A method and system for motion planning of human simulation models based on inverse dynamics

Abstract

This invention discloses a method and system for motion planning of a human simulation model based on inverse dynamics, belonging to the field of human factors engineering assembly simulation technology. The method includes: structuring and lightweighting a pre-constructed human simulation model; constructing human dynamic constraints based on inverse dynamics; constructing the simulation behavior of the human simulation model based on the human dynamic constraints; and performing motion planning on the human simulation model using an inverse dynamics solver based on the simulation behavior and human dynamic constraints. Structuring enables the human simulation model to easily handle associated skeletal information during simulation, providing a foundation for the inverse dynamics solver; lightweighting improves the processing efficiency of the simulation engine; and using physical dynamics equations to perform dynamic analysis of human posture minimizes the number of tool changes by assembly personnel, increases effective working time, improves planning quality and efficiency, and overcomes the shortcomings of traditional posture prediction and planning processes, such as repetitiveness, prediction complexity, and low efficiency.

Description

Technical Field

[0001] This invention relates to a method and system for motion planning of human body simulation models based on inverse dynamics, belonging to the field of human factors engineering assembly simulation technology. Background Technology

[0002] With the continuous development of society and the economy, we yearn for computers to further change people's lifestyles, and for this reason, we are constantly trying to research computer technology. In recent years, the idea of ​​combining computer technology with other technologies to create simulation technology has been proposed to solve a series of problems encountered in actual industrial manufacturing during human development. In the field of computer vision, 3D reconstruction is a technology with great potential in scientific research, industry, and service industries. 3D reconstruction models have great research and application value in streaming media, building preservation, industrial manufacturing, and other fields. However, regarding the 3D reconstruction and motion planning methods for human movements, due to the rich variety of human postures, how to build simulation models to simulate and plan human movements has become a major challenge.

[0003] Most existing human motion planning methods are based on the estimation of expression parameters and the model itself, without extracting the corresponding physical information, which makes it difficult to accurately and in detail describe the corresponding human motion. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for motion planning of human simulation models based on inverse dynamics, overcoming the shortcomings of traditional posture prediction and planning processes such as repetitiveness, complex prediction, and low efficiency.

[0005] To achieve the above objectives, the present invention employs the following technical solution:

[0006] In a first aspect, the present invention provides a motion planning method for a human simulation model based on inverse dynamics, comprising:

[0007] Structuring and lightweighting of pre-built human simulation models;

[0008] Constructing dynamic constraints for the human body based on inverse dynamics;

[0009] The simulation behavior of constructing a human body simulation model based on the aforementioned human body dynamic constraints;

[0010] Based on the simulated behavior and human dynamic constraints, motion planning is performed on the human simulation model using an inverse dynamics solver.

[0011] In conjunction with the first aspect, the human body simulation model is further obtained through the following method:

[0012] Obtain historical human body parameters and construct a human body simulation model in 3D modeling software based on these parameters.

[0013] In conjunction with the first aspect, further, the structuring of the pre-constructed human simulation model includes:

[0014] The relationships between bones in the human simulation model are stored in a pre-defined data structure to complete the structuring;

[0015] The lightweighting of the pre-built human simulation model includes:

[0016] The human body simulation model is discretized into a triangular mesh; redundant high-precision information and invisible content are removed if the human body simulation model meets the rendering requirements; and holes in the human body simulation model are repaired or deleted.

[0017] In conjunction with the first aspect, the human body dynamic constraint further includes a matching constraint, which is used to constrain a certain part of the human body simulation model to a certain surface of the workpiece to be assembled.

[0018] Compatibility constraints are implemented using the following methods:

[0019] Calculate the angle between the surface normal vector of the hand in the human simulation model and the surface of the workpiece to be assembled, and calculate the distance between the hand in the human simulation model and the surface of the workpiece to be assembled. When the distance is less than a preset threshold and the angle is 90 degrees, the hand and the component to be assembled form a matching constraint.

