Femur-like fiber composite locator and design method

Abstract

The application discloses a femur fiber composite material positioner and a design method, and belongs to the technical field of positioners.The femur fiber composite material positioner is compared with a traditional metal positioner under the condition of equal rigidity, and the femur fiber composite material positioner is lightened by about 40%, thereby effectively reducing the influence of flow-receiving mass.Under the condition of equal mass, the rigidity and corrosion resistance of the femur fiber composite material positioner are improved.The design method of the femur fiber composite material positioner is based on the combination of ABAQUS simulation and BP neural network, thereby greatly reducing the time cost, improving the adaptability and service life of the positioner, and reducing the maintenance cost.In the use process, with the further expansion of the database, the time cost is greatly reduced, the adaptability and service life of the positioner are improved, the maintenance cost is reduced, and the mapping relationship accuracy of the structural parameters and deformation is improved.

Description

Technical Field

[0001] This invention discloses a femoral fiber composite material positioner and its design method, belonging to the field of positioner technology. Background Technology

[0002] With the development of rail transit, the requirements for lightweight, efficient, and high-performance materials are becoming increasingly stringent. Carbon fiber composites possess excellent properties such as lightweight, high strength, high stiffness, excellent vibration damping, fatigue resistance, and corrosion resistance. Advanced composite materials, with high-performance carbon fiber composites as a typical example, serve as structural, functional, or integrated structural and functional materials. They are not only irreplaceable in the construction of national defense strategic weapons but will also play a crucial role in green energy development, energy-saving technology advancement, and the promotion of energy diversification.

[0003] Traditional positioners are generally made of aluminum alloy, which has a large overall weight, affecting the quality of current collection. They also have poor fatigue resistance and corrosion resistance, high maintenance costs, and it is difficult to determine the optimal size and specifications of the positioners required for each route.

[0004] Currently, a small number of positioners on the market are made of composite materials. Their internal support structure is filled only with PMI foam, resulting in an unreasonable distribution of pores and an open-cell structure. The number of external carbon fiber layers is also difficult to determine, necessitating further optimization of the positioner's weight and structure. Therefore, how to rationally apply composite materials to railway positioners is a problem that needs to be solved in the current engineering field. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention proposes a femoral fiber composite material positioner and its design method, which solves the problems of the current positioners made of composite materials having an internal support structure filled only with PMI foam, resulting in an unreasonable distribution of internal pores and an open structure, and making it difficult to determine the number of external carbon fiber layers.

[0006] The technical solution of the present invention is as follows:

[0007] According to a first aspect of the present invention, a femoral fibrous composite material positioner is provided, comprising a positioning tube, wherein the positioning tube is connected to a positioning connection point and a positioning clamp at both ends respectively, the positioning tube comprising a femoral shell, wherein a femoral support structure and a filling support structure are provided within the cavity of the femoral shell, and the bone porosity of the femoral support structure gradually decreases from the positioning clamp to the positioning connection point.

[0008] Preferably, the femoral support structure includes a compact bone pore unit region and a cancellous bone pore unit region. The compact bone pore unit region is adjacent to the femoral shell. The pore diameter of the compact bone pore unit region is smaller than that of the cancellous bone pore unit region. Both the pores in the compact bone pore unit region and the pores in the cancellous bone pore unit region are closed pores.

[0009] Preferably, the femoral shell is formed by heating and pressurizing a fiber resin layer, and the filling support structure is formed by secondary processing and curing of carbon fiber bundles and resin.

[0010] According to a second aspect of the present invention, a design method for a femoral fiber composite material positioner is provided, applied to the femoral fiber composite material positioner described in the first aspect, comprising:

[0011] Obtain the design length and maximum design working load of the positioning tube, and determine the value range of the design parameters of the femoral fiber composite positioning tube based on the positioning tube length and maximum working load, respectively.

[0012] The value range of the design parameters of the imitation femoral fiber composite positioning tube is normalized to obtain the value range of the design parameters of the imitation femoral fiber composite positioning tube after preprocessing.

[0013] The value ranges of the pre-processed femoral fiber composite positioning tube design parameters are arranged and combined to obtain multiple sets of femoral fiber composite positioning tube design parameters. The multiple sets of femoral fiber composite positioning tube design parameters are then fed into the trained optimal BP neural network model.

[0014] The optimal design parameter set for the femoral fiber composite positioning tube is obtained using the optimal BP neural network model. The positioning tube model is then obtained by modeling based on the optimal design parameter set for the femoral fiber composite positioning tube and the positioning tube length.

[0015] Based on the positioning tube model, design the positioning connection point model and the positioning clamp model respectively, and assemble the positioning tube model, the positioning connection point model and the positioning clamp model to obtain the locator model.

[0016] Preferably, before obtaining the design length of the positioning tube and the maximum design working load, the method further includes:

[0017] Obtain existing datasets of multiple sets of femoral fiber composite positioning tube design parameters and corresponding multiple positioning tube lengths;

[0018] A finite element model of a femoral fiber composite positioning tube is established. Based on the existing dataset of multiple sets of design parameters of the femoral fiber composite positioning tube and the corresponding multiple positioning tube lengths, a database of design parameters and deformation of the femoral fiber composite positioning tube is constructed by ABAQUS simulation using Python secondary development.

[0019] The design parameters and deformation database of the femoral fiber composite positioning tube are preprocessed and training and testing sets are constructed.

[0020] A BP neural network model for deformation calculation is constructed, and the optimal BP neural network model is obtained by training and testing the BP neural network model using the training set and test set.

