S3: Obtain the surface quality detection results of each local polishing unit after the current path segment is applied. Based on the local polishing units in the path application record table and the current path segment identifier, determine the local polishing units that need to be retested. Calculate the local incremental proximity of the local polishing unit based on the surface quality detection results. Calculate the local effective processing value based on the local incremental proximity. Construct a local effective processing status table based on the local incremental proximity and the local effective processing value, specifically including: This step follows the path action record table generated in S2, further transforming the recorded path coverage, contact execution, and covered fragment relationships into locally effective processing states. The output of S2... , , The path function type, the covered segment number, and the current path segment identifier are all used in this step: Used to determine the local object to which the retest quality belongs. This serves as a reference point to indicate the path action of the local polishing unit at the current stage. This is used to indicate the strength of the objective effect of the current path on the local polishing unit. The path effect type is used to distinguish between complete effect, low-pressure effect, short-dwelling effect, boundary rubbing effect, and confined area effect. The coverage segment sequence number is used to determine the order in which the same local polishing unit is continuously or repeatedly covered. The current path segment identifier is used to associate the current quality change with the specific polishing path segment. The processing goal of this step is to further determine whether "the path has indeed acted on a certain local area" as "whether the effect has brought about effective surface improvement", thereby forming a local effective processing state that can be directly read by subsequent path control.

[0041] After the current path segment finishes execution, the system records the path's function in the table. The local polishing units requiring retesting are identified using the current path segment identifier, and the surface quality after the current path segment has been applied is obtained under the same quality evaluation system. To express the quality change brought about by the current path segment itself, this step records the surface quality before the application of the path segment as follows: The surface quality after the path segment is applied is denoted as .in, During the initial path segment evaluation, S1 is taken. When evaluating subsequent path segments, the surface quality stored after the previous evaluated path segment is completed is taken; Mapping the surface retest results completed by the current path segment to The retest can be performed using online visual inspection, structured light topography inspection, white light interferometry inspection, or contact roughness inspection. The inspection results are mapped to the corresponding local polishing unit positions established in S1. If a certain If multiple coverage segments exist within the same path segment, and S2 has already merged consecutive coverage segments into a single action record, then the retest result after the end of that path segment is used as the quality result corresponding to that merged action record. If coverage segments are recorded separately due to changes in path segment, pressure setting, or tool attitude, this step retains the coverage segment sequence number to distinguish different segments affecting the same action. The order of their effects. , S1 and Using the same scaled quality level, such as those derived from roughness level, texture level, or surface defect level, ensures that the quality changes of the current path segment maintain the same evaluation scale as the target quality.

[0042] When the test results are unstable in a single retest due to polishing fluid, reflection, or dust, the system can perform the same test simultaneously. The results are taken from stable statistical values ​​obtained through multiple retests or after removing obvious outliers. And retain the correspondence between the retest result and the current path segment identifier. If Has reached or exceeded If the unit is close to the target quality in this step, then even if there are slight fluctuations in subsequent detections, they will not be amplified into a higher value due to the small denominator. ;like Below The outer layer value restriction will cause the incremental proximity to not be regarded as a positive effective processing contribution, and will be recorded as no improvement or abnormal retest status in the quality improvement status.

[0043] Specifically, when Has reached or exceeded At this time, the system will no longer substitute the local polishing unit into the incremental proximity calculation as a normal area to be improved, but will record its quality improvement status as met or weakening, and will record the positive path segment corresponding to this path segment. according to The process involves re-entering the normal incremental proximity calculation only when a retest confirms that the unit quality has fallen below the target quality again but is still permissible for processing. This process avoids unstable calculation results when the target difference is zero or less than the detection fluctuation.

[0044] This step first calculates the local increment proximity. The initial source of this formula is the mathematical calculation of the normalized distance ratio, which involves dividing the "state change caused by the current path segment" by the "change required from the current state to the target state" to obtain the contribution ratio of the current path segment to the target approach process. This application introduces the local action reference coefficient obtained by S2 based on this normalized ratio. This is treated as a stable term related to the process state of the local unit. With this treatment, the true improvement in ordinary areas can be preserved, and small detection fluctuations in areas close to the target or sensitive boundary areas will not be amplified into a large effective processing contribution. Therefore: ; in, For local polishing unit The local incremental proximity is generated under the current path segment and written into the local effective processing status table; After applying it to the current path segment The surface quality obtained from the surface quality retest is mapped from the output of the testing equipment to... get; Before applying to the current path segment The surface quality, initially evaluated, is taken as S1. Subsequent evaluations will take the storage quality after the previously evaluated path segment is completed; Write to S1 The target surface quality is obtained by target mapping in the polishing process; The local action reference coefficients obtained from S2 are derived from... , , and form. , , and All calculations are performed using proportionalized values. The derivation of this formula is as follows: numerator The denominator represents the quality improvement brought about by the current path segment. This represents the baseline improvement required to approach the target state from the current path segment's initial state, and a local baseline is superimposed to suppress fluctuation amplification when the target gap is small; the outer layer value constraint is used to limit the incremental approach within the state determination range. Taking a common cavity bottom surface element as an example, if the current path segment's initial state is... Target surface quality S2 is obtained The current path segment was retested and obtained ,but This indicates that the path segment caused the local cell to make a identifiable incremental approach to the target quality. If the same cell is covered again in an adjacent path, the current path segment's effect before... After the current path segment Target surface quality S2 is obtained ,but This indicates that the quality improvement brought about by the second coverage has been significantly weakened.

