A wheelchair customization optimization method and device based on virtual simulation and a medium
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- MAGLITE (SHANGHAI) TECH CO LTD
- Filing Date
- 2026-02-27
- Publication Date
- 2026-06-09
AI Technical Summary
Existing wheelchair customization methods suffer from inefficient design parameter optimization and insufficient accuracy in matching simulation models with actual user needs, resulting in inadequate accuracy and reliability of the customization solutions.
By scanning user posture data and wheelchair component shapes, a set of 3D models is generated. Dimensional data is extracted and bound with unique data codes to form measurement data import records. Combined with user needs mapping, a virtual simulation configuration package is generated. A multi-objective constraint relationship model between design variables, geometric constraints, and design preferences is constructed. Parameter perturbation sampling and simulation feedback iteration are performed to generate a parameter sensitivity ranking sequence. Parameter adjustments and geometric verification are performed to generate a comprehensive evaluation sequence, and finally, a customized wheelchair solution is generated.
It achieves high-precision 3D data acquisition and security identification, improving the accuracy, security and efficiency of customized solutions. The geometric consistency and performance indicators of the design scheme are verified through twin simulation.
Smart Images
Figure CN122174364A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of computer-aided design technology, and in particular to a method, device and medium for customizing and optimizing wheelchairs based on virtual simulation. Background Technology
[0002] In the field of wheelchair customization, conventional methods typically rely on a combination of computer-aided design (CAD) software and anthropometry data to manually or semi-automatically adjust design variables such as seat width, seat depth, and backrest height. These methods, based on standard anthropometric databases or limited manual measurements, utilize virtual simulation to build 3D models to predict the wheelchair's structural performance and user comfort. Twin simulation, a recent trend, allows for the simulation of physical entity behavior in a digital environment, supporting design validation. These methods assist designers in assessing the feasibility of solutions during the preliminary design phase and iteratively optimizing them based on ergonomic principles, thus laying the foundation for personalized wheelchair customization and advancing the fields of rehabilitation engineering and assistive technology.
[0003] However, conventional methods still have certain limitations in practical applications. On the one hand, the design parameter optimization process relies heavily on manual experience and trial-and-error iterations, lacking a systematic automatic screening mechanism, resulting in low optimization efficiency and difficulty in efficiently exploring large-scale design spaces. On the other hand, the matching accuracy between simulation models and users' actual needs is insufficient. Due to the discrepancy between simulation conditions and personalized usage scenarios, the accuracy and reliability of customized solutions are affected. These factors restrict the rapid response and performance optimization of the wheelchair customization process, especially given the increasingly prominent demand for high-precision personalization, and require further improvement. Summary of the Invention
[0004] In view of the aforementioned existing problems, the present invention is proposed.
[0005] Therefore, this invention provides a virtual simulation-based wheelchair customization optimization method to solve the problems of low efficiency in design parameter optimization and insufficient accuracy in matching simulation models with users' real needs.
[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a wheelchair customization optimization method based on virtual simulation, comprising: scanning user sitting posture data and wheelchair component outlines to generate a set of three-dimensional models; extracting dimensional data from the set of three-dimensional models and binding them with unique data codes to form a measurement data import record; performing unit verification and validity verification on the measurement data import record, and generating a virtual simulation configuration package based on user requirements mapping; constructing a multi-objective constraint relationship model between design variables, geometric constraints, and design preferences based on the virtual simulation configuration package, and selecting a set of candidate design parameters under the constraints of the virtual simulation configuration package; evaluating performance response changes based on the set of candidate design parameters through a parameter perturbation sampling and simulation feedback iteration mechanism to generate a parameter sensitivity ranking sequence; performing parameterized adjustment and geometric verification based on the parameter sensitivity ranking sequence to generate a geometric consistency evaluation, and generating a comprehensive evaluation sequence based on the design performance evaluation; selecting the optimal set of design parameters from the comprehensive evaluation sequence, associating it with the data code, and generating a wheelchair customization scheme confirmation form.
[0007] As a preferred embodiment of the virtual simulation-based wheelchair customization and optimization method described in this invention, the steps for generating a set of three-dimensional models by scanning user posture data and wheelchair component shapes are as follows: Based on customer information records, a data code bound to the customer information records is generated through a unique identifier generation rule; The model sequentially scans the seat shape, cushion shape, and backrest shape, and then aggregates them to generate a set of 3D models.
[0008] As a preferred embodiment of the virtual simulation-based wheelchair customization and optimization method described in this invention, the steps of extracting dimensional data and binding it with a unique data code to form a measurement data import record are as follows: The sitting posture dimensions, seat cushion dimensions, and backrest dimensions are extracted from the 3D scan model and then organized into a measurement data set according to the field names. Write the data code and measurement data set into the configuration environment to generate a measurement data import record.
