A PMI-driven method for automatic 3D tolerance modeling
By generating structured PMI data and assembly constraints, distinguishing between persistent constraints and state-dependent constraints, and establishing the activation relationship between PMI and assembly constraints, the problem of assembly constraint state changes caused by PMI changes is solved, thus achieving the continuity and accuracy of the three-dimensional tolerance analysis model.
Patent Information
- Application Number
- CN202611618851.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-20
- Publication Date
- 2026-11-17
AI Technical Summary
In existing 3D tolerance modeling processes, when changes in PMI lead to changes in assembly constraint states, it is difficult to maintain a continuous correspondence between assembly constraints and product functional movements. Furthermore, improper selection of assembly constraints may affect the accuracy of the analysis results.
By acquiring the PMI and assembly relationships in the 3D assembly model, structured PMI data is formed, PMI variation parameters are determined, assembly constraints are formed based on the assembly relationships, persistent constraints and state-dependent constraints are distinguished, activation relationships between PMI and assembly constraints are established, the final set of assembly constraints is screened, a 3D tolerance analysis model is generated, and assembly constraints are updated to maintain the continuity of the model.
It establishes the correspondence between PMI tolerance requirements and the current geometric state, maintains the continuity between assembly constraints and product functional movement, and ensures the accuracy and consistency of the three-dimensional tolerance analysis model.
Smart Images

Figure CN123116508A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of three-dimensional tolerance analysis technology. More specifically, this invention relates to a PMI-driven automatic three-dimensional tolerance modeling method. Background Technology
[0002] Three-dimensional assembly models can record manufacturing requirements such as part dimensions and geometric tolerances through PMI (Product Manufacturing Information), and describe the fit relationships between different part geometric features through assembly relationships. When performing three-dimensional tolerance analysis, the tolerance requirements in the PMI need to be converted into analytical inputs that can reflect changes in geometric features, and the impact of tolerance changes on part pose and assembly results needs to be calculated in conjunction with assembly relationships.
[0003] For assembly structures involving gaps and contact relationships, the restrictive effect of these relationships on relative motion varies with the geometric state. When two geometric features have a gap, relative motion can occur in the corresponding direction within the gap; once the contact boundary is reached, motion in the corresponding direction is restricted; and this restriction may be lifted when subsequent pose changes cause the geometric features to separate. Therefore, changes in size, position, orientation, or surface position caused by PMI not only alter the geometric features themselves but may also change the constraint relationships involved in solving the current assembly state.
[0004] In existing 3D tolerance modeling processes, if the solution is performed based on nominal assembly relationships or pre-fixed constraint sets, assembly constraints are difficult to adjust synchronously with clearance, contact, and separation states. If the assembly relationships selected for the solution are based solely on the restrictive effect of assembly constraints on the degrees of freedom of motion, assembly relationships with the same motion constraints but affected by different PMIs may be merged or discarded, preventing some tolerance action paths from continuing to participate in the solution after changes in assembly state. Furthermore, the selection of assembly constraints must be coordinated with the relative motion that the product itself needs to retain; otherwise, the functional motion of the product may be treated as motion that needs to be restricted.
[0005] Therefore, in three-dimensional assembly structures where changes in PMI can cause changes in assembly constraint states, how to ensure that PMI tolerance input, geometric state, assembly constraints, and part pose solutions remain continuously corresponding to the current assembly state, and how to form a three-dimensional tolerance analysis model suitable for the current assembly state while preserving the product's functional motion, remains a technical problem that needs to be solved. Summary of the Invention
[0006] To address the aforementioned technical problems, this invention provides a PMI-driven automatic 3D tolerance modeling method, comprising: acquiring PMIs and assembly relationships between parts in a 3D assembly model; forming PMI structured data based on the PMIs; and forming candidate assembly relationships based on the assembly relationships; forming PMI variation parameters representing geometric changes of the geometric features marked by the PMIs relative to the nominal state based on the PMI structured data; forming the current assembly state based on the PMI variation parameters; and forming assembly constraints that limit the relative motion of parts based on the candidate assembly relationships; determining the assembly constraints that continuously limit the relative motion of parts as persistent constraints and as active states; determining the assembly constraints that restrict relative motion in the corresponding direction after the corresponding geometric features reach the contact boundary as state-dependent constraints; and determining the state-dependent constraints as active or released states based on the gaps between the corresponding geometric features and the relative motion direction; and setting the constraint conditions for the currently active assembly constraints or the gaps and relative positions required to reach the contact boundary for the currently released state-dependent constraints based on the PMI variation parameters. The system establishes a PMI constraint activation relationship between the Product Assembly Management (PMI) and assembly constraints. Based on the product assembly function, a target permissible motion state is formed. Assembly constraints currently in effect are selected based on this target permissible motion state, retaining those with the same independent motion restrictions but different impacts of PMI variation parameters on the constraint conditions. State-dependent constraints currently in a released state and with PMI constraint activation relationships are also retained, forming a final assembly constraint set. A final effective tolerance variation propagation relationship is then formed based on the assembly constraints currently in effect and with PMI constraint activation relationships. A three-dimensional tolerance analysis model is generated based on the PMI variation parameters, the final assembly constraint set, and the final effective tolerance variation propagation relationship. The model calculates the part pose change, maintaining the corresponding values for pose components not determined by currently effective assembly constraints. When the current state of a state-dependent constraint changes, the assembly constraints involved in the solution and the final effective tolerance variation propagation relationship are updated based on the target permissible motion state until the current part pose corresponds to the current state of the state-dependent constraint, thus obtaining the three-dimensional tolerance analysis results.
[0007] Preferably, the PMI structured data includes the PMI identifier, tolerance type, and geometric features marked by the PMI; the data corresponding to dimensional tolerances also includes the nominal size and upper and lower deviations, or the upper and lower limit sizes; the data corresponding to geometric tolerances also includes geometric tolerance items, tolerance values, and the datum identifier and corresponding geometric features when datum requirements exist; the permissible dimensional changes, positional changes, directional changes, or surface position changes of the geometric features are determined based on the PMI structured data, and one or more values representing the corresponding changes are used to form PMI variation parameters; when the dimensional tolerance specifies the distance between two geometric features, the dimensional deviation of the current distance relative to the nominal distance is determined as the PMI variation parameter, and the PMI variation parameter is used as the relative positional change between the two geometric features along the measurement direction of the distance in the nominal state; when multiple PMI variation parameters jointly represent a PMI, their combination satisfies the permissible conditions of the corresponding tolerance zone of the PMI.
[0008] Preferably, for the state-dependent constraint, when there is a gap between two corresponding geometric features, it is determined to be in a released state; when the two geometric features reach the contact boundary and have a tendency to separate along the direction that increases the gap, it is determined to be in a released state; when it reaches the contact boundary and does not have the tendency to separate, it is determined to be in an active state; in the released state, when the current relative motion continues and will cross the contact boundary, the contact boundary is used as the state switching position to update to the active state; when performing a release judgment in the active state, the state-dependent constraint to be judged is not applied temporarily, and the part pose change is calculated based on the changes in the PMI variation parameters that have not yet been applied and the other assembly constraints that are currently in an active state, and the relative motion direction is determined based on the calculation result. When a tendency to separate along the direction that increases the gap appears, it is updated to a released state.
[0009] Preferably, for assembly constraints currently in an active state, a constraint Jacobian matrix is formed based on the first-order change relationship between the independent constraint conditions and the pose changes of the parts involved in the assembly constraint, and a PMI constraint influence matrix is formed based on the first-order change relationship between the independent constraint conditions and the change in the PMI variation parameters; for state-dependent constraints currently in a released state, a PMI constraint influence matrix is formed based on the first-order change relationship between the corresponding gap or the relative position required to reach the contact boundary and the change in the PMI variation parameters; when the rank of the PMI constraint influence matrix is greater than zero, a corresponding PMI constraint activation relationship is established; after a state-dependent constraint enters an active state, the corresponding constraint Jacobian matrix and PMI constraint influence matrix are re-formed based on the current assembly state when entering the active state.
[0010] Preferably, the independent motions in the target allowed motion state are arranged in the order of their degrees of freedom participating in the joint solution to form the target allowed motion basis matrix; for assembly constraints that are currently in effect and have PMI constraint activation, a constraint matrix is formed based on their first-order constraint relationship on each independent motion degree of freedom; for state-dependent constraints that are currently in effect, each independent motion in the target allowed motion state is judged one by one according to the current constraint direction; when there is an independent motion along the direction that reduces the corresponding gap, it is determined that the state-dependent constraint restricts the target allowed motion state; otherwise, it is determined that the state-dependent constraint does not restrict the target allowed motion state; for assembly constraints other than the state-dependent constraints that are currently in effect, when the product of the constraint matrix and the target allowed motion basis matrix is zero, it is determined that the assembly constraint does not restrict the target allowed motion state; for state-dependent constraints that are currently in the released state and have PMI constraint activation, their constraint information, PMI constraint activation relationship and current state are retained, and the judgment is repeated after they enter the effective state.
[0011] Preferably, the currently active assembly constraints that satisfy the target permissible motion state and are used to provide independent motion restrictions form a selection constraint matrix; when the rank of the selection constraint matrix increases after adding the assembly constraint to be judged, the assembly constraint to be judged is added to the selection constraint set; when the rank does not increase and the constraint matrix of the assembly constraint to be judged can be represented according to the linear relationship of the existing assembly constraints in the selection constraint set, a PMI constraint influence matrix is formed according to the first-order change relationship of the constraint conditions of the assembly constraint to be judged and the constraint conditions of the existing assembly constraints relative to the change amount of the PMI variation parameters, and the corresponding columns are arranged according to a unified PMI variation parameter order; the same set of linear combination coefficients of the linear relationship is used to determine whether the PMI constraint influence matrix maintains the same linear relationship. If it does, the existing assembly constraints represent the corresponding independent motion restrictions and the influence of the PMI variation parameters on the constraint conditions. If it does not, the assembly constraint to be judged is added to the selection constraint set and the corresponding PMI constraint activation relationship is retained, but it is not counted as a newly added independent motion restriction.
[0012] Preferably, when the set of selection constraints still needs to add independent motion restrictions, assembly constraints that satisfy the target allowed motion state and can increase the rank of the selection constraint matrix are selected from the remaining currently active assembly constraints that have not been added to the set of selection constraints as closed supplementary constraints, and added to the selection constraint matrix; when all independent motions other than the target allowed motion state need to be restricted, when the rank of the selection constraint matrix is equal to the difference between the number of degrees of freedom of independent motions participating in the joint solution and the number of independent motions retained in the target allowed motion state, the number of independent motion restrictions is determined to meet the requirements; when the independent motion corresponding to the state-dependent constraint currently in the released state does not need to be restricted by other currently active assembly constraints, the independent motion is not included in the currently restricted independent motions, and the corresponding pose components not determined by the currently active assembly constraints maintain the corresponding values of the current assembly state; for state-dependent constraints that only restrict one motion direction of the corresponding independent motion, when it increases the rank of the selection constraint matrix, it is counted as an independent motion restriction according to the current restriction direction, and its current state is re-evaluated in each assembly state solution stage.