[0045] thus, The source, calculation direction, and usage location are all determined: their source is the same. The proportionalized quality change before and after the path segment is calculated based on the premise that "an increase in value indicates that the quality is approaching the target," and its output enters... Calculation and S4 Calculation. If the original index such as roughness is used, which is "the smaller the value, the better", then it must first be converted into a mass ratio value in the same direction in S1 and this step. The original roughness value cannot be directly substituted into the formula, otherwise the direction of quality improvement will be opposite to the direction of the formula.

[0046] In obtaining Next, this step calculates the local effective processing value. The initial source of this formula is the product-type comprehensive evaluation method in classical process evaluation, which multiplies the proportionalized results of each condition when multiple necessary conditions jointly determine a comprehensive contribution; at the same time, it incorporates the diminishing returns principle in the processing process and introduces a suppression term for continuous repeated coverage. This application, in its basic product term... Add path action type coefficient This is used to distinguish the different contributions of complete action, low-pressure action, short-stay action, boundary grazing action, and confined area action to effective processing; it is then combined with the continuous coverage segment count obtained by statistically analyzing the coverage segment sequence number and the current path segment identifier. Used to represent the same The process law shows that the contribution of newly added effective processing gradually decreases after repeated coverage in a continuous path. Therefore: ; in, For local polishing unit The local valid processing values ​​under the current path segment are recorded and written into the local valid processing status table. The path action completeness obtained by S2 is jointly derived from path coverage, contact execution sufficiency, and sensitive over-reference coverage. This represents the local incremental proximity of the current path segment. This is the path action type coefficient, which is obtained by converting the path action type in S2. When the path action type is a complete action, a higher value is taken; when the path action type is a low-pressure action or a short-stay action, a middle value is taken; and when the path action type is a boundary rubbing action or a confined area action, a lower value is taken. This conversion relationship is configured by the equipment process parameter table. For the same local polishing unit The count of consecutive coverage segments formed in the current path segment and adjacent path segments is obtained by counting the coverage segment number of S2 and the current path segment identifier. , , and All are calculated in a proportionalized or counted state, and the results are... This represents the relatively effective processing contribution of the current path segment to the local polishing unit. The derivation of this formula is as follows: Indicates the fact of path action. This indicates the quality improvement result brought about by the current path segment. The path effect cause correction is represented by the product of the three factors, which yield the basic effective processing contribution; the denominator is the result of the product of the three factors. This indicates duplicate coverage suppression, when the same When the number of consecutive coverages increases and the integrity of each path action is high, the new The value is lowered to express the decrease in the benefit of repeated coverage. Taking the aforementioned ordinary bottom element as an example, if S2 obtains... This step yields The path action type is a complete action and This is the first consecutive coverage segment. ,but If the same cell is covered again in an adjacent path, This step yields The number of consecutive covered segments is ,but This indicates that while the second coverage still has a strong path effect, the newly generated effective processing contribution is significantly reduced. For elements near the fractal line, if , The path action type is boundary scrambling and Count of consecutive covered segments ,but The results indicate that the area exhibits repeated grazing with minimal improvement, and subsequent path control should target this area for reduced coverage or detours.

[0047] The value of is also limited to to The proportionalization coefficient is uniquely determined by the path action type, thus avoiding the same record being interpreted repeatedly in subsequent evaluations; It is a non-negative integer count, derived from the cover segment number and path segment identifier, and does not represent time or length units. It is not the amount of material removed, the reduction in roughness, or the actual processing time, but rather the relatively effective processing contribution used for subsequent path control; therefore, its dimensions are... , and Maintain consistency. This value is accumulated in this step as the effective processing level. In S4, it is decided whether the candidate polishing path should be retained or compressed to avoid computational errors. However, it was not used subsequently.

[0048] This step follows the procedure after each current path segment ends. The effective processing sequence of the same local polishing unit is organized by covering the segment number and the current path segment identifier. For the same If multiple coverage segments within the same path segment have already been merged into the same consecutive action record using S2, then this step uses the corresponding merged record. , and A valid processing judgment is made for this path segment; if the covered segments are recorded separately, then each is calculated separately. And arranged according to the coverage segment number. For adjacent path segments that continuously cover the same In this case, this step retains the information for each instance. The changing trend. If and All remained at a high level, indicating that this path segment brought about a clear quality improvement; if Still relatively high and A decrease indicates that the region has entered a state of diminishing returns; if the path action type continuously exhibits boundary grazing or restricted area action, and If the value remains consistently low, the area is recorded as boundary repetitive inefficiency. This judgment directly corresponds to a typical situation in mold polishing: the tool marks on the bottom surface of a normal cavity significantly weaken after the first few passes of the tool path. The processing time is relatively high; even after the surface is close to the target, re-covering will still consume processing time, but... Decrease; areas near rounded corners or fractal lines, even if only rubbed multiple times by the tool edge, will form a low-yield repetitive state due to the increased count of consecutive covered segments.