[0009] As a preferred embodiment of the virtual simulation-based wheelchair customization and optimization method described in this invention, the steps of performing unit verification and validity validation on the imported measurement data records, and generating a virtual simulation configuration package based on user requirements, are as follows: Based on the imported measurement data, the unit consistency and validity verification of the sitting posture dimensions, seat cushion dimensions, and backrest dimensions are performed and mapped to a set of design boundary parameters. Extract configuration item selections from the configuration environment and map them to a set of design preference parameters; The set of design boundary parameters and the set of design preference parameters are used as geometric constraints and design preference conditions, and integrated into a virtual simulation configuration package.
[0010] As a preferred embodiment of the wheelchair customization optimization method based on virtual simulation described in this invention, the steps include: constructing a multi-objective constraint relationship model among design variables, geometric constraints, and design preferences based on a virtual simulation configuration package, and selecting a set of candidate design parameters under the constraints of the virtual simulation configuration package. Based on the hypergraph constraint network, the design variables in the virtual simulation configuration package are defined as parameter nodes, and the geometric constraints and design preference conditions are defined as constraint nodes. The connection relationship between parameter nodes and constraint nodes is established according to the dependency relationship, and a multi-objective constraint relationship model between design variables, geometric constraints and design preferences is constructed. Under the constraints of the multi-objective constraint relationship model, define the allowed value range and value step size, and generate the design parameter domain definition record; Select a set of candidate design parameters that satisfy the constraints of the virtual simulation configuration package from the design parameter domain definition records.
[0011] As a preferred embodiment of the virtual simulation-based wheelchair customization and optimization method described in this invention, the steps of evaluating performance response changes and generating a parameter sensitivity ranking sequence based on a set of candidate design parameters through parameter perturbation sampling and simulation feedback iteration mechanisms are as follows. Based on the candidate design parameter set, records are defined according to the design parameter domain, single-parameter disturbance sampling is performed, and disturbance sampling records are generated. Based on the disturbance sampling records, the virtual simulation configuration package is driven to perform twin simulation inference and perform differential operations to generate a sequence of performance response changes; The performance response change sequence is normalized to determine the parameter sensitivity, and then sorted to form a parameter sensitivity ranking sequence.
[0012] As a preferred embodiment of the virtual simulation-based wheelchair customization and optimization method of the present invention, the steps of performing parameter adjustment and geometric verification according to the parameter sensitivity ranking sequence to generate a geometric consistency evaluation, and combining it with the design performance evaluation to generate a comprehensive evaluation sequence, are as follows: Based on the candidate design parameter set and parameter sensitivity ranking sequence, the 3D scanning model in the configuration environment is parametrically adjusted to generate a wheelchair twin simulation model; The length and angle of the wheelchair twin simulation model are measured, and the geometric relationship is checked for consistency to generate a geometric consistency evaluation. The attribute fields of the wheelchair twin simulation model are calculated to generate corresponding weight index, size matching index and structural stability index, forming a design performance evaluation. Based on geometric consistency evaluation and design performance evaluation, a comprehensive evaluation sequence is generated by merging them according to the evaluation weight rules.
[0013] As a preferred embodiment of the virtual simulation-based wheelchair customization and optimization method described in this invention, the steps for generating the wheelchair customization plan confirmation form are as follows: The optimal set of design parameters is selected from the comprehensive evaluation sequence and written into the configuration items of the configuration environment to generate a configuration scheme output package; Associate the configuration solution output package with the data code to generate the corresponding wheelchair customization solution confirmation form.
[0014] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the virtual simulation-based wheelchair customization and optimization method described in the first aspect of the present invention.
[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the virtual simulation-based wheelchair customization and optimization method described in the first aspect of the present invention.
[0016] The beneficial effects of this invention are as follows: by scanning user sitting posture data and wheelchair component shapes to generate a set of three-dimensional models, high-precision three-dimensional data acquisition and safety label generation are achieved; parametric adjustment and geometric verification are performed based on the candidate design parameter set to form a comprehensive evaluation sequence, realizing parametric optimization and multi-index evaluation based on virtual simulation; and the geometric consistency and performance indicators of the design scheme are dynamically verified through twin simulation, thereby improving the accuracy, safety and efficiency of customized solutions. Attached Figure Description
[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 A flowchart illustrating a virtual simulation-based wheelchair customization and optimization method.
[0019] Figure 2 A flowchart for generating a configuration package for data acquisition and virtual simulation.
[0020] Figure 3 This is a flowchart for parameter selection and simulation evaluation.
[0021] Figure 4 A flowchart for generating and confirming customized solutions. Detailed Implementation
[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.
[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.
[0025] Reference Figures 1-4 This is one embodiment of the present invention, which provides a wheelchair customization and optimization method based on virtual simulation, including the following steps: S1. Scan the user's sitting posture data and the shape of the wheelchair components to generate a set of 3D models. Extract the size data from the 3D model set and bind a unique data code to form a measurement data import record.