[0013] Preferably, for PMI, a PMI variation source is established that records the PMI identifier, the labeled geometric features, the PMI variation parameters, and their allowable ranges. A currently effective global constraint matrix is formed based on the assembly constraints currently in effect in the final assembly constraint set. The PMI constraint influence matrices corresponding to each PMI variation parameter are then arranged in the same constraint order and PMI variation parameter order to form a currently effective global PMI constraint influence matrix. After giving the change in PMI variation parameters and updating the current state of the assembly constraints, a common solvability judgment is performed on all currently effective assembly constraints in the final assembly constraint set. If there is a part pose change that can jointly satisfy the assembly constraints, the part pose change is determined; otherwise, the corresponding PMI variation parameter combination is determined as a state where the assembly relationship cannot be simultaneously satisfied under the current constraint state. When a state-dependent constraint changes from a released state to an effective state, the corresponding PMI constraint activation relationship is added to the currently used final effective tolerance variation transmission relationship. When changing from an effective state to a released state, the use of the corresponding final effective tolerance variation transmission relationship is stopped.
[0014] Preferably, in each assembly state solution stage, the current assembly state at the start of that stage is used as the reference state; when the current state of a state-dependent constraint changes, the current assembly state at the time of the state change is used as the reference state for the next assembly state solution stage, and the solution relationship is re-formed using the PMI variation parameter changes that have not yet been applied relative to the reference state; when there are multiple PMI variation parameter changes, each PMI variation parameter changes synchronously from the current value corresponding to the reference state to the given value corresponding to this analysis in the same proportion, and the synchronous change proportion that first satisfies the state switching condition is determined, the changes that have occurred when the synchronous change proportion is reached are taken as the applied part, and the remaining changes are taken as the unapplied part; when multiple state-dependent constraints satisfy the state switching condition under the same synchronous change proportion, their current states are updated simultaneously until all the given PMI variation parameter changes are applied and the current pose of the part corresponds to the current state of each state-dependent constraint.
[0015] Preferably, based on the final obtained current pose of the part and the current position, direction, and size of the geometric features, the size, position, direction, or clearance result to be analyzed is determined; when performing extreme value analysis, PMI variation parameters are selected within the allowable range of PMI variation parameters to solve the assembly state; when performing statistical analysis, the probability distribution and distribution parameters are determined based on the manufacturing process statistical data of the geometric changes represented by the corresponding PMI, and the PMI variation parameters are obtained based on the probability distribution and distribution parameters, and the assembly state solution is repeated; when multiple PMI variation parameters jointly represent the same PMI, only the combination of PMI variation parameters that jointly satisfy the allowable conditions of the tolerance zone corresponding to the PMI is used.
[0016] The beneficial effects of this invention are as follows: 1. This invention generates PMI variation parameters based on PMI structured data and applies these parameters to the current size, current position, current orientation, surface position, or relative position between two geometric features. This allows PMI tolerance requirements to be converted into geometric change inputs that form the current assembly state, thereby maintaining the correspondence between the PMI tolerance input and the current geometric state.
[0017] 2. This invention updates the current state based on the gap, contact boundary and relative motion direction for state-dependent constraints. When making a release judgment in the active state, the state-dependent constraints to be judged are not applied. The part pose change is calculated based on the change of PMI parameters that have not yet been applied and the other assembly constraints that are currently in the active state. This makes the judgment of the separation motion trend not obscured by the limiting effect of the state-dependent constraints to be judged, so that the state switching is consistent with the solution conditions of the current assembly state.
[0018] 3. This invention filters assembly constraints while retaining the target's permissible motion state, and judges the independent motion restrictions provided by the assembly constraints and the impact of PMI variation parameters on the constraint conditions. Assembly constraints with the same independent motion restrictions but different impacts of PMI variation parameters on the constraint conditions are retained. At the same time, closed supplementary constraints are used to supplement the independent motion that needs to be restricted at present, and the pose components not determined by the assembly constraints that are currently in effect are kept at the corresponding values of the current assembly state, so that the target's permissible motion state, the impact of PMI variation parameters on assembly constraints, and the assembly state solution are consistent.
[0019] 4. After switching state-dependent constraints, this invention updates the assembly constraints involved in the solution and the final effective tolerance variation propagation relationship, and uses the current assembly state at the time of state switching as the new reference state, so that the unused parts of multiple PMI variation parameters continue to participate in the subsequent solution, avoiding the repeated entry of the already used changes into the subsequent assembly state solution, thereby making the PMI change, assembly constraint state change and the current solution relationship of the three-dimensional tolerance analysis model continuously correspond. Attached Figure Description
[0020] Figure 1 This is a flowchart illustrating a PMI-driven automatic tolerance modeling method according to the present invention; Figure 2 This is a schematic diagram illustrating the synchronous changes of multiple PMI parameters and the phased assembly state solution of the present invention. Figure 3 This is a schematic diagram of the assembly geometry between the first part and the second part in this invention; Figure 4 This is a schematic diagram illustrating the state-dependent constraint of the present invention switching from a released state to an active state. Detailed Implementation
[0021] 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, not all, of the embodiments of the present invention. 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.
[0022] This embodiment provides a PMI-driven automatic 3D tolerance modeling method. By acquiring tolerance information and assembly relationships between parts in a 3D assembly model, the tolerance requirements are converted into PMI variation parameters that can participate in geometric state updates. The current state of assembly constraints is determined based on the geometric state after PMI variation, and the interaction relationship between PMI variation parameters and assembly constraints is further determined. While retaining the motion required by the product assembly function, a final set of assembly constraints and a final effective tolerance variation transmission relationship are formed. A 3D tolerance analysis model is generated based on the formed data and constraint relationships.
[0023] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
[0024] See Figure 1 This is a flowchart of a PMI-driven automatic tolerance modeling method provided in an embodiment of the present invention, as shown below. Figure 1 As shown, the method may include: S1. Obtain the PMI and assembly relationships in the 3D assembly model to form PMI structured data and candidate assembly relationships.
[0025] A three-dimensional assembly model to be modeled for three-dimensional tolerances is obtained. The three-dimensional assembly model includes multiple parts and the position and orientation of each part in the assembly coordinate system. The position and orientation of a part in the assembly coordinate system are defined as the nominal pose of the part. The assembly state formed when multiple parts are in their respective nominal poses is defined as the nominal assembly state. The geometric features in the three-dimensional assembly model include, but are not limited to, planes, cylindrical surfaces, conical surfaces, spheres, axes, center planes, and contours.
[0026] The PMI processed in this embodiment is product manufacturing information in a three-dimensional assembly model that specifies dimensional tolerances or geometric tolerances for geometric features; the geometric tolerances include, but are not limited to, form tolerances, orientation tolerances, positional tolerances, runout tolerances, and profile tolerances.
[0027] For each PMI in the 3D assembly model, obtain the PMI identifier, tolerance type, and geometric features marked by the PMI; when the PMI is a dimensional tolerance, obtain the nominal size, upper deviation, and lower deviation corresponding to the dimensional tolerance, or obtain the upper limit size and lower limit size corresponding to the dimensional tolerance; when the dimensional tolerance specifies the distance between two geometric features, also obtain the measurement direction of the distance in the nominal state; when the PMI is a geometric tolerance, obtain the geometric tolerance items and tolerance values of the geometric tolerance; when the geometric tolerance includes datum requirements, also obtain the datum identifier in the geometric tolerance and the geometric feature corresponding to the datum identifier.
[0028] The acquired data forms PMI structured data, which includes PMI identifier, tolerance type, and geometric features marked by the PMI. When the PMI is a dimensional tolerance, the PMI structured data also includes the nominal size, upper deviation, and lower deviation corresponding to the dimensional tolerance, or the upper limit size and lower limit size corresponding to the dimensional tolerance. When the dimensional tolerance specifies the distance between two geometric features, it also includes the measurement direction of the distance in the nominal state. When the PMI is a geometric tolerance, the PMI structured data also includes the geometric tolerance items and tolerance values of the geometric tolerance, and when the geometric tolerance includes datum requirements, it also includes the datum identifier in the geometric tolerance and the geometric feature corresponding to the datum identifier.
[0029] PMI structured data organizes the tolerance requirements in different types of PMI into data items with clear data meanings, so that the PMI variation parameters formed later can be mapped to specific PMI, specific geometric features and corresponding tolerance requirements, and the size changes, position changes, orientation changes or surface position changes in PMI can participate in the subsequent geometric state update according to a unified data relationship.
[0030] The assembly relationships between parts recorded in the 3D assembly model are obtained. Each assembly relationship includes two parts involved in the assembly relationship, two geometric features belonging to the two parts respectively, and the relationship type between the two geometric features. The assembly relationship can be coaxial, coincident, parallel, perpendicular, distance, angular, surface contact, or shaft-hole fit. Each obtained assembly relationship is determined as a candidate assembly relationship.
[0031] For two parts whose assembly relationship is not recorded in the 3D assembly model, the assembly relationship between the two geometric features is determined based on the position and orientation of the geometric features of the two parts in the nominal assembly state. For example, when the inner cylindrical surface of one part mates with the outer cylindrical surface of another part and their axes coincide, the assembly relationship is a shaft-hole mating relationship; when the planes of two different parts are in contact with each other, the assembly relationship is a surface contact relationship; when the axes of two different parts coincide, the assembly relationship is a coaxial relationship.
[0032] The determined assembly relationship is identified as a candidate assembly relationship. For each candidate assembly relationship, a candidate assembly relationship identifier is set, and the two parts participating in the candidate assembly relationship, the two geometric features belonging to the two parts respectively, the relationship type of the candidate assembly relationship, and the position and orientation of the two geometric features in the nominal assembly state are recorded.
[0033] Candidate assembly relationships thus preserve the geometric correspondences that already exist in the 3D assembly model and the specific geometric features that form the geometric correspondences. When the PMI changes and the position, orientation or size of the corresponding geometric features changes, the geometric state of the two geometric features after the change can be determined according to the same candidate assembly relationship, and subsequent assembly constraints can be established accordingly.
[0034] S2. Generate PMI variation parameters based on PMI structured data, and generate assembly constraints and their current status based on candidate assembly relationships.