[0049] This step is based on , The path action type, covered segment number, and current path segment identifier generate a locally effective processing state. This state includes effective processing degree, quality improvement state, and low-return repetition state. The effective processing degree is determined by the same... In the current and historical path segments Accumulated sequentially along the path, a proportional state is formed. Once the maximum processing limit is reached, the process remains at the limit, retaining each... The corresponding current path segment identifier; the quality improvement status is determined by... Its changing trend is determined, when the current path segment is followed by Maintaining a high level is recorded as a quality improvement. Lower and When the level remains high, it is recorded as a weakening of quality improvement, especially when the path action type is low-pressure action or short-dwelling action. When the level is low, it is recorded as insufficient inefficiency; the low-return repetitive state is... The trend of change, the sequence number of the covered segment, and the number of consecutive covers are jointly determined when the same consecutive low Furthermore, when the number of covered segments increases, it is recorded as a low-yield repetitive state. When this low-yield repetitive state is accompanied by boundary swiping or restricted area effects, it is recorded as boundary repetitive inefficiency. Through this state classification, subsequent steps can distinguish between "areas that need to continue normal processing", "areas that are close to the target and whose repeated coverage benefits are reduced", "areas where sensitive boundaries are repeatedly swiped", and "areas that have not been effectively processed due to pressure or insufficient dwell time".

[0050] This step outputs a table of localized effective processing status. This table is based on localized polishing units. For indexing, each record includes , , The data includes: effective processing level, quality improvement status, low-yield repetition status, covered segment sequence number, and current path segment identifier. Among these, From the path action record table and consistent with the local polishing unit of S1; Surface quality before the current path segment acts Surface quality after the current path segment is applied Target surface quality and S2 get; From S2 This step Path action type corresponding and the results obtained from the statistics of the covered segment numbers Obtained; effective processing degree from the same of The quality improvement state is obtained by accumulating and converting the path segments sequentially into a proportional state; the quality improvement state is obtained by... The changing trend is obtained; the low-yield repetitive state is obtained from The change trend, the covered segment number, and the number of consecutive covers are used to determine the status; the covered segment number and the current path segment identifier are inherited from the record in S2. Therefore, the path action record table is transformed into a locally effective processing status table, and the next step can directly read each... The effective degree of processing, the state of quality improvement, and the state of low-yield repetition determine whether the subsequent mold polishing path should be retained, compressed, bypassed, or the processing intensity reduced.

[0051] S4: Read the candidate polishing paths to be executed. For each local polishing unit covered by the candidate polishing path, calculate the subsequent path retention coefficient based on the local effective processing state table. Determine the control action of the candidate polishing path on the corresponding local polishing unit based on the subsequent path retention coefficient. Adjust the candidate polishing path according to the control action on each local polishing unit to generate the adjusted subsequent mold polishing path and process parameters. Optimize the processing efficiency based on the subsequent mold polishing path and process parameters, specifically including: This step follows up on the local effective processing status table formed in S3, and includes the local polishing unit. The effective processing results are converted into actual path control actions in the subsequent mold polishing process. S3 output , , The following parameters are used in this step: effective processing level, quality improvement status, low-yield repetition status, covered segment sequence number, and current path segment identifier: Used to determine spatial objects that need to be retained, compressed, bypassed, or have their processing intensity reduced for subsequent paths; Used to indicate the approximate incremental quality contribution of the current path segment to the local unit; The following parameters are used to characterize the new effective processing contribution of the current path segment to the unit: Effective processing level is used to determine whether the cumulative effective processing received by the unit has been sufficient; Quality improvement status is used to distinguish whether the unit is still improving, the improvement is weakening, or the effect is insufficient; Low-yield repetition status is used to identify whether continuing to cover the unit will cause time waste or boundary over-polishing risk; Coverage segment sequence number is used to determine the order in which the same unit is repeatedly covered in a continuous path; Current path segment identifier is used to implement control actions to specific subsequent candidate polishing path segments. Through this step, the status judgment in S3 is converted into a polishing path adjustment result that can be executed by the equipment, so that local areas that still have effective processing value continue to receive processing, and local areas that have been fully processed, have decreased yield, or have been repeatedly affected by sensitive boundaries are reduced from repeated coverage.