[0020] In conjunction with the first aspect, the human body dynamic constraints further include contact constraints, which are implemented through the following method: based on the inverse dynamics solver, key parts of the human body and objects are selected, the positions of the key parts of the human body are modified to the positions of the objects, and the position and rotation changes of the skeletal chain in the human body simulation model are calculated through inverse dynamics. A certain part of the human body simulation model is directly constrained to a specific position that needs to be contacted.

[0021] In conjunction with the first aspect, the human body dynamic constraints further include fixed constraints, which are implemented by the following method: based on the inverse dynamics solver, a certain part of the human body simulation model is directly constrained to a preset position or kept in a fixed direction, and the part under fixed constraints will not move with the human body simulation model when it moves.

[0022] In conjunction with the first aspect, the simulated behaviors further include picking and placing, moving forward and backward, and climbing stairs;

[0023] Picking and placing are constructed using the following method: obtaining the initial position of the target object and the initial action of the human simulation model, constructing the human dynamic constraints between the human simulation model and the target object, thereby determining the termination position of the target object and the termination action of the human simulation model, and completing the construction of picking and placing.

[0024] Forward and backward movement are constructed using the following method: the behavior path is constructed using the spline system in the process of building the human simulation model, and the human dynamic constraints of the human simulation model on the spline are created to determine the swing of the hands, walking stride length and speed, thereby completing the construction of forward and backward movement.

[0025] The staircase is constructed using the following method: create a staircase model, determine the length, width, and height of the steps in the staircase model, create a spline trajectory based on the length, width, and height of the steps, construct the constraint relationship between the human simulation model and the spline trajectory, determine the staircase path, and complete the construction of the staircase.

[0026] In conjunction with the first aspect, furthermore, when constructing the picking and placing, a conformity constraint between the human simulation model and the target object is constructed. The initial and final actions of the human simulation model are transitioned through keyframes of the action using an interpolation function. The position and rotation of the bones on the bone chain where the end bone is located during the transition are calculated based on inverse dynamics.

[0027] In conjunction with the first aspect, furthermore, when constructing the staircase, the offset of the feet in the human simulation model is obtained based on the height of the staircase. Combined with the offset of the feet, the positional change of the overall leg skeletal chain is obtained through inverse dynamics, thereby determining the staircase path.

[0028] Secondly, the present invention also provides a motion planning system for a human simulation model based on inverse dynamics, comprising:

[0029] Model normalization module: used to structure and lightweight pre-built human simulation models;

[0030] Human body dynamic constraint construction module: used to construct human body dynamic constraints based on inverse dynamics;

[0031] Simulation Behavior Construction Module: Used to construct the simulation behavior of the human body simulation model based on the aforementioned human body dynamic constraints;

[0032] Motion planning module: Used to perform motion planning on the human simulation model based on the simulated behavior and human dynamic constraints, using an inverse dynamics solver.

[0033] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0034] This invention provides a method and system for motion planning of human simulation models based on inverse dynamics. It structures and lightweights the pre-built human simulation model. Structuring allows the model to easily handle associated skeletal information during simulation, providing a foundation for the inverse dynamics solver. Lightweighting significantly improves the processing efficiency of the simulation engine. By using physical dynamics equations and inverse dynamics to analyze human posture, the method minimizes tool changes by assembly personnel, reducing planning steps and increasing effective working time. This improves planning quality and efficiency, overcoming the shortcomings of traditional posture prediction and planning processes, such as repetitiveness, complex predictions, and low efficiency. Furthermore, using inverse dynamics for planning can greatly enhance the enthusiasm and efficiency of product designers and assembly personnel, thereby improving the competitiveness of enterprises. Attached Figure Description

[0035] Figure 1 This is one of the flowcharts of a human motion planning method based on inverse dynamics provided in an embodiment of the present invention;

[0036] Figure 2 This is the second flowchart of a motion planning method for a human simulation model based on inverse dynamics provided in an embodiment of the present invention;

[0037] Figure 3 This is a flowchart of the compatibility constraint provided in the embodiments of the present invention;

[0038] Figure 4 This is a flowchart illustrating the implementation of the picking and placing simulation behavior provided in this embodiment of the invention;

[0039] Figure 5 This is a flowchart of contact constraints provided in an embodiment of the present invention. Detailed Implementation

[0040] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to illustrate the technical solution of the present invention more clearly, and should not be used to limit the scope of protection of the present invention.