[0035] For each PMI, the allowable changes in size, position, orientation, or surface position of the geometric features labeled by the PMI relative to the nominal state are determined based on the structured data of the PMI, and one or more numerical values that can represent the magnitude of the corresponding changes are used as the PMI variation parameters of the PMI.
[0036] When a PMI can be represented by a single value, the PMI corresponds to a PMI variation parameter. When a PMI requires multiple values to represent a change, multiple PMI variation parameters are recorded in a fixed order, along with the change content and allowable range for each parameter. The change content includes the geometric feature affected by the parameter, as well as the updated current position, orientation, dimension, surface position, or relative position between two geometric features. Simultaneously, the allowable conditions satisfied by multiple PMI variation parameters are recorded according to the tolerance zone corresponding to the PMI, ensuring that the overall combination of these parameters meets the tolerance requirements specified by the corresponding PMI. For PMI variation parameters representing a nominal state, a value of zero indicates that the corresponding geometric state remains in the nominal state.
[0037] PMI structured data records the tolerance information specified by PMI itself. PMI variation parameters further represent the amount of change in the geometric feature or the geometric relationship between geometric features relative to the nominal state when the corresponding tolerance requirements are met. By forming PMI variation parameters, the tolerance specifications in PMI can be converted into input quantities that can directly update the current position, current orientation, current size of geometric features, or the current relative position between geometric features, thereby establishing the correspondence between PMI tolerance requirements and the current geometric state of the part.
[0038] When the PMI specifies the dimensional tolerance of a single geometric feature, the dimensional deviation of the current dimension of the geometric feature indicated by the PMI relative to the nominal dimension is determined as the PMI variation parameter, and the allowable range of the PMI variation parameter is determined based on the upper and lower deviations or the upper and lower limit dimensions. When the dimensional tolerance specifies the distance between two geometric features, the dimensional deviation of the current distance between the two geometric features relative to the nominal distance is determined as the PMI variation parameter, and the PMI variation parameter is used as the relative positional change between the two geometric features along the measurement direction of the distance in the nominal state.
[0039] When the PMI is a positional tolerance, based on the tolerance zone of the positional tolerance and the nominal position of the geometric feature indicated by the PMI, a positional deviation is determined to represent the positional change of the geometric feature in the permissible direction, and the corresponding positional deviation is determined as a PMI variation parameter. For example, when the positional tolerance of an axis allows the axis to change position in a plane perpendicular to the nominal axis, the positional deviations of the axis along two mutually perpendicular directions in the plane can be used as two PMI variation parameters, and the two PMI variation parameters can jointly satisfy the permissible conditions determined by the tolerance zone corresponding to the positional tolerance.
[0040] When the PMI is a directional tolerance, the directional deviation of the geometric feature marked by the PMI relative to the nominal direction is determined based on the geometric tolerance item, the nominal direction of the geometric feature marked by the PMI, and the geometric feature corresponding to the datum mark when there is a datum requirement. The corresponding directional deviation is then determined as the PMI variation parameter. For example, when a plane has a parallelism tolerance relative to the plane corresponding to the datum mark, the rotation angles of the plane relative to the nominal direction around two mutually perpendicular axes can be used to represent the two directional deviations, and the two directional deviations can be made to jointly satisfy the allowable conditions determined by the tolerance zone corresponding to the parallelism tolerance.
[0041] When the PMI is a form tolerance, runout tolerance, or profile tolerance, multiple positions are selected on the geometric feature marked by the PMI to determine the corresponding geometric changes, and the position and orientation of each position in the nominal state are recorded. These multiple positions can be determined based on the discrete representation of the geometric feature in the three-dimensional assembly model. For form tolerances or profile tolerances, the change in the normal position of each position relative to the nominal surface represents the corresponding surface position change. For runout tolerances, the change in the radial or axial position of each position relative to the rotational datum determined by the datum marker represents the corresponding surface position change, and this surface position change is determined as the corresponding PMI variation parameter. When multiple positions undergo different position changes, the position and orientation used for subsequent contact judgment are re-determined based on the updated geometric relationship between adjacent positions.
[0042] Given the current values of each PMI variation parameter, update the current size of the corresponding geometric feature or the current distance and relative position between two geometric features marked by the PMI according to the PMI variation parameter corresponding to the dimensional tolerance; update the current position of the corresponding geometric feature according to the PMI variation parameter corresponding to the position tolerance; update the current orientation of the corresponding geometric feature according to the PMI variation parameter corresponding to the orientation tolerance; and update the corresponding position according to the PMI variation parameter corresponding to the form tolerance, runout tolerance, or profile tolerance. Combine the nominal pose of each part and the relative position change between the two geometric features to determine the position and orientation of each part and each geometric feature in this calculation, and determine the resulting assembly state as the current assembly state.
[0043] For each candidate assembly relationship, one or more assembly constraints are established to limit the relative movement of the two corresponding parts, based on the relationship type of the candidate assembly relationship and the two geometric features involved in the candidate assembly relationship, and an assembly constraint identifier is set for each assembly constraint.
[0044] Assembly constraints that are continuously used to limit the relative motion of two parts under the current analysis condition are defined as persistent constraints, and assembly constraints that restrict the relative motion in the corresponding direction after two geometric features reach the contact boundary are defined as state-dependent constraints. For state-dependent constraints, the corresponding assembly constraint identifier, constraint condition, contact boundary, and processing relationship for determining its current state based on the clearance and relative motion direction are recorded.
[0045] The relative position required to reach the contact boundary refers to the relative position change required for the two corresponding geometric features to move from the current assembly state to the contact boundary along the constraint direction of the state-dependent constraints. When determining the initial state based on the current PMI variation parameters, the relative motion direction is determined according to the relative position change of the two corresponding geometric features when updating from the nominal state to the current assembly state. In the subsequent assembly state solution stage, when releasing state-dependent constraints currently in effect, the pending state-dependent constraints are not applied. The part pose change is calculated based on the changes in the PMI variation parameters that have not yet been applied and the remaining assembly constraints currently in effect, and the relative motion direction is determined based on the calculation results. The former is used for initial geometric state determination, and the latter is used for release determination after state-dependent constraints have participated in the assembly state solution.
[0046] For persistent constraints, their current state is determined as the active state. For state-dependent constraints, the gap along the limiting direction is determined based on the position, orientation, and size of the two corresponding geometric features in the current assembly state, as well as the current distance when distance is involved. When a gap exists between two geometric features, the current state of the corresponding state-dependent constraint is determined as the released state. When two geometric features reach the contact boundary and have a tendency to separate along the direction that increases the gap, the current state of the corresponding state-dependent constraint is determined as the released state. When two geometric features reach the contact boundary and do not have a tendency to separate along the direction that increases the gap, the current state of the corresponding state-dependent constraint is determined as the active state. When the state-dependent constraint is in the released state, if the current relative movement continues and crosses the contact boundary, the contact boundary is used as the state switching position, and the corresponding state-dependent constraint is updated to the active state. When the state-dependent constraint is in the active state, if two geometric features have a tendency to separate along the direction that increases the gap, the corresponding state-dependent constraint is updated to the released state. When two geometric features still have a tendency to continue to approach each other, the corresponding state-dependent constraint remains in the active state.
[0047] The assembly constraint identifier and constraint conditions of state-dependent constraints remain unchanged between the active and released states; what changes is the current state of the state-dependent constraint. This allows the same assembly constraint to be used to determine whether a geometric feature has entered an active state even after a PMI change alters the gap between geometric features, while maintaining the PMI interaction relationship and constraint data corresponding to the same technical object before and after the state switch.
[0048] The current state of the state-dependent constraints determined based on the current PMI variation parameters and the nominal pose of each part is the initial state; after the final assembly constraint set is formed, the current state of the state-dependent constraints is re-determined based on the relative motion direction before solving the assembly state.
[0049] S3. Determine the PMI constraint activation relationship based on the impact of PMI variation parameters on assembly constraints.
[0050] For each assembly constraint, based on the PMI variation parameters related to the current position, current orientation, current dimension, or surface position that form the geometric relationship corresponding to the assembly constraint, the PMI to be evaluated is determined, and the handling method for the impact of the PMI variation parameters on the assembly constraint is determined according to the current state of the assembly constraint. The PMIs involved in this analysis are indexed in a fixed order, where i is the PMI index and its value corresponds to each PMI involved in this analysis; the assembly constraints processed in this analysis are indexed in a fixed order, where j is the assembly constraint index and its value corresponds to each assembly constraint processed in this analysis.
[0051] For an assembly constraint currently in effect, the translational changes along the three coordinate axes and the rotational changes around the three coordinate axes of the two parts involved in the assembly constraint are respectively composed of the part pose change vectors of the two parts in a fixed order; for the first... Each assembly constraint is used to form a constraint Jacobian matrix based on the first-order relationship between each independent constraint condition and the part pose change vectors of the two parts. and According to the independent constraint condition relative to the first The first-order relationship between the changes in the PMI parameters of each PMI forms the PMI constraint influence matrix. Establish a formula to represent the relationship between changes in PMI parameters, pose changes of two parts, and currently active assembly constraints:
[0052] in, and They respectively represent participation in the first The part pose change vector of two parts under assembly constraints. and They represent the first The assembly constraints are formed by the constraint Jacobian matrix relative to the part pose change vectors of the two parts. Indicates the first The change in the PMI parameter of each PMI. Indicates the first The PMI change parameter of the first PMI is related to the second PMI. The PMI constraint influence matrix generated by each assembly constraint is calculated as follows: This represents matrix multiplication; in the formula... This represents the zero vector corresponding to each independent constraint. The rows of the constraint Jacobian matrix and the PMI constraint influence matrix are formed according to the physical dimensions of the corresponding independent constraints, ensuring that each calculation term corresponding to the same independent constraint has a consistent physical dimension.
[0053] The above calculation formula is used for assembly constraints that are currently in effect. It indicates that after the PMI parameters change, causing the corresponding geometric features to change, the two parts involved in the assembly constraint jointly maintain the independent constraint conditions in the assembly constraint through their own pose changes. When multiple PMIs affect the same assembly constraint, they respectively form corresponding PMI constraint influence matrices, and jointly participate in the assembly state solution according to the order of PMI parameter changes when forming the currently effective global PMI constraint influence matrix.
[0054] For state-dependent constraints currently in a released state, determine the corresponding gaps and contact boundaries based on the position, orientation, and dimensions of the two corresponding geometric features in the current assembly state, and calculate the first... When each PMI parameter undergoes a small unit change, the first-order change in the gap or the relative position required to reach the corresponding contact boundary is calculated to form the... The PMI relative to the first PMI constraint influence matrix for each state-dependent constraint .