[0052] During subsequent path control, the system reads the next segment or several candidate polishing paths to be executed, and, following the same path mapping method as S2, pre-projects the tool contact area of ​​the candidate polishing paths onto the local polishing unit established in S1. The system obtains the set of local units that the candidate polishing path will cover. For each candidate polishing path segment, the system determines the local unit set to be covered based on its expected coverage area. Retrieve the local effective processing state table from S3 of , The parameters include effective processing level, quality improvement status, low-yield repetitive status, covered segment number, and current path segment identifier. In this way, a candidate polishing path can produce different control results in different local regions. For example, when a path crosses the cavity bottom surface and the rounded corner boundary, if the cavity bottom surface unit is still in a continuously improving state, the candidate polishing path is retained in that region; if the rounded corner boundary unit has already shown boundary repetitive inefficiency, the candidate polishing path reduces pressure, shortens dwell time, or locally detours in that region. For multiple consecutive candidate polishing paths covering the same... In such cases, the system identifies whether the coverage is a continuous repeated coverage based on the coverage segment number and the current path segment identifier, and incorporates it into the calculation of subsequent control strength.

[0053] This step records the effective processing level in S3 as... ,in For the same Local effective processing value The proportionalized effective processing level, accumulated sequentially along the path, is maintained at the upper limit of full processing after reaching it; the quality improvement status is then converted into an improvement retention value. ,in Only representing the trend of quality improvement, a higher value is used for continuous improvement, a middle value is used for weakening improvement, and a lower value is used for inefficiency due to insufficient effect; low-return repetitive states are converted into repetition inhibition values. ,in This only represents the degree of low benefit caused by repeated coverage. Lower values ​​are used for states with no low-return repetition, higher values ​​for consecutive low-return repetition, and even higher values ​​for boundary-repetitive inefficiency. This conversion can be accomplished using equipment process parameter tables; for example, higher values ​​correspond to states of continuous improvement. Improving the weakening state corresponds to the middle. Insufficient function corresponds to lower efficiency No low-yield repetitive states correspond to lower... A continuous low-yield repetition state corresponds to a higher Boundary repetitive inefficiency corresponds to higher Then, the subsequent path retention coefficient is calculated. , used to indicate the candidate polishing path in the local polishing unit The formula initially derives from the proportional adjustment approach in control engineering and the normalized weighted evaluation method in mathematics, which determines subsequent control quantities based on the gap between the current state and the target state. This application introduces a low-return repetition suppression term to automatically reduce the intensity of subsequent paths in regions approaching full processing and where repetition returns are decreasing, while retaining regions that still contribute to quality improvement and are in a state of continuous improvement. ; in, For candidate polishing paths in local polishing units The subsequent path retention coefficient is used to determine whether the dwell time, contact pressure, or path coverage of the corresponding path segment of the unit is retained. The proportionalized effective processing degree obtained for S3 is derived from the same Local effective processing value Formed by accumulating in the order of the path; The improvement retention value corresponding to the quality improvement state is obtained from the quality improvement state transition recorded in S3. The local incremental proximity obtained by S3; The repetition suppression value corresponding to the low-yield repetitive state is obtained from the low-yield repetitive state transition recorded in S3. , , and All calculations use proportionalized states. In this formula, This indicates the processing strength that should be retained when effective processing is not yet sufficient and quality is still improving. This indicates the supplementary processing intensity that should be retained when the current path segment still generates incremental quality that is similar and the repetitive low returns are not significant. The outer layer value restriction makes... It can be directly used as the proportion for subsequent path control. The calculation chain of the formula forms a continuous relationship with S3: S3 is first obtained from the path action completeness and incremental proximity. Then by Accumulated And obtained from the quality improvement state and the low-return repetitive state respectively. and This step will then... , , and synthesis This allows the local effective processing state to enter the actual path control.

[0054] in, , and The conversion table is pre-configured by the equipment process parameter table and associated with the mold area type; all three are limited to... to ,therefore It is also a dimensionless proportional value. It does not directly represent a specific absolute pressure or absolute speed, but rather serves as a proportion or classification basis for subsequent path segment control parameters; for example, dwell time can be calculated as... The ratio is shortened, and the pressure can be adjusted accordingly. The path coverage can be adjusted by selecting whether to maintain, reduce, or bypass the specified interval. The decision to retain, compress, or detour a state with low returns is determined by the repetitive state. This ensures the formula's results can be explicitly used by equipment instructions, rather than remaining at an abstract scoring level.

[0055] Taking a typical local polishing unit on the bottom surface of a cavity as an example, if the effective machining degree of this unit... The quality improvement status is continuous improvement and has been transformed into... Local increment proximity Low-yield repetitive states are weak and transform into ,but This indicates that the unit retains its normal path function, and the equipment executes the process according to the original candidate polishing path's dwell time and contact pressure. If the effective processing degree of a certain rounded corner unit... The quality improvement status is that the improvement is weakening and then transforms into... Local increment proximity The low-yield repetitive state is obvious and transforms into ,but This indicates that the subsequent path of this unit should be significantly compressed or detoured. If the effective processing degree of a sidewall unit... The quality improvement state is characterized by insufficient effectiveness and inefficiency, and is thus transformed into... Local increment proximity Low-yield repetitive states are weak and transform into ,but This indicates that the current path retention coefficient in this area is low, but there is still unfinished processing demand. Subsequent control measures will include it in the re-action area based on the reasons for its insufficient effect, rather than treating it as a fully processed area.