[0041] Example 1

[0042] like Figure 1 As shown in the figure, an embodiment of the present invention provides a motion planning method for a human simulation model based on inverse dynamics, comprising the following steps:

[0043] S1. Structuring and lightweighting of pre-built human simulation models.

[0044] Historical human body parameters are collected, and based on these parameters, a human body simulation model is constructed in 3D modeling software, using the GJB2873-1997 human factors standardization document.

[0045] Then, the human simulation model is standardized, namely, structured and lightweighted. Structured processing uses a pre-defined data structure to store the relationships between bones in the human simulation model. Lightweighting includes discretizing the human simulation model into a triangular mesh, deleting redundant high-precision information and invisible content while ensuring the human simulation model meets rendering requirements, and repairing or deleting holes in the human simulation model.

[0046] Lightweighting a 3D model primarily involves reducing its size and improving its simulation efficiency within the simulation engine, without altering its external features, while maintaining accuracy and overall system performance. This is achieved by removing unnecessary geometric information and decreasing the number of points, lines, and surfaces, all while closely adhering to the original model's characteristics. In industry standards, rendering model data can be made very small, even down to 1 / 20th of its original size. Given that current machine performance cannot be significantly improved, reducing model content will make virtual simulation smoother and more efficient. Directly using high-poly models for rendering and simulation would increase the machine's workload and slow down viewing and operation.

[0047] The specific operation of lightweighting can be described as follows:

[0048] (1) Triangulation of parametric curves and surfaces: The speed of model display and operation and the reduction of computer memory can be achieved by triangulation of surface and edge data. Triangular mesh models can greatly improve efficiency. By discretizing the human simulation model into a triangular mesh model, the appearance of the original model can be approximated in a more accurate way.

[0049] (2) Combine objects of the same material: In order to improve the efficiency of the graphics card and CPU, objects of the same material can be combined. Imported objects with the same material information will display redundant point and surface information. The combined content will effectively reduce the content displayed. Taking the combination of cars of the same material as an example, the number of objects is reduced from tens of thousands to thousands, which effectively improves the overall simulation frame rate.

[0050] (3) Organizing Invisible Content: The virtual model (human simulation model) constructed using high-precision 3D modeling software contains a large number of points, lines, and surfaces to display the high-precision information of the model. However, during the simulation process, this extra high-precision information is not needed. Therefore, we need to organize and delete these meaningless contents before importing them into the scene. Under the premise that the model meets the features required for rendering, we should delete the redundant points, lines, and surfaces. Secondly, we should also organize and delete some invisible content.

[0051] (4) Handling of details such as holes and openings: Holes in the model will also occupy a lot of data information if they are not repaired or deleted. Repairing and deleting them will effectively reduce the number of points, lines and surfaces when there is no need to display them.

[0052] (5) Processing of duplicate model content: Duplicate component models can be created by cloning. The principle of cloning is to obtain the copy of the original object information, rather than simply copying the object to generate redundant point, line and surface information, which can effectively reduce system calls.

[0053] Structured design allows human simulation models to easily process associated skeletal information during simulation, providing a foundation for the inverse dynamics solver, while lightweight design improves the efficiency of the simulation engine.

[0054] S2. Constructing human body dynamic constraints based on inverse dynamics.

[0055] Human body dynamic constraints include engagement constraints, contact constraints, and fixed constraints. Engagement constraints are used to constrain a part of the human body simulation model to a surface of the workpiece to be assembled. Contact constraints are used to directly constrain a part of the human body simulation model to a specific position that needs to be contacted. Fixed constraints are used to constrain a part of the human body simulation model to a specific position or a specific direction.