[0055] When currently in the release state, the PMI constraint influence matrix is used to indicate whether the PMI variation parameter can change the gap of the state-dependent constraint or the relative position required to reach the contact boundary, so that the PMI tolerance change and the state-dependent constraint may subsequently switch states. After the state-dependent constraint enters the effective state, when it continues to cross the contact boundary according to the current relative motion, the geometric state when the two corresponding geometric features reach the contact boundary is taken as the current geometric state of entering the effective state. Then, according to the current position, current direction and current size of the two geometric features when entering the effective state, the corresponding constraint Jacobian matrix and PMI constraint influence matrix are re-determined, and the assembly constraints currently in the effective state participate in the subsequent assembly state solution.
[0056] The PMI constraint influence matrix corresponds to the first under different current states. The PMI and the first The interaction relationship between assembly constraints; when the first assembly constraint... When an assembly constraint is currently in effect, the PMI constraint influence matrix is determined based on the first-order influence of the PMI variation parameters on the independent constraints within the assembly constraint. When an assembly constraint is a state-dependent constraint currently in a released state, the PMI constraint influence matrix is determined based on the first-order influence of the PMI variation parameters on the corresponding gap or the relative position required to reach the contact boundary; after the current state of the state-dependent constraint changes, the corresponding PMI constraint influence matrix is re-formed based on the changed geometric state.
[0057] For assembly constraints that are currently in effect or state dependency constraints that are currently in a released state, the rank of the corresponding PMI constraint influence matrix is used to determine the first... Can the PMI change the first At least one independent constraint in the assembly constraints or a change in the first assembly constraint State determination conditions for each state-dependent constraint are established, and:
[0058] in, Denotes the rank of a matrix; when the rank is... When an assembly constraint is currently in effect and the above conditions are met, it indicates that the first assembly constraint is in effect. At least one PMI variation parameter of a PMI can change at least one independent constraint condition in the assembly constraints; when the first PMI When the assembly constraint is a state dependency constraint currently in a released state and the above conditions are met, it indicates that the first assembly constraint is a state dependency constraint. At least one PMI variation parameter of a PMI can change the gap corresponding to the state-dependent constraint or the relative position required to reach the contact boundary. The PMI and the first The correspondence between assembly constraints is determined as the PMI constraint activation relationship.
[0059] The PMI constraint activation relationship maps the geometric changes specified by the PMI to the assembly constraints affected by those geometric changes. For assembly constraints currently in an active state, the mapping relationship is used to determine which current assembly constraints the PMI change participates in the part pose solution through; for state-dependent constraints currently in a released state, the mapping relationship is used to preserve the action path between the PMI change and subsequent contact state changes, and to continue participating in the current assembly state solution after the state-dependent constraint becomes active.
[0060] For each PMI constraint activation relationship, record the PMI identifier, the affected PMI change parameter, the corresponding assembly constraint identifier, and the candidate assembly relationship identifier corresponding to the assembly constraint; when multiple PMIs affect the same assembly constraint, establish corresponding PMI constraint activation relationships respectively.
[0061] S4. Select assembly constraints based on the target allowable motion state to form the final assembly constraint set and the final effective tolerance variation transmission relationship.
[0062] Based on the product's assembly function, the relative motions of the parts that need to be retained in the three-dimensional tolerance analysis model are determined, and the relative motions of the parts that need to be retained are defined as the target allowable motion states. The target allowable motion states can be determined based on the assembly relationships in the three-dimensional assembly model, the product assembly definition, or pre-given motion requirements. For example, the target allowable motion state of a revolute joint may include rotation about a specified axis, and the target allowable motion state of a prismatic joint may include translation along a specified direction. Parts that need to be fixed do not retain independent relative motions.
[0063] The target permissible motion state is used to clarify the functional motions of the product that should be retained in the final assembly constraints. During subsequent screening of assembly constraints, other motions that need to be restricted in the current analysis condition are limited while retaining the target permissible motion state. This ensures that the resulting three-dimensional tolerance analysis model simultaneously reflects the permissible functional motions of the product and the motion constraints arising from actual assembly relationships. Parts or assembly components requiring assembly degree-of-freedom closure are indexed in a fixed order. This is an index for a part or assembly component, and its value corresponds to each part or assembly component for which assembly degree of freedom closure needs to be performed this time.
[0064] For the For a part or an assembly assembly consisting of multiple parts that needs to perform assembly degree-of-freedom closure, determine the number of independent motion degrees of freedom participating in the joint solution, and represent each independent motion in the target allowed motion state as a motion vector in the same order as the independent motion degrees of freedom. Then, use all the independent motion vectors that need to be retained as columns to form the target allowed motion basis matrix. .
[0065] For each assembly constraint that is currently active and corresponds to the PMI constraint activation relationship, the assembly constraint is adjusted according to the assembly constraint. The first-order constraints on the independent degrees of freedom of motion in each part or assembly form a constraint matrix. And establish a calculation formula for determining whether the assembly constraints restrict the target's permissible motion state:
[0066] in, Indicates the first The constraint matrix corresponding to each assembly constraint. Indicates the first The target permissible motion basis matrix corresponding to a part or assembly component, in the calculation formula This represents matrix multiplication; when the above formula holds true, it indicates that the first... The assembly constraint does not restrict any independent motion in the target's allowed motion state, and the assembly constraint continues to be used to determine subsequent independent constraint conditions. For state-dependent constraints that are currently in effect, the product being zero is not used as the sole criterion; instead, each independent motion in the target's allowed motion state is determined one by one based on the current restriction direction. When there is an independent motion along the direction that reduces the corresponding gap, it is determined that the state-dependent constraint restricts the target's allowed motion state; otherwise, it is determined that the state-dependent constraint does not restrict the target's allowed motion state.
[0067] For state-dependent constraints that are currently in a released state and have established PMI constraint activation relationships, retain the assembly constraint identifier, constraint conditions, PMI constraint activation relationships, and current state of the state-dependent constraints; after the state-dependent constraints enter the effective state, determine whether the assembly constraints meet the target allowable motion state and whether they participate in the current assembly state solution based on the geometric state when entering the effective state.
[0068] The assembly constraints that already meet the target allowed motion state and are currently in effect are arranged into a selection constraint matrix according to a uniform order of independent degrees of freedom. For assembly constraints that are currently in effect and are to be added, the assembly constraints are added to the selection constraint matrix to form the selection constraint matrix to be judged. and establish:
[0069] in, This represents the selection constraint matrix before adding the current assembly constraint to be judged. This represents the selection constraint matrix after adding the current assembly constraint to be judged; when the above conditions are met, it means that the assembly constraint to be added provides an independent motion restriction that does not exist in the previous selection constraint set, and the assembly constraint is added to the selection constraint set.
[0070] When the above conditions are not met, based on the linear relationship between the constraint matrix of the assembly constraint to be added and the existing selection constraint matrix, the relationship between the PMI constraint influence matrix corresponding to the assembly constraint and the PMI constraint influence matrix corresponding to the existing assembly constraints in the selection constraint set is checked. When the constraint matrix can be represented according to the linear relationship of the existing assembly constraints in the selection constraint set, the PMI constraint influence matrices participating in the comparison are arranged in the corresponding columns according to the unified PMI variation parameter order, and the same set of linear combination coefficients representing the linear relationship of the constraint matrix are used to determine whether the corresponding PMI constraint influence matrices maintain the same linear relationship. When the corresponding PMI constraint influence matrices maintain the same linear relationship, the existing assembly constraints in the selection constraint set have already represented the same motion restriction and PMI action relationship. When the corresponding PMI constraint influence matrices do not maintain the same linear relationship, the assembly constraint is added to the selection constraint set and the corresponding PMI constraint activation relationship is retained, and the assembly constraint participates in the subsequent common solvability judgment, but it is not counted as the number of newly added independent motion restrictions.
[0071] This allows for the separate handling of motion constraint effects and PMI tolerance effects within assembly constraints. Assembly constraints with increased constraint matrix rank are used to supplement new independent motion constraints. Assembly constraints without increased matrix rank but with different PMI effects retain their corresponding tolerance action paths, ensuring that assembly relationships with the same motion constraints but affected by different PMIs can undergo consistency checks in subsequent assembly state solutions.
[0072] After the assembly constraints currently in effect corresponding to the PMI constraint activation relationship have completed the above screening, if the selection constraint set still needs to add independent motion restrictions, continue to select assembly constraints that satisfy the target allowed motion state and can increase the rank of the selection constraint matrix from the remaining assembly constraints currently in effect corresponding to the already formed candidate assembly relationships that have not been added to the selection constraint set. The assembly constraints added to supplement the independent motion restrictions for solving the current assembly state are determined as closed supplementary constraints, and the closed supplementary constraints are added to the selection constraint matrix.
[0073] The closed supplementary constraint supplements the independent motion constraints required for solving the current assembly state. The assembly constraint corresponding to the PMI constraint activation relationship maintains the action path of PMI tolerance change. Together, they constitute an assembly constraint system that can participate in three-dimensional tolerance analysis.
[0074] The first The number of independent motions that need to be retained in the target allowed motion state of a part or assembly is denoted as . , will the The number of independent degrees of freedom of motion of a part or assembly participating in a joint solution is denoted as . Given that, in the current assembly state, all independent motions except those required to be retained in the target allowed motion state are subject to assembly constraints, a calculation formula is established to determine whether the number of independent motion constraints in the current assembly state solution meets the requirements:
[0075] in, Indicates the first The number of independent motion degrees of freedom of each part or assembly component participating in the joint solution. This indicates the number of independent motions that need to be retained in the target's allowed motion state; when the above condition is met, it means that the number of independent motion restrictions provided by the selected constraint set corresponds to the number of independent motions that need to be restricted in the current assembly state solution.
[0076] When a state-dependent constraint in the released state does not restrict the corresponding independent motion, and the independent motion does not need to be restricted by other assembly constraints in the current assembly state, the independent motion is not counted as an independent motion that needs to be restricted by the selection constraint set. The determination is made based on whether the rank of the selection constraint matrix corresponds to the number of independent motions that need to be restricted.
[0077] For state-dependent constraints in the selection constraint set, when the state-dependent constraint only restricts one direction of the corresponding independent motion and remains released in the opposite direction, and when the state-dependent constraint increases the rank of the selection constraint matrix, it determines that it provides independent motion constraints in the current contact state and the currently restricted relative motion direction; the independent motion constraints are used as constraints for solving the current assembly state, and after obtaining the new part pose, the current state of the state-dependent constraint is re-determined based on the gap between the corresponding geometric features.