[0026] Based on customer information records, a data code bound to the customer information records is generated through a unique identifier generation rule.
[0027] Furthermore, the customer's name, contact information, and sitting posture scan batch identifier fields are collected as customer information records. These records are then normalized and sequentially concatenated to form a feature string. The scan timestamp (rounded to the nearest second) is concatenated with the scan resolution field to form a perturbation source string. This perturbation source string is then subjected to fixed-length encoding to obtain a temporal perturbation factor. This factor is injected into the feature string to construct an anti-replay feature expression, enabling the same customer information record to generate different data codes under different scan batches and timestamps. Using the anti-replay feature expression as input and constrained by fixed parameters (e.g., a 256-bit key length, a 16-byte salt value, and 10,000 iterations), an irreversible mapping is performed to obtain a digest value. This digest value is then alphanumeric encoded and truncated to a fixed length (e.g., a 32-bit alphanumeric combination) to form a data code. Finally, a one-to-one binding relationship is established between the data code and the primary key of the customer information record, and the data code is written to the configuration environment.
[0028] It should be noted that all the above information has been agreed to by the user and is used for legitimate purposes; the configuration environment is the configuration data carrying space created after the user logs in, used to write measurement data import records and carry virtual simulation configuration packages, configuration item values and configuration scheme output packages.
[0029] The model sequentially scans the seat shape, cushion shape, and backrest shape, and then aggregates them to generate a set of 3D models.
[0030] Furthermore, using the data code as the scanning session index, the sitting posture shape scan, seat cushion shape scan, and backrest shape scan are performed sequentially to obtain the sitting posture shape point cloud model, seat cushion shape point cloud model, and backrest shape point cloud model. Using the sitting posture shape point cloud model as the spatial reference and the seat cushion shape point cloud model as the alignment reference, spatial registration and scale consistency fusion are performed in a unified coordinate system to form a three-dimensional model set, for example, the point cloud density is unified to 1000 points per square centimeter.
[0031] The sitting posture dimensions, seat cushion dimensions, and backrest dimensions are extracted from the 3D scan model and then organized into a unified measurement data set according to the field names.
[0032] Furthermore, the shoulder point, hip point, and lower limb support point are determined by an automatic anatomical reference point positioning algorithm, and the axial projection distance from the cross-sectional profile to the sitting posture reference plane is obtained along the main axis of the skeleton to obtain the sitting posture dimensions. Using the sitting posture reference plane in the sitting posture dimensions as a constraint, the contact area boundary line (the contact area is the area adjacent to the seat cushion point cloud and the sitting posture reference plane) is filtered and extracted according to the distance threshold. The support spine line and lateral boundary are located by the neighborhood normal change rate, and the seat cushion width, seat cushion depth, and support profile dimensions are obtained to obtain the seat cushion shape dimensions. The backrest midpoint is extracted layer by layer from the sitting posture reference plane and fitted as the spinal midline. The backrest height is obtained by performing endpoint height projection on the spinal midline, and the backrest curvature radius is obtained by sagittal plane arc least square fitting. Combined with the normal angle of the sitting posture reference plane, the tilt angle related dimensions are obtained and packaged to form the backrest shape dimensions. The sitting posture dimensions, seat cushion shape dimensions, and backrest shape dimensions are uniformly packaged into a measurement data set, for example, the size unit is uniformly set to millimeters.
[0033] It should be noted that the sitting posture reference plane is obtained from the pelvic posture plane determined by the anatomical reference points of the shoulder and hip points in the sitting posture shape point cloud model; the distance threshold is determined by taking the median distance from the cushion point cloud to the sitting posture reference plane and combining it with a safety margin, wherein the safety margin is an integer multiple of the scan resolution (e.g., 1 to 3 times), and the exemplary distance threshold range is 1 mm to 5 mm.
[0034] Write the data code and measurement data set into the configuration environment to generate a measurement data import record.
[0035] S2. Perform unit verification and validity validation on the imported measurement data records, and generate a virtual simulation configuration package based on user requirements.
[0036] Based on the imported measurement data, the unit consistency and validity verification of the sitting posture dimensions, seat cushion dimensions, and backrest dimensions are performed and mapped to a set of design boundary parameters.
[0037] Furthermore, the measurement data imported into the record for sitting posture dimensions, seat cushion dimensions, and backrest dimensions are converted to a unified unit, with lengths uniformly converted to mm and angles uniformly converted to rad. Based on the ergonomic constraint range, the validity of each dimension value is checked, and abnormal values that exceed the ergonomic constraint range and the feasible range of the wheelchair structure are removed. Based on the dimension values that pass the interval validity check, an expansion coefficient is combined to generate upper limit boundary values and lower limit boundary values. The sitting posture dimensions are mapped to the boundaries of seat width, seat depth, and support height; the seat cushion dimensions are mapped to the boundaries of support width and support depth; and the backrest dimensions are mapped to the boundaries of backrest height, radius of curvature, and tilt angle. A set of design boundary parameters is then generated, with the dimension unit uniformly converted to millimeters.