[0056] like Figure 3 As shown, the coherence constraint is achieved by the following method: calculating the angle between the surface normal vector of the hand in the human simulation model and the surface of the workpiece to be assembled, calculating the distance between the hand in the human simulation model and the surface of the workpiece to be assembled, and when the distance is less than a preset threshold and the angle is 90 degrees, the hand and the assembly to be assembled form a coherence constraint; the arm state after the coherence constraint is formed is solved by inverse dynamics, and the inverse dynamics solver will solve the position and rotation of the entire skeletal chain where the end bone is located by the end bone position.

[0057] like Figure 5 As shown, contact constraints are achieved through the following method: Based on the inverse dynamics solver, key parts of the human body are first selected, then objects are selected. After selection, the position of the key parts of the human body is modified to the position of the objects. Then, the skeletal chain is solved according to inverse dynamics, so that a certain part of the human simulation model is directly constrained to a specific position that needs to be contacted; and the position and rotation of the bones in the skeletal chain will change based on inverse dynamics.

[0058] Fixed constraints are achieved through the following method: Based on the inverse dynamics solver, a certain part of the human simulation model is directly constrained to a preset position or kept in a fixed direction, and the part under fixed constraints will not move with the human simulation model when it moves.

[0059] S3. Simulation behavior of constructing a human body simulation model based on the aforementioned human body dynamic constraints.

[0060] Simulated behaviors include picking and placing, moving forward and backward (i.e.) Figure 2 Walking), climbing stairs, such as Figure 2 As shown, it can also include bending over.

[0061] Picking and placing are constructed using the following method: obtaining the target object ( Figure 4 The initial position of the object to be picked up or placed and the initial movement of the human simulation model are used to construct the human dynamic constraints between the human simulation model and the target object (i.e., Figure 4 The system creates a nominal constraint relationship between the hand and the object, thereby determining the termination position of the target object and the termination action of the human simulation model, and completing the construction of picking and placing.

[0062] When constructing the picking and placing process, the human simulation model and the target object are constrained to match. The initial and final actions of the human simulation model are transitioned through keyframes using interpolation functions. The position and rotation of the bones on the bone chain where the end bone is located during the transition are calculated based on inverse dynamics.

[0063] Forward and backward movement are constructed using the following method: the behavior path is constructed using the spline system during the human simulation model construction process, and the human dynamic constraints of the human simulation model on the spline are created to determine the swing of the hands, walking stride length and speed, thus completing the construction of forward and backward movement.

[0064] The staircase is constructed using the following method: create a staircase model, determine the length, width, and height of the steps in the staircase model, create a spline trajectory based on the length, width, and height of the steps, construct the constraint relationship between the human simulation model and the spline trajectory, determine the staircase path, and complete the construction of the staircase.

[0065] When constructing the staircase, the offset of the feet in the human simulation model is obtained based on the height of the staircase. Combined with the offset of the feet, the positional change of the overall leg skeletal chain is obtained through inverse dynamics, thereby determining the staircase path and generating the standard skeletal movements for climbing the staircase.

[0066] S4. Based on the simulated behavior and human dynamic constraints, perform motion planning on the human simulation model using an inverse dynamics solver.

[0067] By constructing the human body dynamic constraints in step S2, and based on the inverse dynamics solver, the planning and simulation of the human body simulation model's movements are realized, and the movement planning of the human body simulation model is generated based on the simulation behavior in step S3.

[0068] The human motion planning method based on inverse kinematics (IK) effectively overcomes the limitations of inaccurate posture prediction in traditional human simulation models, reduces the cost of posture prediction, and improves the prediction and planning efficiency of staff. Simultaneously, by incorporating planning activities during the actual prediction process, the quality and efficiency of prediction planning can be improved. The inverse kinematics-based human motion planning method effectively solves the problems existing in traditional human simulation model posture prediction, allowing users to obtain information from the real-world posture environment while observing the posture model. During system operation, only the posture prediction model needs to be rendered, which can greatly save the workload of entity modeling and the resource consumption of system operation. Furthermore, while improving prediction perception, the inverse kinematics-based human motion planning method can significantly increase the enthusiasm and efficiency of product designers and prediction personnel, thereby enhancing the competitiveness of enterprises.