[0078] When determining whether the current assembly state meets the constraints required for solving the current assembly state, it is also determined whether the rank of the constraint matrix and the direction of motion to be restricted in the current contact state are restricted by the corresponding assembly constraints. When the new part pose causes the geometric feature corresponding to a certain state-dependent constraint to separate along the direction of increasing gap, the state-dependent constraint is updated to the release state, and the solution relationship is re-formed based on the updated assembly constraint current state.
[0079] Perform a nominal assembly state consistency check on the selected constraint set and closed supplementary constraints that satisfy the current assembly state solution conditions. Set each PMI variation parameter to its corresponding nominal value and set each part to its own nominal pose. Check whether the geometric features involved in each candidate assembly relationship satisfy the assembly relationship recorded by the 3D assembly model in the nominal assembly state according to the selected constraint set and closed supplementary constraints. Also check whether the selected constraint set and closed supplementary constraints continue to retain the target's allowed motion state.
[0080] The nominal assembly state consistency check is used to re-map the constraint combination formed by motion screening and independent constraint screening to the nominal geometric state in the three-dimensional assembly model, so that the formed assembly constraint system can maintain the original assembly relationship of the product under the nominal tolerance state.
[0081] Further check whether the assembly constraints corresponding to each PMI constraint activation relationship are retained in the selected constraint set or as state-dependent constraints that need to be retained; retain the PMI constraint activation relationships corresponding to the assembly constraints so that the PMI action path entering the final three-dimensional tolerance analysis model corresponds to the final retained assembly constraints.
[0082] The set of selection constraints, closed supplementary constraints, and state-dependent constraints that meet the retention conditions, processed as described above, are collectively determined as the final assembly constraint set. For each assembly constraint in the final assembly constraint set, its assembly constraint identifier, current state, and corresponding PMI constraint activation relationship are recorded. The assembly constraints that are currently in effect and correspond to the PMI constraint activation relationship are determined as the final effective constraints. The correspondence between the corresponding PMI identifier, PMI variation parameters, and the assembly constraint identifier of the final effective constraints is determined as the final effective tolerance variation propagation relationship.
[0083] For state-dependent constraints currently in a released state and retained in the final assembly constraint set, the corresponding PMI constraint activation relationships are retained. When a state-dependent constraint subsequently enters an effective state, it is identified as a new final effective constraint based on the same assembly constraint identifier, and the corresponding PMI constraint activation relationship is added to the currently used final effective tolerance variation transitivity. When a state-dependent constraint is updated from an effective state to a released state, the final effective tolerance variation transitivity relationship corresponding to the state-dependent constraint is discontinued, but the corresponding PMI constraint activation relationship is retained.
[0084] S5. Generate and solve a three-dimensional tolerance analysis model based on the PMI variation parameters, the final assembly constraint set, and the final effective tolerance variation transfer relationship.
[0085] For each PMI, a PMI change source is established. The PMI change source records the PMI identifier, the geometric features marked by the PMI, the corresponding PMI change parameters, the change content represented by each PMI change parameter, and the allowable range. When multiple PMI change parameters jointly represent the same PMI, the allowable conditions jointly satisfied by the multiple PMI change parameters are also recorded.
[0086] For each assembly constraint in the final assembly constraint set, the assembly constraint identifier, the corresponding candidate assembly relationship identifier, the two parts involved in the assembly constraint, the two geometric features, the corresponding constraint conditions, and the current state are recorded. For state-dependent constraints, the corresponding contact boundary and the processing relationship for determining the current state based on the gap and relative motion direction are also recorded.
[0087] For each final effective tolerance variation propagation relationship, record the corresponding PMI identifier, the affected PMI variation parameters, and the assembly constraint identifier of the corresponding final effective constraint; for state-dependent constraints currently in the released state, continue to retain the corresponding PMI constraint activation relationship according to the assembly constraint identifier, and do not include the corresponding PMI constraint activation relationship in the currently used final effective tolerance variation propagation relationship, so that the PMI variation parameters participating in the solution of the current assembly state can be determined after the state-dependent constraint enters the effective state.
[0088] The PMI variation source determines which PMI the tolerance change originates from and the corresponding allowable range of change. The final assembly constraint set determines the assembly geometry that needs to be maintained in the current model and the current state of each state-dependent constraint. Finally, the effective tolerance variation propagation relationship determines which currently effective assembly constraints affect the assembly state of related parts through which PMI variation parameters affect the assembly state of related parts. All of the above data together form a continuous action chain from PMI tolerance input to changes in the assembly state of parts.
[0089] In a specific assembly state solution, the pose components corresponding to the allowable motion of the target are taken as values pre-given for this analysis condition. This pre-given value does not change the allowable properties of the target's allowable motion in the 3D tolerance analysis model. For the joint solution of multiple parts, the position and orientation of a selected part are kept unchanged, or the relative pose changes of other parts with respect to the selected part are used for the solution, ensuring that the translations and rotations occurring together in the entire assembly do not enter the pose change quantity to be determined. After completing the above-mentioned pose component assignment and excluding overall rigid body motion, for pose components not yet determined by the currently effective assembly constraints, they are maintained at the values corresponding to the current assembly state.
[0090] When solving the assembly state using a first-order linearization method, the unknown pose changes, excluding the pose components held at a given position, are arranged in a fixed order to form the part pose change vector. The assembly constraints that are in effect in the final assembly constraint set under the current assembly state are arranged into a currently effective global constraint matrix according to the order of the part pose change vectors. The PMI constraint influence matrix generated by each PMI change parameter on the currently effective assembly constraints is formed by arranging the PMI constraint influence matrix in the same constraint order and PMI change parameter order. The changes in PMI variation parameters participating in the joint solution will be used to form... A formula is established to represent how, after the PMI parameters change and the geometric features change, each part, through its own pose change, jointly satisfies the assembly constraints that are currently in effect:
[0091] Among them, superscript Indicates the current effective status. This represents the unknown pose changes of each part involved in the joint solution, excluding the pose components that remain at a given position. This represents the change in the PMI variation parameter involved in the joint solution. This represents the currently active global constraint matrix. This represents the current effective global PMI constraint impact matrix, calculated in the formula. This represents matrix multiplication.
[0092] In each assembly state solution stage and The current assembly state at the time when the currently effective global constraint matrix and the currently effective global PMI constraint influence matrix of this stage are formed is used as the reference state, where, This represents the change in PMI parameters that has not yet taken effect relative to the current assembly state. The change in PMI parameters is the amount of change required for the current value of the PMI parameter in this stage to reach the given value of the corresponding PMI parameter in this analysis.
[0093] like Figure 2 As shown, when the current assembly state solution includes multiple PMI parameter changes, the current value of each PMI parameter corresponding to the current assembly state is used as the starting value for this stage. Each PMI parameter changes synchronously from its current value to the corresponding given value in this analysis at the same proportion. Along this synchronous change process, it is sequentially determined whether each state dependency constraint meets the state switching condition. When a state dependency constraint that meets the state switching condition first appears, the changes in each PMI parameter before reaching the state switching position are considered the already implemented changes, and the changes that have not yet occurred are considered the unimplemented changes. The current assembly state formed at the state switching position is used as the reference state for subsequent assembly state solutions, and the unimplemented changes continue to change synchronously at the same proportion. When multiple state dependency constraints meet the state switching condition under the same synchronous change proportion, the current state of the corresponding state dependency constraints is updated simultaneously, and the subsequent solution relationship is re-formed according to the updated current state until all the given PMI parameter changes have been implemented.
[0094] For state dependency constraints that are currently in effect, they are only used to restrict the corresponding geometric features from continuing to move closer to each other along the current restriction direction in the current assembly state solution; when the corresponding geometric features have a tendency to separate along the direction of increasing gap, the state dependency constraints are updated to the release state and the current solution relationship is re-formed.
[0095] The current effective global PMI constraint influence matrix is formed based on the assembly constraints currently in effect and their PMI constraint activation relationships. Therefore, a PMI change parameter participates in the part pose solution through the PMI action path in effect under the current assembly state. State-dependent constraints in the released state remain in the final assembly constraint set, and their corresponding assembly constraints participate in the formation of the current effective global constraint matrix and the current effective global PMI constraint influence matrix after entering the effective state.
[0096] For assembly constraints whose rank of the constraint matrix has not increased but are retained due to different PMI constraint influence relationships, after giving the change amount of the current PMI variation parameters, the assembly constraint is solved together with other currently effective assembly constraints for assembly state determination; when there is a part pose change vector that can simultaneously satisfy all currently effective assembly constraints, the part pose change vector is determined based on the result of jointly satisfying all currently effective assembly constraints; when there is no part pose change vector that can simultaneously satisfy all currently effective assembly constraints, the current PMI variation parameter combination is determined as a state that cannot simultaneously satisfy the assembly relationship under the current constraint state, and the corresponding PMI variation parameter combination and the assembly constraints that cannot be jointly satisfied are recorded.
[0097] For state-dependent constraints that are currently in effect, before performing the final solution for the current assembly state, the constraint rows corresponding to the state-dependent constraints are not included in the currently effective global constraint matrix. The changes in the PMI parameters and the other currently effective assembly constraints are used to attempt to determine the part pose changes, and the relative motion direction of the corresponding geometric features is determined based on the trial results. When the relative motion direction is along the direction that increases the corresponding gap, the state-dependent constraint is updated to the released state. When the relative motion direction causes the corresponding geometric features to continue to approach each other, the state-dependent constraint remains in the effective state.
[0098] For state-dependent constraints currently in the release state, when the corresponding geometric feature reaches the contact boundary and does not have a tendency to separate along the direction that increases the gap, or when it will cross the contact boundary if it continues to move according to the current relative motion, the contact boundary is used as the state switching position, and the geometric state when the corresponding geometric feature reaches the contact boundary is used as the current geometric state for entering the effective state.
[0099] When a state transition occurs before all PMI parameter changes that have not yet taken effect in this stage have been fully implemented, the PMI parameter changes that have reached the state transition position for the corresponding geometric features will be applied to the current assembly state according to the above-described synchronous change process for PMI parameters. The assembly state at the state transition position will then be used as the current assembly state for subsequent assembly state solutions. The PMI parameter changes that have already been applied to the current assembly state will no longer participate in subsequent solutions, while the PMI parameter changes that have not yet taken effect will continue to be used as the PMI parameter changes in subsequent assembly state solutions.
[0100] Based on the updated assembly constraint state, the currently effective global constraint matrix and the currently effective global PMI constraint influence matrix are reconstructed. The assembly constraints currently in effect are reassessed to determine whether they meet the target allowable motion state and whether the number of independent motion restrictions meets the requirements. If additional independent motion restrictions are needed, the closed supplementary constraints and selection constraint matrix are updated. If, after completing the state update, there are still no part pose changes that can jointly satisfy all currently effective assembly constraints, the combination of PMI change parameters is determined to be a state that cannot simultaneously satisfy the current assembly relationship.