[0056] In obtaining Then, the system converts it into a candidate polishing path in the corresponding The specific control actions on it. For For higher-level local polishing units, the candidate polishing path coverage on that unit maintains the original path form and is executed according to the contact pressure and residence time set in the current process procedure; for For localized polishing units in the intermediate range, the candidate polishing path retains space for interaction, but the dwell time on the unit is shortened or the contact pressure is reduced, allowing it to continue improving in a lighter manner; for For localized polishing units with significant low-yield repetitive states, the system locally compresses or detours candidate polishing paths, ensuring the tool contact area avoids the unit or passes through it with only a very short stop; for A local polishing unit with a lower but improved quality status is considered an inefficient local polishing unit with insufficient effect. The system includes this unit in the re-application area and adjusts subsequent application conditions according to the insufficient effect type recorded in S3. For example, low-pressure application corresponds to improved contact stability, and short-dwell application corresponds to extended necessary dwell time or reduced local feed speed. The above control actions directly affect subsequent path segments: path compression is achieved by deleting candidate polishing paths that cover this... The path is circumvented by reconnecting the repeated sampling segments on the local surface. The path points on both sides are achieved. The contact pressure is adjusted by modifying the pressure setting of the path segment in the equipment control program. The dwell time is adjusted by modifying the feed speed or local dwell command of the path segment.

[0057] When a candidate polishing path spans multiple local polishing units, the system classifies adjacent polishing paths according to the path segment identifier. of Sequential processing is performed to ensure that subsequent paths reflect local state differences while maintaining the continuity of executable paths for the device. If a candidate polishing path passes through multiple... For higher local units, the path segment continues to execute continuously; if a small number of units appear in the middle of the path... For lower local units, the system employs methods such as local acceleration, decompression, or contact range offset to control both path continuity and local over-throw risk; if the path continuously passes through multiple For local units exhibiting significant low-yield repetitive states, the path segment is deleted or the path is routed around the region. For boundary-repetitive inefficient local units, the system prioritizes rerouting, reducing pressure, or altering the tool contact range. For low-yield repetitive states in ordinary, flat regions, the system prioritizes compressing path density or shortening dwell time. For underutilized inefficient regions, the system retains their processing requirements and adjusts their subsequent path segments to more stable contact methods. In this way, ordinary regions, boundary regions, and underutilized regions receive different controls, and candidate polishing paths are differentiated based on their local effective processing state before execution.

[0058] At the device execution level, the system will Write the control parameter table for subsequent path segments, and generate the adjusted subsequent mold polishing path and process parameters. If the candidate polishing path passes through a certain... The original residence time is given by the equipment process program, then the actual residence time is executed according to... Adjust the original dwell time proportionally; if the candidate polishing path passes through a certain The original contact pressure is given by the equipment process program, then the actual execution pressure is according to... The interval is adjusted in stages, for example Maintain the original pressure when it is higher. Reduce stress during moderate periods. When inefficiencies are low and involve boundary repetition, the tool is redirected or bypassed non-contactly. For path breaks caused by compression or detours in candidate polishing paths, the system generates transition segments between adjacent retained path points. These transition segments are smoothly connected according to the allowable angle and attitude changes of the equipment, enabling continuous execution by the robot or CNC polishing equipment. For local units included in the re-action area, the system retains them in subsequent paths. The processing requirements are identified, and the reason for inefficiency is indicated as low-pressure action or short-stay action, so that the action conditions can be adjusted according to the cause during subsequent execution.

[0059] The actual pressure and residence time retain their original physical units before being output to the equipment. For example, the pressure setting, feed rate, and residence time in the original process program are first determined by this step. The system generates adjustment ratios or graded actions, which are then converted by the equipment control program into specific pressure setpoints, speed commands, and dwell commands. Path detours or path compressions output new path point sequences, path segment identifiers, and transition sections. If the adjusted pressure, speed, or path angle exceeds the equipment's allowable range, the system performs safety actions such as limiting the amplitude or using pressure reduction or local detours according to equipment constraints. This ensures that the optimization results reflect both the effective local processing state and can be executed by the actual polishing equipment.

[0060] Taking a candidate polishing path that spans the bottom surface of the cavity and the rounded corner boundary as an example, if the bottom surface unit... near This path maintains normal contact pressure and feed rate in the bottom area; if the rounded corner unit... near Furthermore, if the low-yield repetitive state is a boundary repetitive inefficiency, the system will reduce pressure in advance near the rounded corner of the path and cause the tool contact range to deviate from the rounded corner, or directly bypass the rounded corner unit; if a certain sidewall unit If the path is in the middle range, indicating that it is close to the target but still has slight room for improvement, the system retains the path and increases the feed rate to shorten the dwell time. Through this process, the same candidate polishing path is split into a normal execution segment, a reduced execution segment, and a detour segment, ultimately forming a subsequent polishing path that can be executed by the equipment.