[0038] It should be noted that the ergonomic constraint range is a set of pre-set human measurement reference data for sitting posture in the configuration environment, used to limit the allowable range of sitting posture dimensions, seat cushion dimensions, and backrest dimensions (e.g., the allowable range of seat width is 360 mm to 520 mm); the expansion factor refers to the proportional coefficient that expands the dimension value to the lower limit boundary value. For example, the lower limit expansion factor is 0.95 to 0.99, and the upper limit expansion factor is 1.01 to 1.05.
[0039] Extract configuration item selections from the configuration environment and map them to a set of design preference parameters.
[0040] Furthermore, using the data code in the measurement data import record as an index, the configuration item selection record (including seat width configuration item, seat depth configuration item, backrest height configuration item, backrest tilt angle configuration item, and profile control configuration item) is read from the configuration environment; the discrete values of the seat width configuration item, seat depth configuration item, backrest height configuration item, and backrest tilt angle configuration item are read and converted into corresponding continuous preference values according to a preset discrete-continuous mapping relationship table; the level values of the profile control configuration item are read and mapped into weight coefficients according to a preset level-weight mapping relationship table; the range of the continuous preference values and weight coefficients is validated and the field order is uniformly organized to form a set of design preference parameters.
[0041] It should be noted that the discrete-continuous mapping relationship table is based on the seat width boundary, seat depth boundary, backrest height boundary, and backrest tilt angle boundary settings in the design boundary parameter set. When the configuration environment is created, the discrete value of each configuration item is bound to the target value within the boundary interval and written into the configuration item mapping rule record to form the discrete-continuous mapping relationship table. The fields of the discrete-continuous mapping relationship table are defined as configuration item name, discrete level, target continuous value, applicable boundary interval, and version number (for example, the seat width configuration item value "medium" is bound to 440 mm and meets the 360 mm to 520 mm boundary). The level-weight mapping relationship table is based on the evaluation weight rules. When the configuration environment is created, the profile control configuration item level value is bound to the weight coefficient in the range of 0 to 1 and written into the weight mapping rule record to form the level-weight mapping relationship table. The fields of the level-weight mapping relationship table are defined as configuration item name, level value, weight coefficient, applicable boundary interval, and version number. For example, level 2 is bound to 0.5 for the preference weight of profile control design variables (including seat width, seat depth, backrest height, backrest tilt angle, and profile control).
[0042] The set of design boundary parameters and the set of design preference parameters are used as geometric constraints and design preference conditions, and integrated into a virtual simulation configuration package.
[0043] Furthermore, the design boundary parameter set is mapped to the upper and lower limits of each design variable by field name and marked as geometric constraints; the continuous preference values and weight coefficients in the design preference parameter set are mapped to preference intensity parameters by field name and marked as design preference conditions, and then packaged in order to form a virtual simulation configuration package that can be directly used for subsequent twin simulation parameter screening and evaluation.
[0044] S3. Based on the virtual simulation configuration package, construct a multi-objective constraint relationship model between design variables, geometric constraints, and design preferences, and select a set of candidate design parameters under the constraints of the virtual simulation configuration package.
[0045] Based on the hypergraph constraint network, the design variables in the virtual simulation configuration package are defined as parameter nodes, and the geometric constraints and design preference conditions are defined as constraint nodes. The connection relationship between parameter nodes and constraint nodes is established according to the dependency relationship, and a multi-objective constraint relationship model between design variables, geometric constraints and design preferences is constructed.
[0046] Furthermore, based on the design boundary parameter set and the design preference parameter set, initial node connection weights of the hypergraph constraint network are randomly generated. Multiple sample pairs are extracted from the historical wheelchair customization case library. Through hypergraph convolution operation, the constraint satisfaction score of the current hypergraph network is obtained, and the node connection weights are dynamically adjusted according to the gradient descent method to improve the constraint satisfaction. During training iterations, the hyperedge structure is automatically reconstructed according to the constraint satisfaction change rate. The high-order relation structure is optimized by merging high-similarity constraint nodes and splitting low-relevance constraint nodes. When the constraint satisfaction improvement of multiple consecutive iterations (e.g., 5 times) is less than the satisfaction threshold or the maximum number of iterations (e.g., 100 times) is reached, the training process stops, and the trained hypergraph constraint network is obtained.