[0101] The common solvability judgment enables assembly constraints with the same motion restrictions but different PMI action relationships to participate in the current assembly state verification. Thus, when changes in PMI cause inconsistencies between different assembly relationships, the inconsistencies are represented as assembly constraint conflicts under the current PMI change state, rather than the pose results that violate some assembly constraints as normal three-dimensional tolerance analysis results.
[0102] When there exists a part pose change vector that can simultaneously satisfy all currently active assembly constraints, the part pose change vector is obtained based on the change amount of PMI variation parameters given by each PMI variation source, and the position and orientation of each part, as well as the current position, orientation, and dimension of the corresponding geometric features, are updated based on the obtained results.
[0103] When the final assembly constraint set contains state-dependent constraints, the gap and relative motion direction between the corresponding geometric features of the state-dependent constraints currently in the released state are re-determined based on the updated current position, current direction, and current size of the geometric features, and it is determined whether the state-dependent constraints currently in the released state meet the conditions for entering the effective state; whether the state-dependent constraints currently in the effective state enter the released state is determined before the final solution is executed in the next assembly state solution stage, in a manner that the state-dependent constraints to be judged are not applied.
[0104] For state-dependent constraints that transition from a released state to an active state, when the corresponding geometric feature reaches the contact boundary and does not have a tendency to separate along the direction that increases the gap, or when the current relative motion continues and will cross the contact boundary, the contact boundary is used as the state switching position. Based on the current position, current direction, and current size of the geometric feature after entering the active state, the first-order relationship between each independent constraint condition in the assembly constraint and the part pose change, as well as the first-order influence of the PMI variation parameter on the independent constraint condition, are re-determined, and the corresponding constraint row and PMI constraint influence term are re-formed. For state-dependent constraints that transition from an active state to a released state, when re-forming the currently active global constraint matrix and the currently active global PMI constraint influence matrix, the corresponding constraint row and PMI constraint influence term are stopped, and the final effective tolerance variation propagation relationship corresponding to the state-dependent constraint is stopped.
[0105] Based on the updated state of all assembly constraints, a new global constraint matrix and a global PMI constraint influence matrix are formed. The current assembly state at the time of forming these matrices are used as the new starting state for the solution. The part pose change relative to the current assembly state is calculated based on the changes in PMI parameters that have not yet been applied to the current assembly state. The calculated part pose change is then applied to the current assembly state. The PMI parameter changes that have already been applied to previous assembly states are no longer included in the subsequent assembly state solution. This process is repeated until all the given PMI parameter changes have been applied, and the current assembly state formed based on the calculated part pose corresponds to the current state of each state-dependent constraint.
[0106] Through the above state update, the gap changes, contact formation and contact release caused by PMI changes can be further altered to change the assembly constraints currently involved in the solution, so that the three-dimensional tolerance analysis model corresponding to different combinations of PMI change parameters adopts the constraint relationship corresponding to the corresponding geometric state.
[0107] Based on the final obtained positions and orientations of each part, as well as the current positions, orientations, and dimensions of each geometric feature, the dimensions, positions, orientations, or clearance results of interest in the three-dimensional tolerance analysis are determined.
[0108] When performing extreme value analysis, appropriate values are selected within the allowable range of each PMI variation parameter for solution; when multiple PMI variation parameters jointly represent the same PMI, the selected multiple PMI variation parameters are made to jointly satisfy the allowable conditions determined by the tolerance zone corresponding to the PMI.
[0109] During statistical analysis, each PMI variation parameter is obtained based on the pre-defined probability distribution, distribution parameters, and statistical sample size. Assembly state solutions are then performed for each set of obtained PMI variation parameters. The probability distribution and distribution parameters are determined based on manufacturing process statistical data of the geometric changes represented by the corresponding PMI, and the statistical sample size is pre-determined according to the requirements of this statistical analysis. When multiple PMI variation parameters represent the same PMI, only combinations of PMI variation parameters that commonly satisfy the tolerance zone allowable conditions of the corresponding PMI are used to obtain the range or distribution result of the analysis result as the PMI changes.
[0110] In one embodiment, taking a three-dimensional assembly structure with two mutually perpendicular parallel relationships and corresponding surface contact relationships between the first part and the second part as an example, the implementation process of the above-mentioned automatic three-dimensional tolerance modeling method is explained: 1. Obtain PMI and assembly relationships and form basic data.
[0111] like Figure 3 As shown, the three-dimensional assembly model to be analyzed includes a first part and a second part. The first part has a first plane and a second plane that are perpendicular to each other. The second part has a third plane corresponding to the first plane and a fourth plane corresponding to the second plane. The third plane is parallel to the first plane, and the fourth plane is parallel to the second plane. The first plane and the third plane, as well as the second plane and the fourth plane, form corresponding parallel relationships and surface contact relationships, respectively. The second part can move relative to the first part along the intersection of the first plane and the second plane, and this movement is determined to be the product functional movement that needs to be retained in this implementation process.
[0112] The 3D assembly model records the parallel relationships between the first and third planes, the parallel relationships between the second and fourth planes, the surface contact relationships between the first and third planes, and the surface contact relationships between the second and fourth planes. The parallel relationship between the first and third planes is designated as the first candidate assembly relationship, the parallel relationship between the second and fourth planes as the second candidate assembly relationship, the surface contact relationship between the first and third planes as the third candidate assembly relationship, and the surface contact relationship between the second and fourth planes as the fourth candidate assembly relationship. Candidate assembly relationship identifiers are set for each relationship.
[0113] The first candidate assembly relationship records the position and orientation of the first part, the second part, the first plane, the third plane, and the two planes in the nominal assembly state. The second candidate assembly relationship records the position and orientation of the first part, the second part, the second plane, the fourth plane, and the two planes in the nominal assembly state. The third candidate assembly relationship records the position and orientation of the first part, the second part, the first plane, the third plane, and the two planes in the nominal assembly state. The fourth candidate assembly relationship records the position and orientation of the first part, the second part, the second plane, the fourth plane, and the two planes in the nominal assembly state.
[0114] The PMI in the three-dimensional assembly model includes a first dimensional tolerance PMI for the distance between the first plane and the third plane, and a second dimensional tolerance PMI for the distance between the second plane and the fourth plane. The first dimensional tolerance PMI includes the nominal distance between the first plane and the third plane, the upper and lower deviations corresponding to the nominal distance, and the measurement direction of the distance dimension in the nominal state. The second dimensional tolerance PMI includes the nominal distance between the second plane and the fourth plane, the upper and lower deviations corresponding to the nominal distance, and the measurement direction of the distance dimension in the nominal state.
[0115] 2. Generate PMI variation parameters and determine the current state of assembly constraints.
[0116] Structured PMI data are generated based on the first dimensional tolerance PMI and the second dimensional tolerance PMI, respectively. A first PMI variation parameter is generated based on the first dimensional tolerance PMI, indicating that the distance between the first plane and the third plane changes relative to the corresponding nominal distance. A second PMI variation parameter is generated based on the second dimensional tolerance PMI, indicating that the distance between the second plane and the fourth plane changes relative to the corresponding nominal distance. The allowable ranges of the first PMI variation parameter and the second PMI variation parameter are determined based on the upper and lower deviations corresponding to each dimensional tolerance.
[0117] The relative position and current distance between the first plane and the third plane are updated according to the measurement direction of the distance dimension corresponding to the first dimensional tolerance PMI in the nominal state based on the first PMI variation parameter. The relative position and current distance between the second plane and the fourth plane are updated according to the measurement direction of the distance dimension corresponding to the second dimensional tolerance PMI in the nominal state based on the second PMI variation parameter. The current assembly state is formed by combining the nominal pose of the first part and the second part with the updated relative position.
[0118] 3. Determine the activation relationship between PMI and assembly constraints.
[0119] Based on the first candidate assembly relationship, establish a first assembly constraint that keeps the first plane and the third plane parallel. Based on the second candidate assembly relationship, establish a second assembly constraint that keeps the second plane and the fourth plane parallel. Based on the third candidate assembly relationship, establish a third assembly constraint that restricts the first plane and the third plane from continuing to approach each other after they reach the contact boundary. Based on the fourth candidate assembly relationship, establish a fourth assembly constraint that restricts the second plane and the fourth plane from continuing to approach each other after they reach the contact boundary. Set assembly constraint identifiers for each constraint.
[0120] The first and second assembly constraints maintain the directional relationship between the two corresponding planes under the current analysis condition in this implementation process. The first and second assembly constraints are determined as persistent constraints and their current state is determined as the effective state. The third and fourth assembly constraints change their current state according to the gap between the corresponding planes. The third and fourth assembly constraints are determined as state-dependent constraints.
[0121] For the third assembly constraint, record the contact boundary between the first and third planes and the processing relationship for determining the current state based on their current gap and relative motion direction; for the fourth assembly constraint, record the contact boundary between the second and fourth planes and the processing relationship for determining the current state based on their current gap and relative motion direction.
[0122] In one analysis of this implementation process, the current value of the first PMI variation parameter causes the first plane and the third plane to reach the contact boundary, and the first plane and the third plane do not have a tendency to separate along the direction that increases the gap between them. The current value of the second PMI variation parameter causes the second plane and the fourth plane to maintain a gap greater than zero. Based on the current geometric state, the current state of the third assembly constraint is determined to be the active state, and the current state of the fourth assembly constraint is determined to be the released state.
[0123] The first and second assembly constraints maintain the orientation of the second part relative to the first part, so that when the second part moves along the intersection of the first and second planes, its third and fourth planes remain parallel to the first and second planes, respectively; the third assembly constraint restricts the second part from continuing to move toward the first plane in the current contact direction, and the fourth assembly constraint is currently in a released state, so that the second part still has the movement space provided by the current gap between the fourth and second planes.
[0124] For the third assembly constraint, a corresponding PMI constraint influence matrix is formed based on the first-order influence of the first PMI variation parameter on the contact condition between the first plane and the third plane. When the third assembly constraint is currently in effect, the PMI constraint influence matrix represents the first-order influence of the first PMI variation parameter on the corresponding independent constraint condition in the third assembly constraint. When the rank of the PMI constraint influence matrix is greater than zero, a PMI constraint activation relationship is established between the first dimensional tolerance PMI and the third assembly constraint.