[0069] This invention provides a motion planning method for human simulation models based on inverse dynamics. Guided by human factors engineering theory, virtual simulation engine technology, and assembly simulation theory, this method is designed with inverse dynamics as its focus. It uses physical dynamic equations to analyze the dynamics of human posture. This method can minimize the number of tool changes and predictions required by assembly personnel during the assembly process, thereby increasing effective working time and improving the quality and efficiency of posture planning. It overcomes the shortcomings of traditional technologies, such as the repetitive planning process, complex prediction, and low efficiency.

[0070] Example 2

[0071] This invention provides a motion planning method for a human simulation model based on inverse dynamics, comprising the following steps:

[0072] S1. Constructing human dynamic constraints based on inverse kinematics, including coincidence constraints, contact constraints, and fixed constraints:

[0073] The three methods for implementing human body dynamic constraints are the same as those described in Example 1.

[0074] Human body dynamic constraints can be transformed into relationships between points and lines in a coordinate system, where α0 and l0 represent the critical values ​​for angle and distance recognition, respectively, and α and l represent the angle and distance, respectively. The constraint rules are as follows:

[0075] (1) When the constraint is identified as a surface:

[0076] When l-l0≤0 and |α|-α0≤0, the system recognizes the alignment information;

[0077] When l-l0≤0 and |π-α|-α0≤0, the system recognizes the fitting information;

[0078] When l-lx≤0 and |π-α|-α0≤0, the system recognizes the offset l0 information.

[0079] (2) When the constraint is identified as a straight line:

[0080] When l - l0 ≤ 0 and |cosα| - |cosα0| ≤ 0, alignment information is detected;

[0081] When l-l0≤0 and |cosα|-|cosα0|≥0, the offset l0 information is detected.

[0082] (3) When the constraint is identified as a line and a surface:

[0083] When l - l0 ≤ 0, if Then, coincident constraints are identified;

[0084] When |π-α|-α0≤0 or |α|-α0≤0, vertical information is detected;

[0085] When l-l0≤L0 and |cosα|-|cosα0|≥0, the offset L0 information is detected;

[0086] When l-l0≤0, if |π-α|-α0≤θ or |α|-α0≤θ, then θ information is identified.

[0087] (4) When the constraint is identified as a cylindrical surface:

[0088] When l-l0≤0 and |cosα|-|cosα0|≥0 and |r1-r2|≤r0, the inserted information is recognized.

[0089] S2, Human Body Transport Simulation UI (User Interface) Construction:

[0090] Human body handling simulation can name the activity. During handling, first set the reaching action. The reaching action needs to set an initial position, and then determine the end position of the reaching. The determination of the reaching position is based on the human body dynamic constraints in step S1.

[0091] Set the reaching animation trajectory based on the initial and ending positions as key points. When the hand and object are in the right position, create constraints between the hand and the object, and then construct the lifting motion. The lifting motion only requires setting the ending position of the object, and the hand will make the lifting action based on the inverse dynamics solver according to the position of the object.

[0092] S3, Construction of the reaching motion:

[0093] The reaching motion is also based on IK calculation. In the process, we need to first set up the IK solver for the arm and set the IK end bones to the left and right hand bones. This way, the position of the entire skeletal chain can be changed at runtime by adjusting the position of the hand bones.

[0094] Next, the initial motion of the reaching hand is set. Whether it is frame animation or interpolation function, key points are needed to realize the position transition. Therefore, the initial position of the reaching hand motion is located at the beginning keyframe of the motion, where two variables will be used to store the initial positions of the left and right hands.

[0095] Then, locate the end position of the reaching motion. The two hands may touch different positions on the object, so they are located separately. Taking the left hand as an example, when you want to locate the end position of the left hand, turn on ray detection. Ray detection will select the position to contact the object. Once determined, the position of the left hand will be constrained to the specified position on the object.

[0096] Next is the animation of the reaching hand action. The Timeline function of the 3D simulation engine does not support custom time curves at runtime, so this article uses an interpolation function to complete the animation effect at two key action positions.