[0047] Using the trained hypergraph constraint network, the seat width, seat depth, backrest height, backrest tilt angle, and contour control in the virtual simulation configuration package are defined as discrete-continuous hybrid parameter nodes, and their allowed value ranges and value step sizes are written into them. The upper and lower limits of the values obtained by mapping the design boundary parameter set and the preference intensity parameters obtained by mapping the design preference parameter set are defined as constraint nodes, respectively. A one-to-one bidirectional connection relationship is established between the parameter nodes and the constraint nodes, and the weight coefficients determined by the design preference parameter set are written into the connection relationship to represent the influence of the preference intensity parameters in the constraint satisfaction determination. After completing the connection of all parameter nodes and constraint nodes, a multi-objective constraint relationship model between design variables, geometric constraints, and design preferences is formed.
[0048] It should be noted that the satisfaction threshold is based on historical wheelchair customization cases. The optimal threshold is determined by calculating the average number of convergence iterations and constraint satisfaction under each threshold. An exemplary value range is 0.1% to 1.0%. The historical wheelchair customization case library is a structured database that stores historical wheelchair customization data (including design variable values and constraint satisfaction fields).
[0049] Under the constraints of the multi-objective constraint relationship model, the allowed value range and value step size are defined, and the design parameter domain definition record is generated.
[0050] Furthermore, under the constraints of the multi-objective constraint relationship model, the set of constraint nodes participating in the constraint satisfaction determination for each parameter node is determined by the connection relationship, and weighted arc consistent pruning is performed on the initial allowable value range of the parameter node. Specifically, for each value point within the initial allowable value range, the corresponding constraint node is substituted to complete the satisfaction determination. Value points that fail to satisfy the determination are directly removed, and the minimum and maximum values of the set of value points are retained as the pruned allowable value range. Based on the length of the pruned allowable value range, the value step size is obtained by combining the discrete level (e.g., 32 levels), and encapsulated to form a design parameter domain definition record (e.g., the allowable value range of the seat width design variable is 360 mm to 520 mm and the value step size is 5 mm). The discrete level refers to the number of consecutive design parameter ranges discretized into a finite number of sampling levels. The larger the discrete level, the more refined the parameter search.
[0051] It should be noted that the step size rule is set based on the geometric resolution requirements and simulation accuracy requirements corresponding to the design variables in the multi-objective constraint relationship model. The step size is determined by the length of the allowed value range and the number of discrete levels (the exemplary values are 20 to 80 levels). For example, if the seat width design variable is in the range of 360 mm to 520 mm, 33 levels are taken, corresponding to a step size of 5 mm.
[0052] Select a set of candidate design parameters that satisfy the constraints of the virtual simulation configuration package from the design parameter domain definition records.
[0053] Furthermore, based on the value range and value step size of each design variable in the design parameter domain definition record, parameter combinations are constructed, and geometric constraints and design preference conditions are directly embedded. Only parameter combinations that simultaneously satisfy boundary intervals, preference weight thresholds, and variable association constraints are retained, and parameter combinations that do not satisfy any constraint conditions are pruned immediately, thus forming a set of candidate design parameters.
[0054] It should be noted that the preference weight threshold is calculated based on the corresponding weight coefficients within the design preference parameter set, converted at a fixed ratio (e.g., 0.6), and used as the lower limit for screening. The larger the preference weight threshold, the smaller the candidate set size, and the higher the risk of false screening. An exemplary value range is 0 to 1. The variable association constraint is based on the boundary interval given by the design boundary parameter set, and combined with the preference weights in the design preference parameter set to trigger segmented contraction to form a linkage constraint (e.g., when the seat width is greater than 480 mm, the backrest tilt angle is not less than 90 degrees).
[0055] S4. Based on the candidate design parameter set, the performance response changes are evaluated through parameter perturbation sampling and simulation feedback iteration mechanism to generate a parameter sensitivity ranking sequence.
[0056] Based on the candidate design parameter set, records are defined according to the design parameter domain, and single-parameter disturbance sampling is performed to generate disturbance sampling records.
[0057] Furthermore, each parameter combination in the candidate design parameter set is used as a baseline parameter group. The values of the remaining design variables, except for the first design variable, are fixed. According to the value step size of the corresponding design variable in the design parameter domain definition record, the first design variable is shifted sequentially within the allowable value range to generate disturbance values, forming a single-parameter disturbance parameter group sequence around the baseline parameter group. The single-parameter disturbance process is sequentially performed on the seat width, seat depth, backrest height, backrest tilt angle, and contour control, while keeping the values of the remaining design variables unchanged. The baseline parameter group identifier, design variable identifier, disturbance value, value step size, and parameter group sequence number are encapsulated to generate a disturbance sampling record. For example, the seat width is in the range of 360 mm to 520 mm, and a disturbance sampling sequence is formed with a step size of 5 mm.
[0058] Based on the disturbance sampling records, the virtual simulation configuration package is driven to perform twin simulation inference and perform differential operations to generate a sequence of performance response changes.