[0061] This step outputs the adjusted and executed subsequent mold polishing path and process parameters, as well as the corresponding local polishing units. The corresponding control action records. The adjusted and executed subsequent mold polishing path and process parameters include the retained path segment, compression path segment, bypass path segment, adjusted contact pressure, adjusted dwell time, and re-application area; the control action records include... , The corresponding control action, the identifier of the adjusted path segment, and the identifier of the executed device command. Among these, Local effective processing state table from S3 Depend on , , and The calculation shows that the corresponding control action is... The scope, quality improvement status, and low-yield repetitive status are jointly determined. The adjusted path segment identifier adopts the current path segment identifier and corresponds to the subsequent candidate polishing path segment. The equipment instruction identifier is generated by the polishing equipment control program after the path adjustment. As a result, the local effective processing status table is transformed into the actual execution of subsequent path control results. Repetitive invalid coverage, excessive boundary rubbing, and low-yield continued processing in the mold polishing process are reduced, and local areas that still have effective processing value continue to receive necessary processing.

[0062] A system corresponding to the above method may include a local element creation module, an initial quality mapping module, a path action recording module, an effective processing evaluation module, a subsequent path control module, and an equipment instruction output module. The local element creation module is used to read the mold CAD model, scanned point cloud, and process area identifiers, and generate a mapping function bound to the mold surface coordinates. The initial quality mapping module is used to convert roughness detection, visual texture detection, or shape detection results into... , , and The path action recording module is used to receive data from the motion controller, force control module, and process program, and generate... , , , Path function type and coverage segment number; the effective processing evaluation module is used to receive the retest quality and generate , Effective processing level, quality improvement status, and low-yield repetitive status; the subsequent path control module is used to calculate... This generates control actions such as retention, compression, detour, pressure reduction, shortened dwell time, and re-action. The equipment instruction output module converts these control actions into executable path points, pressure settings, feed rates, dwell commands, and execution logs for the robot or CNC polishing equipment. All data objects in each module are... It is associated with path segment identifiers to ensure that there is a correspondence between method steps and system structure.

[0063] In one specific implementation, the system first divides the bottom surface, sidewalls, fillets, and near the parting line of an injection mold cavity into local units, and then converts the visual texture detection results before polishing into... Convert the target mirror finish grade or target roughness grade in the process document to During robotic polishing, the controller provides the path segment and orientation, while the constant force actuator provides the actual contact pressure. The system is derived from the overlap relationship between the tool contact range and local units. It is obtained from the pressure achievement ratio and the residence achievement ratio. After the current path segment is completed, the system re-measures the surface quality of the corresponding local unit and calculates... and Furthermore, units that are still undergoing improvement on the ordinary bottom surface are retained for normal path operation, while low-yield repetitive units near the parting line are converted to pressure reduction or bypass operations, and units on the sidewalls that are insufficient in effect due to short dwell times are included in the re-operation area. In this embodiment, the initial state table, path operation record table, local effective processing state table, and subsequent path control results can all be traced from the same equipment log, which can explain how this solution reduces invalid coverage and repeated boundary wiping in the actual polishing process.

[0064] The efficiency optimization described in this invention does not merely manage manual scheduling or time statistics, but rather utilizes mold geometry data, detection data, equipment motion data, and force control data to form a closed-loop control process that can be executed by the polishing equipment. The proportionalized states in each step are derived from actual detection results, process documents, or equipment controller data, and are continuously transmitted within the same local polishing unit Ui, avoiding situations where a variable is only used for intermediate calculations but not entered into subsequent control. The final output path compression, detour, pressure adjustment, dwell time adjustment, and re-action area all correspond to the path segments and process parameters of the robot, CNC polishing equipment, or constant force polishing mechanism; therefore, this application can be applied to the execution level of specific processing equipment.

[0065] refer to Figure 2 As shown, another aspect of this application embodiment proposes a system for optimizing the processing efficiency of a mold polishing process, characterized by comprising: The status table construction module is used to acquire detection data of the mold surface to be polished, determine the effective range of the mold surface to be polished based on the detection data, divide the effective range into multiple local polishing units, assign a unit identifier to each local polishing unit, calculate the boundary sensitivity level of the local polishing unit, determine the initial processing allowable state of each local polishing unit according to the unit identifier and boundary sensitivity level of each local polishing unit, and generate the local polishing unit status table based on the initial processing allowable state of all local polishing units. The action record table construction module is used to acquire path execution data, unify the path execution data into the mold coordinate system, map the path sampling points to the local polishing unit, determine the actual action area of ​​the current path on the local polishing unit, calculate the coverage ratio of the actual action area, calculate the local action reference coefficient according to the local polishing unit status table, calculate the path action integrity based on the coverage ratio and the local action reference coefficient, determine the path action type of the local polishing unit, and summarize the local polishing unit status table, local action reference coefficient, path action integrity and path action type to construct a path action record table; The processing status table construction module is used to obtain the surface quality detection results of each local polishing unit after the current path segment is applied, determine the local polishing unit that needs to be retested based on the local polishing unit in the path application record table and the current path segment identifier, calculate the local incremental proximity of the local polishing unit based on the surface quality detection results, calculate the local effective processing value based on the local incremental proximity, and construct a local effective processing status table based on the local incremental proximity and the local effective processing value. An efficiency optimization module is used to read candidate polishing paths to be executed, calculate a subsequent path retention coefficient for each local polishing unit covered by the candidate polishing path based on the local effective processing state table, determine the control action of the candidate polishing path on the corresponding local polishing unit according to the subsequent path retention coefficient, adjust the candidate polishing path according to the control action on each local polishing unit, generate the adjusted subsequent mold polishing path and process parameters, and optimize the processing efficiency based on the subsequent mold polishing path and process parameters.