[0097] S4. Select objects in the scene to create transport constraints:

[0098] Once the reaching action is created, click the "Create Hand and Object Constraints" dialog box in the human body handling simulation UI to select objects in the scene for constraints. All objects in the scene that can interact with the operator are recorded in the array ArrayMovableItems. We only need to obtain the corresponding array index position from the dialog box, assign it a value, and use world space coordinates as spatial rules to define the parent-child relationship between the object to be constrained and the human body, as well as the constraint relationship with the human hand.

[0099] S5. Picking Action Construction: Similar to the reaching action construction, the picking action is also based on two key positions, the initial position and the ending position, to create keyframes, and then use an interpolation function to make a smooth transition. The difference is that the constraint relationship between the hand and the object to be picked has been defined in the picking action. Therefore, when defining the ending position of the picking action, the change of the initial position and the ending position of the object to be picked will affect the change of the hand position, generating the IK picking action animation.

[0100] Through the steps S1 to S5 described above, by classifying and weighting, we not only consider the impact of each evaluation indicator on planning efficiency, but also the impact of the interrelationships between the indicators.

[0101] This invention utilizes the method to obtain the sequence of actions at different positions. Addressing the shortcomings of current virtual assembly systems, which are entirely virtual and cannot interact with the real environment in real time during the planning process, this invention designs the overall framework of the system based on the functional requirements of virtual assembly and the characteristics of augmented reality technology. Virtual simulation is selected as the development tool according to the scheme, and relevant modeling and optimization are completed. The model is then imported into the development platform. This planning method reduces the number of planning iterations, thereby increasing effective working time and improving planning quality and efficiency.

[0102] Example 3

[0103] This invention provides a motion planning system for a human simulation model based on inverse dynamics, comprising:

[0104] Model normalization module: used to structure and lightweight pre-built human simulation models;

[0105] Human body dynamic constraint construction module: used to construct human body dynamic constraints based on inverse dynamics;

[0106] Simulation Behavior Construction Module: Used to construct the simulation behavior of the human body simulation model based on the aforementioned human body dynamic constraints;

[0107] Motion planning module: Used to perform motion planning on the human simulation model based on the simulated behavior and human dynamic constraints, using an inverse dynamics solver.

[0108] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0109] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0110] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0111] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0112] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for motion planning of a human simulation model based on inverse dynamics, characterized in that, The application relates to a method for constructing a human simulation model, comprising the following steps: Structuring and lightening a pre-constructed human simulation model; Constructing human dynamic constraints based on inverse dynamics; Constructing simulation behaviors of the human simulation model based on the human dynamic constraints; Planning actions of the human simulation model through an inverse dynamics solver according to the simulation behaviors and the human dynamic constraints; The human dynamic constraints comprise a mutual constraint, which is used for constraining a part of the human simulation model to a surface of an assembly workpiece; The mutual constraint is realized by the following method: Calculating an angle between a surface normal vector of a hand in the human simulation model and the surface in the assembly workpiece, and calculating a distance between the hand in the human simulation model and the surface in the assembly workpiece; when the distance is less than a preset threshold value and the angle is 90 degrees, the hand and the assembly workpiece form the mutual constraint; The human dynamic constraints comprise a contact constraint, which is realized by the following method: Based on the inverse dynamics solver, selecting a human key part and an object, modifying a position of the human key part to the object, and inversely dynamically solving position and rotation changes of a bone chain in the human simulation model, so that a part of the human simulation model is directly constrained to a specific position needing contact; The human dynamic constraints comprise a fixed constraint, which is realized by the following method: Based on the inverse dynamics solver, a part of the human simulation model is directly constrained to a preset position or kept in a fixed direction, and the part in the fixed constraint does not move with the human simulation model during movement of the human simulation model; The simulation behaviors comprise picking and placing, advancing and retreating, and climbing stairs; The picking and placing is constructed by the following method:

2. The method of claim 1, wherein the method is based on inverse dynamics. Obtaining initial positions of target objects and initial actions of the human simulation model, constructing human dynamic constraints of the human simulation model and the target objects, so as to determine terminal positions of the target objects and terminal actions of the human simulation model, and completing construction of the picking and placing; The advancing and retreating is constructed by the following method:

3. The method of claim 1, wherein the method further comprises: Constructing a behavior path through a spline line system in a construction process of the human simulation model, and creating human dynamic constraints of the human simulation model on the spline line, so as to determine swinging conditions of both hands, walking step length and speed, and complete construction of the advancing and retreating; The climbing stairs is constructed by the following method: Creating a stair model, determining length, width and height of a stair in the stair model, creating a spline line track according to the length, width and height of the stair, constructing a constraint relationship between the human simulation model and the spline line track, determining a climbing stair path, and completing construction of the climbing stairs. The human simulation model is obtained by the following method: Obtaining historical human parameters, and constructing the human simulation model in a three-dimensional modeling software according to the historical human parameters. The structuring of the pre-constructed human simulation model comprises the following steps: Saving relationships between bones in the human simulation model in a preset data structure, and completing structuring; The lightening of the pre-constructed human simulation model comprises the following steps: Discretizing the human simulation model into a triangular mesh form, deleting redundant high-precision information and invisible contents under a condition that the human simulation model meets rendering requirements, and repairing or deleting holes in the human simulation model.

4. The method of claim 1, wherein, In the construction of pick and place, the body simulation model and the target object are combined with the constraint, the initial action and the final action of the body simulation model are transitioned through the interpolation function, and the position and rotation of the bones in the bone chain where the end bone is located during the transition are calculated based on the inverse dynamics.

5. The method of claim 1, wherein, In the construction of climbing stairs, the offset of the foot of the body simulation model is obtained according to the height of the stairs, the position change of the overall leg bone chain is obtained through the inverse dynamics combined with the offset of the foot, and the climbing stairs path is determined.

6. A motion planning system for a human simulation model based on inverse dynamics, characterized in that, It comprises: A model normalization processing module is used to structure and lighten the pre-constructed body simulation model. A human dynamic constraint construction module is used to construct human dynamic constraints based on inverse dynamics. A simulation behavior construction module is used to construct the simulation behavior of the body simulation model based on the human dynamic constraints. An action planning module is used to plan the action of the body simulation model through the inverse dynamics solver based on the simulation behavior and human dynamic constraints. The human dynamic constraints include the combination constraint, which is used to constrain a part of the body simulation model to a surface of the workpiece to be assembled. The combination constraint is realized by the following method: Calculate the angle between the surface normal vector of the hand in the body simulation model and the surface in the workpiece to be assembled, and calculate the distance between the hand in the body simulation model and the surface in the workpiece to be assembled. When the distance is less than the preset threshold and the angle is 90 degrees, the hand and the workpiece form a combination constraint. The human dynamic constraints include the contact constraint, which is realized by the following method: based on the inverse dynamics solver, select the key parts of the human body and the object, modify the position of the key parts of the human body to the object, and calculate the position and rotation change of the bone chain in the body simulation model through the inverse dynamics solver. A part of the body simulation model is directly constrained to a specific contact position. The human dynamic constraints include the fixed constraint, which is realized by the following method: based on the inverse dynamics solver, a part of the body simulation model is directly constrained to a preset position or a fixed direction, and the part under the fixed constraint will not move with the body simulation model during the movement of the body simulation model. The simulation behavior includes pick and place, forward and backward, and climbing stairs. Pick and place is constructed by the following method: obtaining the initial position of the target object and the initial action of the body simulation model, constructing the human dynamic constraints of the body simulation model and the target object, determining the final position of the target object and the final action of the body simulation model, and completing the construction of pick and place. Forward and backward is constructed by the following method: constructing the behavior path through the spline line system in the body simulation model construction process, creating the human dynamic constraints of the body simulation model on the spline line, and determining the swinging situation of the hands, the walking step and the speed, to complete the construction of forward and backward. Climbing stairs is constructed by the following method: creating a stair model, determining the length, width and height of the stairs in the stair model, creating a spline trajectory according to the length, width and height of the stairs, constructing the constraint relationship between the body simulation model and the spline trajectory, determining the climbing stairs path, and completing the construction of climbing stairs.

Images