[0059] Furthermore, the disturbance design parameter groups are read sequentially according to the parameter group numbers in the disturbance sampling record, and each disturbance design parameter group is written into the design variable value field in the virtual simulation configuration package. This drives the corresponding wheelchair twin simulation model to complete a full twin simulation derivation, obtaining the performance response that corresponds one-to-one with the disturbance design parameter group. The performance response corresponding to the baseline parameter group that has not undergone single-parameter disturbance sampling is used as the baseline performance response value. A term-by-term difference operation is performed on the performance response under different disturbance values of the same design variable to obtain the performance response change. The performance response change is sequentially encapsulated according to the parameter group order in the disturbance sampling record to generate a performance response change sequence.
[0060] The performance response change sequence is normalized to determine the parameter sensitivity, and then sorted to form a parameter sensitivity ranking sequence.
[0061] Furthermore, using the baseline performance response value as the normalization benchmark, each change in the performance response change sequence is normalized to obtain a dimensionless change rate sequence; the mean of the dimensionless change rate sequences corresponding to the same design variable is taken to obtain the parameter sensitivity; and the parameter sensitivity is sorted from largest to smallest to obtain the parameter sensitivity ranking sequence.
[0062] S5. Perform parameterized adjustments and geometric verification based on the parameter sensitivity sorting sequence to generate a geometric consistency evaluation, and combine it with the design performance evaluation to generate a comprehensive evaluation sequence.
[0063] Based on the candidate design parameter set and parameter sensitivity ranking sequence, the 3D scanning model in the configuration environment is parametrically adjusted to generate a wheelchair twin simulation model.
[0064] Furthermore, the priority of design variable updates is determined based on the parameter sensitivity ranking sequence. According to the candidate design parameter set, the 3D scanning model in the configuration environment is transformed into a parametric control mesh model, and the seat width, seat depth, backrest height, and backrest tilt angle are mapped to control point displacement and rotation, respectively. Weighted linear deformation is performed on the vertices of the parametric control mesh model to complete the parametric update. Using contour control as the curvature adjustment coefficient, curvature continuity constraints are applied to the seat cushion shape point cloud and backrest shape point cloud to generate a wheelchair twin simulation model with the corresponding candidate parameter combination.
[0065] It should be noted that curvature continuity constraint refers to limiting the rate of curvature change of adjacent mesh surfaces within the contact area between the seat cushion and backrest point cloud models to a preset curvature change threshold, thus ensuring a continuous surface transition. The curvature change threshold is set based on the scanning noise scale and the control mesh resolution. It is determined by statistically analyzing the local curvature change rate distribution in the contact area between the seat cushion and backrest point cloud models and taking the upper quantile value (e.g., the 0.90 quantile) as the threshold. An exemplary value range is 0.02 mm. -1 Up to 0.10mm -1 The curvature is calculated as the average curvature.
[0066] The length and angle of the wheelchair twin simulation model are measured, and the geometric relationship is checked for consistency to generate a geometric consistency evaluation.
[0067] Furthermore, under a unified coordinate system, endpoint distance measurements are performed in the seat width, seat depth, and backrest height directions, and angle measurements are performed based on the sagittal plane normal to obtain the corresponding length and angle values. The measured length and angle values are then compared item by item with the geometric constraints in the virtual simulation configuration package to verify whether each design variable falls within the allowable value range and satisfies the variable association constraints. All wheelchair twin simulation models that pass the verification are marked as having passed the geometric consistency status and are aggregated to form a geometric consistency evaluation, which is used for design performance evaluation and comprehensive evaluation sequence generation.
[0068] The attribute fields of the wheelchair twin simulation model are calculated to generate corresponding weight index, size matching index and structural stability index, forming a design performance evaluation.
[0069] Furthermore, based on the wheelchair twin simulation model that has passed the geometric consistency evaluation, using the parametric control mesh model and attribute fields as input, numerical integration is performed on the volume fields of each structural component in the parametric control mesh model, combined with the material density table, to obtain the weight index in mm³. Based on the weight index, the seat width, seat depth, backrest height, and backrest tilt angle of the wheelchair twin simulation model are subjected to item-by-item deviation calculation with the corresponding sitting posture dimensions to obtain deviation values. The deviation values are then normalized according to the ergonomic constraint interval to obtain the size matching index. Based on the size matching index, and based on the geometric topology and contact area distribution of the wheelchair twin simulation model, under standard load conditions, the center of gravity projection position is obtained through orthogonal projection. The minimum Euclidean distance from the center of gravity projection position to the boundary of the supporting polygon is used as the rollover margin, and a rollover margin of 0 is used as the rollover critical condition. The rollover margin is mapped to the structural stability index according to the normalized interval. Finally, the weight index, size matching index, and structural stability index are encapsulated in a unified field order to form a design performance evaluation.