[0125] For the fourth assembly constraint currently in a released state, the current gap between the second and fourth planes is determined based on the current distance between them and the corresponding contact boundary. The first-order change in the current gap or the relative position required to reach the contact boundary is calculated when the second PMI variation parameter changes. Based on this change, a PMI constraint influence matrix of the second dimensional tolerance PMI relative to the fourth assembly constraint is formed. When the rank of the PMI constraint influence matrix is greater than zero, a PMI constraint activation relationship is established between the second dimensional tolerance PMI and the fourth assembly constraint, and the PMI constraint activation relationship and the information that the fourth assembly constraint is currently in a released state are retained.
[0126] 4. Filter assembly constraints and form the final assembly constraint set.
[0127] Based on the product function, the translation of the second part along the intersection of the first and second planes is determined as the target permissible motion state. With the position and orientation of the first part unchanged, the second part has six independent degrees of freedom relative to the first part, and the target permissible motion state retains one of these translational motions.
[0128] A corresponding constraint matrix is formed based on the first and second assembly constraints, and it is determined whether the two assembly constraints restrict the target's allowed motion state. Since the first and second assembly constraints maintain the directional relationship between the third and fourth planes and allow the second part to translate along the intersection of the first and second planes, the two assembly constraints are used to determine independent constraint conditions.
[0129] The third assembly constraint, which is currently in effect, is added to the selection constraint set, and its unidirectional restriction on the corresponding movement direction of the second part is determined based on the current contact direction of the third assembly constraint. The fourth assembly constraint is currently in a released state, and its corresponding movement direction continues to retain the movement space provided by the clearance. Therefore, the fourth assembly constraint is not included in the selection constraint matrix.
[0130] In the current state described above, it is determined that both the first and second assembly constraints satisfy the target allowable motion state and can increase the rank of the selection constraint matrix. Therefore, the first and second assembly constraints are identified as closed supplementary constraints and added to the selection constraint matrix. The first, second, and third assembly constraints are used together to determine the number of independent motion restrictions in the current assembly state and the currently restricted motion direction. The direction corresponding to the fourth assembly constraint still has motion space provided by the gap. The assembly constraint identifier, constraint conditions, and PMI constraint activation relationship between the fourth assembly constraint and the second dimensional tolerance PMI are retained, so that the second PMI variation parameter continues to participate in the state judgment of the fourth assembly constraint.
[0131] like Figure 4 As shown, when the fourth assembly constraint is in the released state, there is a gap between the second and fourth planes. The fourth assembly constraint does not restrict the second part from moving relative to the second part within the gap range in the direction that reduces the gap. As the second dimensional tolerance PMI corresponding to the second PMI variation parameter changes, the current distance between the second and fourth planes decreases, and the gap decreases accordingly. When the second and fourth planes reach the contact boundary, and the second part still has a tendency to move the fourth plane closer to the second plane, the contact boundary is used as the state switching position, and the fourth assembly constraint is updated from the released state to the active state. After entering the active state, the fourth assembly constraint restricts the second part from moving in the direction that makes the fourth plane closer to the second plane. When the subsequent assembly state solution stage obtains a tendency for the fourth plane to separate from the second plane in the direction of increasing gap without applying the fourth assembly constraint, the fourth assembly constraint re-enters the released state.
[0132] If there are still unused parts of the second PMI variation parameter change after the fourth assembly constraint enters the effective state, the current assembly state when the second plane and the fourth plane reach the contact boundary is used as the starting state for solving the subsequent assembly state. The unused parts of the second PMI variation parameter change will continue to participate in the subsequent assembly state solution, and the parts of the change that have been used to form the contact boundary state will not participate in the subsequent solution again.
[0133] When multiple PMI parameter changes are given simultaneously in this analysis, each PMI parameter changes synchronously in the same proportion from its current value to the corresponding given value in this analysis, and the contact boundary reached first is determined along this synchronous change process. The changes that have occurred in each PMI parameter when the contact boundary is reached are considered as the already implemented changes, and the remaining changes are considered as the unimplemented changes in the subsequent assembly state solution. When multiple state-dependent constraints satisfy the state switching condition at the same change position, the current state of the corresponding state-dependent constraints is updated simultaneously.
[0134] The state change of the fourth assembly constraint causes the distance change generated by the second dimensional tolerance PMI to further form the motion constraint for solving the current assembly state through the gap change between the second plane and the fourth plane, realizing the continuous transmission between the second PMI variation parameters, the current geometric state and the current state of the assembly constraint.
[0135] After the fourth assembly constraint enters the active state, the first-order constraint relationship of the fourth assembly constraint relative to the pose change of the second part is re-determined according to the current position and direction of the second plane and the fourth plane, and the PMI constraint influence matrix is re-formed according to the first-order influence of the second PMI variation parameter on the contact constraint condition.
[0136] The fourth assembly constraint, after entering the effective state, is used to determine the target's permissible motion state. The fourth assembly constraint restricts the second part from continuing to approach the second plane along the current contact direction perpendicular to the second plane, while preserving the target's permissible motion of the second part along the common extension direction of the first and second planes. Therefore, the fourth assembly constraint continues to be used in the formation of the constraint matrix.
[0137] A new selection constraint matrix is formed based on the first assembly constraint, the second assembly constraint, and the currently active third and fourth assembly constraints. The first and second assembly constraints provide corresponding directional restrictions, while the third and fourth constraints, in their respective current contact states, restrict the second part from continuing to move along its corresponding contact direction. The third and fourth assembly constraints are used as the current constraints in this assembly state solution, and the selection constraint matrix is used to determine whether the number of independent motion constraints required for this assembly state solution has been met.
[0138] When the rank of the selection constraint matrix reaches the number of independent motion constraints required for solving the current assembly state, excluding the target's allowed motion, and all motion directions that need to be restricted in the current contact state are restricted by the corresponding assembly constraints, the current assembly state solution is executed according to the first assembly constraint, the second assembly constraint, the third assembly constraint, and the fourth assembly constraint. The third and fourth assembly constraints provide unidirectional constraints for the current contact direction in this solution, and continue to determine the current state based on the gap between the corresponding planes and the relative motion direction in subsequent assembly state solution stages.
[0139] The first assembly constraint, the second assembly constraint, the third assembly constraint, and the fourth assembly constraint are determined as the final set of assembly constraints. Based on the current state of each assembly constraint and the established PMI constraint activation relationship, the final effective tolerance variation transfer relationship is formed.
[0140] 5. Generate and solve the three-dimensional tolerance analysis model.
[0141] Based on the first dimensional tolerance PMI and the second dimensional tolerance PMI, PMI variation sources are established respectively. Based on the final assembly constraint set, the assembly constraint identifier, corresponding candidate assembly relationship, constraint condition and current state of the four assembly constraints are recorded, and the corresponding PMI constraint activation relationship is recorded, thereby forming the three-dimensional tolerance analysis model corresponding to the assembly structure.
[0142] During the assembly state solution process, the position and orientation of the first part are kept unchanged, and the target permissible motion of the second part along the intersection of the first and second planes is maintained at the position given in this analysis. The pose changes of the second part other than the target permissible motion are then calculated. For the pose components of the second part that have not yet been determined by the currently effective assembly constraints, the corresponding values at the beginning of this assembly state solution phase are retained.
[0143] The first, second, third, and fourth assembly constraints that are currently in effect are arranged in a unified order of part pose changes to form a currently effective global constraint matrix. The currently effective global PMI constraint influence matrix is then formed based on the PMI constraint activation relationship corresponding to the currently effective assembly constraints.
[0144] Given the change in PMI parameters, before performing the final solution for the current assembly state, for the third and fourth assembly constraints that are currently in effect, the constraint rows corresponding to the assembly constraints to be judged are temporarily not included in the current effective global constraint matrix. Based on the changes in PMI parameters that have not yet been applied and the remaining assembly constraints that are currently in effect, the pose change of the second part is tentatively calculated. When the tentative result causes the third or fourth plane to separate from the first or second plane along the direction of the corresponding increase in clearance, the corresponding state-dependent constraints are updated to the released state. When the current effective global constraint matrix and the current effective global PMI constraint influence matrix are re-formed, the corresponding constraint rows, PMI constraint influence items, and the corresponding final effective tolerance change transmission relationships are stopped. After the release judgment is completed, all currently effective assembly constraints are used together to solve the assembly state. When there is a second part pose change vector that can simultaneously satisfy all currently effective assembly constraints, the position and orientation of the second part are updated according to the common satisfaction result. When there is no second part pose change vector that can simultaneously satisfy all currently effective assembly constraints, the combination of PMI change parameters is determined as a state in which the assembly relationship cannot be satisfied simultaneously under the current constraint state.
[0145] After updating the position and orientation of the second part, the system re-determines whether the third and fourth assembly constraints, which are currently in a released state, meet the conditions for entering an effective state, based on the current gap between the first and third planes and the current gap between the second and fourth planes.
[0146] When the third or fourth assembly constraint re-enters the active state from the released state, if the corresponding plane reaches the contact boundary and does not have a tendency to separate along the direction that increases the corresponding gap, or if it will cross the contact boundary if it continues according to the current relative motion, the contact boundary is used as the state switching position. The corresponding first-order constraint relationship and the corresponding PMI constraint influence matrix are re-determined according to the current position and current direction of the corresponding plane after entering the active state. The currently active global constraint matrix and the currently active global PMI constraint influence matrix are re-formed. The corresponding PMI constraint activation relationship is added to the currently used final effective tolerance variation transmission relationship. The target allowable motion state and the number of independent motion restrictions are re-evaluated to see if they meet the requirements.
[0147] Repeat the state judgment and assembly state solution until the current pose of the second part corresponds to the current state of the third assembly constraint and the fourth assembly constraint.
[0148] After completing the above processing, the three-dimensional tolerance analysis results corresponding to the PMI variation parameter combination are obtained based on the final position and orientation of the second part, the current distance between the first and third planes, the current distance between the second and fourth planes, and the corresponding current clearance.
[0149] In this specific implementation, after the second PMI variation parameter corresponding to the second dimensional tolerance PMI changes, the current distance between the second plane and the fourth plane changes, and the corresponding gap decreases accordingly. The fourth assembly constraint changes from the released state to the active state, and the assembly constraints currently involved in solving the assembly state change. The solution relationship is re-formed based on the new current assembly state. In each assembly state solution stage, the current state of the third assembly constraint and the fourth assembly constraint is determined based on the relative motion direction and the updated gap, thereby fully realizing the continuous transmission between PMI changes, geometric state changes, state-dependent constraint switching, and three-dimensional tolerance analysis model updates.
[0150] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.