[0066] The following specific examples illustrate this application: In the verification of robotic constant-force polishing of an injection mold cavity, four local records near the same candidate path were selected for recalculation. The verification objects included initial coverage of the bottom surface of a common cavity, repeated coverage of the bottom surface of a common cavity, coverage of the fillet near the parting line, and short-dwell coverage of the deep cavity sidewall. The polishing equipment recorded the tool coverage ratio, pressure and dwell time, original dwell time, and contact pressure for each path segment. The re-testing device re-tested the surface quality of the corresponding local polishing unit after the path segment ended and converted the re-test results into a proportionalized quality value consistent with the aforementioned formula. In this embodiment, qipre, qipost, and All values ​​are proportional values, with larger values ​​indicating closer proximity to the target quality. Ci, Ri, ri, gi, hi, Ai, Di, and Bi are obtained from the sources described in the preceding steps or converted from the process parameter table. Referring to Table 1, after recalculation according to the aforementioned formula, the following local states and subsequent control results are obtained: Table 1 Local state and subsequent control results ; As can be seen from the above examples, during the initial coverage of a normal bottom surface, both the path integrity and the local effective processing value are high, and the system retains a high level of subsequent path strength. When the same bottom surface area is covered again, the path integrity is still close to the initial coverage, but the local incremental proximity and local effective processing value are significantly reduced, and the system accordingly compresses repeated dwell times. Although the coverage area of ​​the rounded corner area near the parting line is small, the boundary sensitivity level is high and the subsequent path retention coefficient is low, so the system treats it as an object for reducing pressure or local detour. The low effective processing value of the deep cavity sidewall area mainly comes from insufficient contact execution, so the system does not simply delete the path in this area, but retains its unfinished processing requirements and includes it in the re-action area. This result is consistent with the technical objective of this invention, namely, to distinguish between different states of "path passed," "actually acted upon," "formed effective processing," and "requires continued control," thereby reducing invalid repeated coverage and lowering the risk of boundary over-throw.

[0067] refer to Figure 3 As shown, Figure 3 This shows the local effective processing values ​​and subsequent path retention coefficients for the four local records mentioned above. (First coverage of the ordinary bottom surface) for The value was significantly higher than the other records, indicating that this path segment had a high effective processing contribution; the repeated coverage of the ordinary bottom surface, the fillet near the parting line, and the short dwell time on the deep cavity sidewalls... They are respectively , and All were significantly lower than initial coverage. Combined It can be further observed that the subsequent path retention coefficients corresponding to repeated coverage and fillet radius adjacent to the parting line are reduced to [values ​​to be filled in]. and This demonstrates that the system can compress or bypass low-yield repetitive coverage and boundary-sensitive coverage; short dwell time on the deep cavity sidewalls. for Although its retention coefficient is low, because the quality improvement state is identified as insufficient, subsequent control will include it in the re-processing area, rather than misjudging it as a fully processed repetitive area. Therefore, this embodiment supports the present invention's ability to improve mold polishing efficiency and control over-polishing risk through locally effective processing states.

[0068] The above embodiments are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for optimizing the processing efficiency of mold polishing process, characterized in that, include: Acquire detection data of the mold surface to be polished, determine the effective range of the mold surface to be polished based on the detection data, divide the effective range into multiple local polishing units, assign a unit identifier to each local polishing unit, calculate the boundary sensitivity level of the local polishing unit, determine the initial processing allowable state of each local polishing unit according to the unit identifier and boundary sensitivity level of each local polishing unit, and generate the local polishing unit state table based on the initial processing allowable state of all local polishing units. Acquire path execution data, unify the path execution data to the mold coordinate system, map the path sampling points to the local polishing unit, determine the actual action area of ​​the current path on the local polishing unit, calculate the coverage ratio of the actual action area, calculate the local action reference coefficient according to the local polishing unit status table, calculate the path action integrity based on the coverage ratio and the local action reference coefficient, determine the path action type of the local polishing unit, and summarize the local polishing unit status table, local action reference coefficient, path action integrity and path action type to construct a path action record table; After obtaining the surface quality detection results of each local polishing unit after the current path segment is applied, the local polishing units that need to be retested are determined according to the local polishing units in the path application record table and the current path segment identifier. The local incremental proximity of the local polishing unit is calculated based on the surface quality detection results. The local effective processing value is calculated based on the local incremental proximity. A local effective processing status table is constructed based on the local incremental proximity and the local effective processing value. The candidate polishing path to be executed is read. For each local polishing unit covered by the candidate polishing path, the subsequent path retention coefficient is calculated based on the local effective processing state table. The control action of the candidate polishing path on the corresponding local polishing unit is determined according to the subsequent path retention coefficient. The candidate polishing path is adjusted according to the control action on each local polishing unit to generate the adjusted subsequent mold polishing path and process parameters. The processing efficiency is optimized based on the subsequent mold polishing path and process parameters.