[0070] It should be noted that the material density table contains preset material density parameters in the configuration environment, storing material density values according to material number (e.g., the density of aluminum alloy is 2.7 grams per cubic centimeter); the standard load conditions are the vertical downward concentrated and distributed forces applied to the wheelchair twin simulation model by the load parameters in the virtual simulation configuration package (e.g., the load parameter is 800 Newtons), where the concentrated force is applied to the center of the seat surface and the distributed force is applied to the seat cushion contact area, for example, the concentrated force accounts for 60% and the distributed force accounts for 40%; the geometric topology relationship is the connection and adjacency relationship between geometric points in the wheelchair twin simulation model (including wheel contact points, foot pedal contact points and anti-roll wheel contact points).
[0071] The formula for calculating the size matching index is: ; in, For size matching indicators; The total number of the dimensions involved in the calculation (including seat width, seat depth, backrest height, and backrest tilt angle); This is the index subscript for the dimension item; The first in the wheelchair twin simulation model Measurement values for each dimension item; Import measurement data into the record and the first The sitting posture dimension values corresponding to each dimension item; The first in the ergonomic constraint interval Upper limit boundary value for each dimension item; The first in the ergonomic constraint interval The lower boundary value of each dimension item.
[0072] Based on geometric consistency evaluation and design performance evaluation, a comprehensive evaluation sequence is generated by merging them according to the evaluation weight rules.
[0073] Furthermore, based on geometric consistency evaluation and design performance evaluation, the weight index, size matching index and structural stability index in the design performance evaluation are weighted and fused according to the evaluation weight rules. Among them, geometric consistency evaluation is used as a hard constraint screening condition to directly eliminate the unqualified items. Among the qualified items, the weight index, size matching index and structural stability index are normalized and weighted and summed according to the evaluation weight rules to obtain the comprehensive evaluation value. The comprehensive evaluation sequence is then organized according to the combination order of candidate design parameters.
[0074] It should be noted that the evaluation weight rules are set based on the correspondence between the preference weights in the design preference parameter set and the evaluation items (weight index, size matching index and structural stability index) in the virtual simulation configuration package. For example, the weight index has a weight of 0.2, the size matching index has a weight of 0.4, and the structural stability index has a weight of 0.4. The sum of the weights is 1, and the weight values are in the range of [0, 1].
[0075] S6. Select the optimal set of design parameters from the comprehensive evaluation sequence, associate them with the data code, and generate a wheelchair customization plan confirmation form.
[0076] The optimal set of design parameters is selected from the comprehensive evaluation sequence and written into the configuration items of the configuration environment to generate a configuration scheme output package.
[0077] Furthermore, the candidate design parameter combinations are sorted from high to low according to the comprehensive evaluation values. The design parameter combination that passes the geometric consistency evaluation and has the highest comprehensive evaluation value is taken as the optimal design parameter set. The seat width, seat depth, backrest height, backrest tilt angle and contour control in the optimal design parameter set are written back to the corresponding configuration item value fields in the configuration environment according to the field name, forming a configuration scheme output package, which contains the values of each configuration item in the optimal design parameter set.
[0078] Associate the configuration solution output package with the data code to generate the corresponding wheelchair customization solution confirmation form.
[0079] Furthermore, the final values of the seat width configuration item, seat depth configuration item, backrest height configuration item, backrest tilt angle configuration item, and contour control configuration item contained in the configuration scheme output package are bound one-to-one with the data code. The data code is written into the scheme identifier field of the configuration scheme output package, and the bound configuration scheme output package is solidified into a customized scheme record according to the field order, generating a wheelchair customized scheme confirmation form that simultaneously contains data code, configuration item values, and scheme identifier information.
[0080] This embodiment also provides a computer device applicable to the wheelchair customization and optimization method based on virtual simulation, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the wheelchair customization and optimization method based on virtual simulation as proposed in the above embodiment.
[0081] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.
[0082] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the wheelchair customization and optimization method based on virtual simulation as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.
[0083] In summary, this invention achieves high-precision 3D data acquisition and safety label generation by: scanning user posture data and wheelchair component shapes to generate a set of 3D models; performing parametric adjustments and geometric verification based on the candidate design parameter set to form a comprehensive evaluation sequence, realizing parametric optimization and multi-index evaluation based on virtual simulation; and dynamically verifying the geometric consistency and performance indicators of the design scheme through twin simulation, thereby improving the accuracy, safety, and efficiency of customized solutions.
[0084] It should be noted that 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 preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A wheelchair customization and optimization method based on virtual simulation, characterized in that: include, Scan user posture data and wheelchair component shapes to generate a set of 3D models, extract dimensional data from the 3D model set and bind a unique data code to form a measurement data import record; The imported measurement data records undergo unit verification and validity validation, and a virtual simulation configuration package is generated based on user requirements. Based on the virtual simulation configuration package, a multi-objective constraint relationship model is constructed between design variables, geometric constraints, and design preferences, and a set of candidate design parameters is selected under the constraints of the virtual simulation configuration package. Based on the candidate design parameter set, the performance response changes are evaluated through parameter perturbation sampling and simulation feedback iteration mechanism, and a parameter sensitivity ranking sequence is generated. Based on the parameter sensitivity ranking sequence, parameter adjustments and geometric verification are performed to generate a geometric consistency evaluation, and combined with the design performance evaluation, a comprehensive evaluation sequence is generated. The optimal set of design parameters is selected from the comprehensive evaluation sequence and associated with the data code to generate a wheelchair customization solution confirmation form.