Claims
1. A PMI-driven automatic tolerance modeling method, characterized in that, include: Obtain the Parts Assembly Indicators (PMIs) and assembly relationships between parts in the 3D assembly model; generate PMI structured data based on the PMIs; and generate candidate assembly relationships based on the assembly relationships. Based on the PMI structured data, PMI variation parameters are generated to represent the geometric changes of the geometric features marked by the PMI relative to the nominal state. The current assembly state is generated based on the PMI variation parameters, and assembly constraints that limit the relative motion of parts are generated based on the candidate assembly relationships. Assembly constraints that continuously limit the relative motion of parts are determined as continuous constraints and are determined as active states. Assembly constraints that restrict the relative motion in the corresponding direction after the corresponding geometric feature reaches the contact boundary are determined as state-dependent constraints. The state-dependent constraints are determined as active states or released states based on the gap between the corresponding geometric features and the direction of relative motion. Based on the influence of the PMI variation parameters on the constraint conditions of the currently active assembly constraints, or on the gap and relative position required to reach the contact boundary of the currently released state-dependent constraints, establish the PMI constraint activation relationship between PMI and assembly constraints. Based on the product assembly function, a target allowable motion state is formed. Based on the target allowable motion state, the assembly constraints that are currently in effect are filtered. Assembly constraints with the same independent motion limit but different effects of PMI variation parameters on the constraint conditions are retained. State-dependent constraints that are currently in the released state and have PMI constraint activation relationships are also retained. A final set of assembly constraints is formed. Based on the assembly constraints that are currently in effect and have PMI constraint activation relationships, a final effective tolerance variation transfer relationship is formed. A three-dimensional tolerance analysis model is generated based on the PMI variation parameters, the final assembly constraint set, and the final effective tolerance variation propagation relationship. The model is used to calculate the part pose change. For pose components not determined by the currently effective assembly constraints, the model maintains the corresponding values of the current assembly state. When the current state of the state-dependent constraint changes, the assembly constraints and the final effective tolerance variation propagation relationship involved in the solution are updated according to the target allowed motion state until the current pose of the part corresponds to the current state of the state-dependent constraint, thus obtaining the three-dimensional tolerance analysis result.
2. The PMI-driven automatic tolerance modeling method according to claim 1, characterized in that, The PMI structured data includes the PMI identifier, tolerance type, and geometric features marked by the PMI; the data corresponding to dimensional tolerances also includes the nominal size and upper and lower deviations, or the upper and lower limit sizes; the data corresponding to geometric tolerances also includes geometric tolerance items, tolerance values, and the datum identifier and corresponding geometric features when datum requirements exist. Based on the PMI structured data, the permissible dimensional, positional, directional, or surface positional changes of the geometric features are determined, and one or more numerical values representing the corresponding changes are used to form PMI variation parameters. When the dimensional tolerance specifies the distance between two geometric features, the dimensional deviation of the current distance relative to the nominal distance is determined as the PMI variation parameter, and the PMI variation parameter is used as the relative positional change between the two geometric features along the measurement direction of the distance in the nominal state. When multiple PMI variation parameters jointly represent a PMI, their combination satisfies the permissible conditions of the corresponding tolerance zone of the PMI.
3. The PMI-driven automatic tolerance modeling method according to claim 1, characterized in that, For the state-dependent constraint, when there is a gap between two corresponding geometric features, it is determined to be in a released state; when the two geometric features reach the contact boundary and have a tendency to separate along the direction that increases the gap, it is determined to be in a released state; when they reach the contact boundary and do not have the tendency to separate, it is determined to be in an active state. In the released state, when the current relative motion continues and crosses the contact boundary, the state is updated to the active state using the contact boundary as the state switching position. When performing a release judgment in the active state, the dependent constraints of the state to be judged are not applied temporarily. The part pose change is calculated based on the changes in the PMI variation parameters that have not yet been applied and the other assembly constraints that are currently in the active state. The relative motion direction is determined based on the calculation results. When a motion trend of separation along the direction that increases the gap occurs, the state is updated to release.
4. The PMI-driven automatic tolerance modeling method according to claim 1, characterized in that, For an assembly constraint that is currently in effect, a constraint Jacobian matrix is formed based on the first-order change relationship between its independent constraint condition and the pose change of the part involved in the assembly constraint, and a PMI constraint influence matrix is formed based on the first-order change relationship between the independent constraint condition and the change amount of the PMI variation parameter. For state-dependent constraints currently in the release state, a PMI constraint influence matrix is formed based on the first-order relationship between the corresponding gap or the relative position required to reach the contact boundary and the change in the PMI variation parameter. When the rank of the PMI constraint influence matrix is greater than zero, the corresponding PMI constraint activation relationship is established; after the state-dependent constraint enters the effective state, the corresponding constraint Jacobian matrix and PMI constraint influence matrix are re-formed according to the current assembly state when it enters the effective state.
5. The PMI-driven automatic tolerance modeling method according to claim 1, characterized in that, The independent motions in the target's allowed motion states are arranged in the order of their degrees of freedom in the joint solution to form the target's allowed motion basis matrix; For assembly constraints that are currently in effect and have PMI constraint activation, a constraint matrix is formed based on their first-order constraint relationship on each independent motion degree of freedom. For state-dependent constraints that are currently in effect, each independent motion in the target's allowed motion state is judged one by one according to the current constraint direction. When there is an independent motion along the direction that reduces the corresponding gap, it is determined that the state-dependent constraint restricts the target's allowed motion state; otherwise, it is determined that the state-dependent constraint does not restrict the target's allowed motion state. For assembly constraints other than state-dependent constraints that are currently in effect, when the product of the constraint matrix and the target's allowed motion basis matrix is zero, it is determined that the assembly constraint does not restrict the target's allowed motion state. For state-dependent constraints that are currently in a released state and have PMI constraint activation relationships, retain their constraint information, PMI constraint activation relationships, and current state, and re-perform the aforementioned judgment after they enter an effective state.
6. The PMI-driven automatic tolerance modeling method according to claim 5, characterized in that, The selection constraint matrix is composed of the currently active assembly constraints that satisfy the target allowed motion state and are used to provide independent motion restrictions; When the rank of the selection constraint matrix increases after adding the assembly constraint to be judged, the assembly constraint to be judged is added to the selection constraint set. When the rank does not increase and the constraint matrix of the assembly constraint to be judged can be represented according to the linear relationship of the existing assembly constraints in the selection constraint set, a PMI constraint influence matrix is formed according to the first-order change relationship between the constraint conditions of the assembly constraint to be judged and the constraint conditions of the existing assembly constraints relative to the change in PMI variation parameters, and the corresponding columns are arranged according to a unified order of PMI variation parameters. The same set of linear combination coefficients of the linear relationship is used to determine whether the PMI constraint influence matrix maintains the same linear relationship. If it does, the existing assembly constraints represent the corresponding independent motion restrictions and the influence of PMI variation parameters on the constraint conditions. If it does not, the assembly constraint to be judged is added to the selection constraint set and the corresponding PMI constraint activation relationship is retained, but it is not counted as a newly added independent motion restriction.
7. The PMI-driven automatic tolerance modeling method according to claim 6, characterized in that, When the set of selection constraints still needs to add independent motion restrictions, select assembly constraints that satisfy the target allowed motion state and can increase the rank of the selection constraint matrix from the remaining currently effective assembly constraints that have not been added to the set of selection constraints, and add them as closed supplementary constraints to the selection constraint matrix. When all independent motions other than the target's allowed motion state need to be restricted, the number of independent motion restrictions is determined to meet the requirements when the rank of the constraint matrix is equal to the difference between the number of degrees of freedom of the independent motions participating in the joint solution and the number of independent motions retained in the target's allowed motion state. When the independent motion corresponding to the state-dependent constraint currently in the released state does not need to be restricted by other assembly constraints currently in the active state, the independent motion is not included in the independent motion that needs to be restricted, and the corresponding pose components not determined by the assembly constraints currently in the active state retain the values corresponding to the current assembly state. For state-dependent constraints that restrict only one direction of independent motion, when the rank of the selection constraint matrix increases, the current restriction direction is counted as an independent motion constraint, and its current state is reassessed in each assembly state solution stage.
8. The PMI-driven automatic tolerance modeling method according to claim 1, characterized in that, For PMI, establish a record of PMI identifier, labeled geometric features, PMI variation parameters and their allowable ranges. Based on the assembly constraints currently in effect in the final assembly constraint set, form a currently effective global constraint matrix. Then, form a currently effective global PMI constraint influence matrix by combining the PMI constraint influence matrices corresponding to each PMI variation parameter with the same constraint order and PMI variation parameter order. After giving the change amount of PMI variation parameters and completing the current state update of assembly constraints, a common solvability judgment is made on all assembly constraints currently in effect in the final assembly constraint set. If there is a part pose change that can jointly satisfy the assembly constraint, the part pose change is determined. If not, the corresponding PMI variation parameter combination is determined as a state in which the assembly relationship cannot be satisfied simultaneously under the current constraint state. When a state-dependent constraint moves from the released state to the active state, the corresponding PMI constraint activation relationship is added to the currently used final effective tolerance variation transitive relationship. When it moves from the active state to the released state, the corresponding final effective tolerance variation transitive relationship is stopped.
9. The PMI-driven automatic tolerance modeling method according to claim 8, characterized in that, In each assembly state solution stage, the current assembly state at the beginning of that stage is used as the reference state. When the current state of a state-dependent constraint changes, the current assembly state at the time of the state change is used as the reference state for the next assembly state solution stage, and the solution relationship is re-formed using the change in PMI parameters that have not yet been applied relative to the reference state. When there are multiple PMI change parameters, each PMI change parameter is made to change synchronously from the current value corresponding to the reference state to the given value corresponding to the current analysis in the same proportion. The synchronous change proportion that first meets the state switching condition is determined. The changes that have occurred when the synchronous change proportion is reached are taken as the part that has been acted, and the remaining changes are taken as the part that has not yet been acted. When multiple state-dependent constraints meet the state switching conditions under the same synchronous change ratio, their current states are updated simultaneously until all the changes in the given PMI variation parameters have been completed and the current pose of the part corresponds to the current state of each state-dependent constraint.
10. The PMI-driven automatic tolerance modeling method according to claim 1, characterized in that, Based on the final obtained current pose of the part and the current position, orientation, and size of the geometric features, determine the size, position, orientation, or gap result to be analyzed; When performing extreme value analysis, PMI variation parameters are selected within the allowable range of PMI variation parameters to solve the assembly state; when performing statistical analysis, the probability distribution and distribution parameters are determined based on the manufacturing process statistical data of the geometric changes represented by the corresponding PMI, and the PMI variation parameters are obtained based on the probability distribution and distribution parameters, and the assembly state is solved repeatedly; when multiple PMI variation parameters jointly represent the same PMI, only the combination of PMI variation parameters that jointly satisfy the allowable conditions of the tolerance zone corresponding to the PMI is used.