2. The method for optimizing processing efficiency based on mold polishing process according to claim 1, characterized in that, The detection data includes three-dimensional geometric data, polishing process targets, and initial surface quality detection results. The initial surface quality detection results are obtained by contact roughness detection, white light interferometry detection, structured light morphology detection, or visual texture detection.

3. The method for optimizing processing efficiency based on mold polishing process according to claim 2, characterized in that, When the initial surface quality detection results exist in the form of multiple sampling points, each sampling point is mapped to its corresponding local polishing unit; When the initial surface quality detection result is a region image, the region image is first matched with the position of the three-dimensional curved surface, and then the initial quality state at the corresponding position is read.

4. The method for optimizing processing efficiency based on mold polishing process according to claim 1, characterized in that, The coverage ratio is obtained by the surface overlap relationship between the tool contact range within the current path segment and the local polishing unit; When consecutive sampling times cover the same local polishing unit, the overlap of each sampling time is merged according to the temporal continuity to form a proportional coverage of the local polishing unit.

5. The method for optimizing processing efficiency based on mold polishing process according to claim 1, characterized in that, The completeness and type of the path action are recorded using the local polishing unit as an index.

6. The method for optimizing processing efficiency based on mold polishing process according to claim 1, characterized in that, The control action for determining the candidate polishing path on the corresponding local polishing unit based on the subsequent path retention coefficient specifically includes: When the subsequent path retention coefficient is higher than the first threshold, the control action is to retain the path and execute according to the contact pressure and residence time set in the current process. When the subsequent path retention coefficient is between the first threshold and the second threshold, the control action is to shorten the dwell time or reduce the contact pressure. When the subsequent path retention coefficient is lower than the second threshold and the low-return repetition state is continuous low-return repetition or boundary repetition type inefficiency, the control action is local compression or detour. When the subsequent path retention coefficient is lower than the second threshold and the quality improvement status is inefficient due to insufficient effect, the control action is to include it in the re-action area and adjust the subsequent action conditions according to the type of insufficient effect.

7. The method for optimizing processing efficiency based on mold polishing process according to claim 1, characterized in that, The subsequent mold polishing path and process parameters include the retained path segment, the compressed path segment, the bypass path segment, the adjusted contact pressure, the adjusted dwell time, and the re-action area.

8. The method for optimizing processing efficiency based on mold polishing process according to claim 1, characterized in that, The sensitivity level is calculated and generated using a scaled proximity value, a scaled surface change value, a process sensitivity identifier, and a preset weight.

9. The method for optimizing processing efficiency based on mold polishing process according to claim 1, characterized in that, The local effective processing status table is constructed using local polishing units as indexes. The local effective processing status table records local polishing units, local incremental proximity, local effective processing values, effective processing degree, quality improvement status, low-yield repetition status, covered segment sequence number, and current path segment identifier.

10. A system for optimizing the processing efficiency of mold polishing process, characterized in that, include: The status table construction module is used to acquire detection data of the mold surface to be polished, determine the effective range of the mold surface to be polished based on the detection data, divide the effective range into multiple local polishing units, assign a unit identifier to each local polishing unit, calculate the boundary sensitivity level of the local polishing unit, determine the initial processing allowable state of each local polishing unit according to the unit identifier and boundary sensitivity level of each local polishing unit, and generate the local polishing unit status table based on the initial processing allowable state of all local polishing units. The action record table construction module is used to acquire path execution data, unify the path execution data into the mold coordinate system, map the path sampling points to the local polishing unit, determine the actual action area of ​​the current path on the local polishing unit, calculate the coverage ratio of the actual action area, calculate the local action reference coefficient according to the local polishing unit status table, calculate the path action integrity based on the coverage ratio and the local action reference coefficient, determine the path action type of the local polishing unit, and summarize the local polishing unit status table, local action reference coefficient, path action integrity and path action type to construct a path action record table; The processing status table construction module is used to obtain the surface quality detection results of each local polishing unit after the current path segment is applied, determine the local polishing unit that needs to be retested based on the local polishing unit in the path application record table and the current path segment identifier, calculate the local incremental proximity of the local polishing unit based on the surface quality detection results, calculate the local effective processing value based on the local incremental proximity, and construct a local effective processing status table based on the local incremental proximity and the local effective processing value. An efficiency optimization module is used to read candidate polishing paths to be executed, calculate a subsequent path retention coefficient for each local polishing unit covered by the candidate polishing path based on the local effective processing state table, determine the control action of the candidate polishing path on the corresponding local polishing unit according to the subsequent path retention coefficient, adjust the candidate polishing path according to the control action on each local polishing unit, generate the adjusted subsequent mold polishing path and process parameters, and optimize the processing efficiency based on the subsequent mold polishing path and process parameters.

Images