2. The wheelchair customization and optimization method based on virtual simulation as described in claim 1, characterized in that: The steps for generating a set of 3D models from the scanned user posture data and wheelchair component shapes are as follows: Based on customer information records, a data code bound to the customer information records is generated through a unique identifier generation rule; The model sequentially scans the seat shape, cushion shape, and backrest shape, and then aggregates them to generate a set of 3D models.
3. The wheelchair customization and optimization method based on virtual simulation as described in claim 2, characterized in that: The steps for extracting dimensional data and binding it with a unique data code to form a measurement data import record are as follows. The sitting posture dimensions, seat cushion dimensions, and backrest dimensions are extracted from the 3D scan model and then organized into a measurement data set according to the field names. Write the data code and measurement data set into the configuration environment to generate a measurement data import record.
4. The wheelchair customization and optimization method based on virtual simulation as described in claim 1, characterized in that: The steps for performing unit verification and validity validation on the imported measurement data records, and generating a virtual simulation configuration package based on user requirements, are as follows: Based on the imported measurement data, the unit consistency and validity verification of the sitting posture dimensions, seat cushion dimensions, and backrest dimensions are performed and mapped to a set of design boundary parameters. Extract configuration item selections from the configuration environment and map them to a set of design preference parameters; The set of design boundary parameters and the set of design preference parameters are used as geometric constraints and design preference conditions, and integrated into a virtual simulation configuration package.
5. The wheelchair customization and optimization method based on virtual simulation as described in claim 1, characterized in that: The process involves constructing a multi-objective constraint relationship model among design variables, geometric constraints, and design preferences based on a virtual simulation configuration package, and then selecting a set of candidate design parameters under the constraints of the virtual simulation configuration package. The steps are as follows. Based on the hypergraph constraint network, the design variables in the virtual simulation configuration package are defined as parameter nodes, and the geometric constraints and design preference conditions are defined as constraint nodes. The connection relationship between parameter nodes and constraint nodes is established according to the dependency relationship, and a multi-objective constraint relationship model between design variables, geometric constraints and design preferences is constructed. Under the constraints of the multi-objective constraint relationship model, define the allowed value range and value step size, and generate the design parameter domain definition record; Select a set of candidate design parameters that satisfy the constraints of the virtual simulation configuration package from the design parameter domain definition records.
6. The wheelchair customization and optimization method based on virtual simulation as described in claim 1, characterized in that: The process involves evaluating performance response changes based on the candidate design parameter set through parameter perturbation sampling and simulation feedback iteration mechanisms, and generating a parameter sensitivity ranking sequence. The steps are as follows: Based on the candidate design parameter set, records are defined according to the design parameter domain, single-parameter disturbance sampling is performed, and disturbance sampling records are generated. Based on the disturbance sampling records, the virtual simulation configuration package is driven to perform twin simulation inference and perform differential operations to generate a sequence of performance response changes; The performance response change sequence is normalized to determine the parameter sensitivity, and then sorted to form a parameter sensitivity ranking sequence.
7. The wheelchair customization and optimization method based on virtual simulation as described in claim 1, characterized in that: The steps for parametric adjustment and geometric verification based on the parameter sensitivity sorting sequence to generate a geometric consistency evaluation, and combining this with the design performance evaluation to generate a comprehensive evaluation sequence, are as follows: Based on the candidate design parameter set and parameter sensitivity ranking sequence, the 3D scanning model in the configuration environment is parametrically adjusted to generate a wheelchair twin simulation model; The length and angle of the wheelchair twin simulation model are measured, and the geometric relationship is checked for consistency to generate a geometric consistency evaluation. The attribute fields of the wheelchair twin simulation model are calculated to generate corresponding weight index, size matching index and structural stability index, forming a design performance evaluation. Based on geometric consistency evaluation and design performance evaluation, a comprehensive evaluation sequence is generated by merging them according to the evaluation weight rules.
8. The wheelchair customization and optimization method based on virtual simulation as described in claim 1, characterized in that: The steps for generating the wheelchair customization plan confirmation form are as follows: The optimal set of design parameters is selected from the comprehensive evaluation sequence and written into the configuration items of the configuration environment to generate a configuration scheme output package; Associate the configuration solution output package with the data code to generate the corresponding wheelchair customization solution confirmation form.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the wheelchair customization and optimization method based on virtual simulation as described in any one of claims 1 to 8.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the virtual simulation-based wheelchair customization and optimization method as described in any one of claims 1 to 8.