[0021] Preferably, the design parameters of the femoral fiber composite positioning tube include: femoral shell radius, number of femoral shell carbon fiber layers, proportion of cancellous bone pore unit area, and bone pore reduction coefficient. The BP neural network model includes: an input layer neural network, a hidden layer neural network, and an output layer neural network, wherein:

[0022] The input layer neural network includes: a femoral shell radius input neural unit, a femoral shell carbon fiber layer number input neural unit, a cancellous bone pore unit area proportion input neural unit, and a bone pore reduction coefficient input neural unit. The hidden layer neural network includes: 10 hidden neural units and a hidden layer bias. The output layer neural network includes: a deformation output neural unit and an output layer bias.

[0023] Preferably, the step of training and testing the BP neural network model using the training set and test set to obtain the optimal BP neural network model includes:

[0024] The design parameters and deformation of the femoral fiber composite positioning tube in the training set are weighted and summed with the connection weights of each hidden neural unit, and the values ​​of each hidden neural unit are obtained through the activation function of the input layer. The values ​​of each hidden neural unit are weighted and summed with the connection weights of the corresponding output neural units, and the first output deformation is obtained through the activation function of the output layer.

[0025] The first error value is obtained by subtracting the first output deformation from the deformation in the training set. The first error value is then backpropagated through a BP neural network model to update the connection weights of each hidden neural unit and the connection weights of the output neural unit.

[0026] The updated connection weights of each hidden neural unit and the connection weights of the output neural unit are forward propagated again to obtain the second output deformation. The difference between the second output deformation and the deformation in the training set is used to obtain the second error value. It is then determined whether the second error value is greater than the target error value.

[0027] Yes, the second error value is backpropagated through the BP neural network model again to update the connection weights of each hidden neural unit and the connection weights of the output neural unit. The process is iterated until the error value is less than or equal to the target error value, and then the next step is executed.

[0028] No, proceed to the next step;

[0029] Save the connection weights of each hidden layer neural unit and the connection weights of the output neural unit to complete the training of the neural network and obtain the optimal BP neural network model. Then, test the performance of the optimal BP neural network model through a test set.

[0030] Preferably, the activation function of the input layer is the logsig function, and the activation function of the output layer is the sigmoid function.

[0031] Preferably, the step of updating the connection weights of each hidden neural unit and the connection weights of the output neural unit by backpropagating the first error value through a BP neural network model includes:

[0032] The value obtained by the first error value LM learning algorithm is multiplied by the learning rate to obtain the adjustment value of the connection weight of each hidden neuron and the adjustment amount of the connection weight of the output neuron, respectively.

[0033] The adjusted connection weights of each hidden neural unit and the adjusted connection weights of the output neural unit are added to the current connection weights of the hidden neural unit and the output neural unit, respectively, to obtain the updated connection weights of the hidden neural unit and the output neural unit.

[0034] Preferably, the step of obtaining the optimal design parameter set for the femoral fiber composite positioning tube using the optimal BP neural network model includes:

[0035] Multiple deformation values ​​are obtained using the optimal BP neural network model, and the minimum deformation data is obtained by comparing the multiple deformation values.

[0036] Based on the minimum deformation data, multiple sets of femoral fiber composite positioning tube design parameter sets are traversed to obtain the corresponding femoral fiber composite positioning tube design parameter set, and the corresponding femoral fiber composite positioning tube design parameter set is the optimal femoral fiber composite positioning tube design parameter set.

[0037] The beneficial effects of this invention are as follows:

[0038] This invention provides a femoral fiber composite material positioner and its design method. The femoral fiber composite material positioning tube, under the same stiffness conditions, is approximately 40% lighter than traditional metal positioners, effectively reducing the impact of flow mass. Under the same mass conditions, it improves stiffness and corrosion resistance. The design method, based on a combination of ABAQUS simulation and BP neural network, significantly reduces time costs, improves the positioner's adaptability and service life, and lowers maintenance costs. Furthermore, with the further expansion of the database during use, the accuracy of the mapping relationship between various structural parameters and deformation amounts is improved.

[0039] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of a femoral fiber composite material positioner according to an exemplary embodiment.

[0041] Figure 2 This is a schematic diagram of a positioning tube with a cancellous bone pore unit area accounting for 80% in a femoral fiber composite material positioner according to an exemplary embodiment.

[0042] Figure 3 This is a schematic diagram of a positioning tube with a cancellous bone pore unit area accounting for 80% in a femoral fiber composite material positioner according to an exemplary embodiment.

[0043] Figure 4 This is a longitudinal section along A-A' showing the position tube cancellous bone pore unit area accounting for 80% of the total area in an exemplary femoral fiber composite locator.

[0044] Figure 5 This is a schematic diagram of a positioning tube with a cancellous bone pore unit area accounting for 50% in a femoral fiber composite material positioner according to an exemplary embodiment.

[0045] Figure 6 This is a schematic diagram of a positioning tube with a cancellous bone pore unit area accounting for 50% in a femoral fiber composite material positioner according to an exemplary embodiment.

[0046] Figure 7 This is a longitudinal section along B-B' of a positioning tube in a femoral fiber composite material positioning device, showing 50% of the cancellous bone pore unit area.

[0047] Figure 8This is a schematic diagram of a positioning tube with a cancellous bone pore unit area accounting for 40% in an exemplary femoral fiber composite material positioner.

[0048] Figure 9 This is a schematic diagram of a positioning tube with a cancellous bone pore unit area accounting for 40% in an exemplary femoral fiber composite material positioner.

[0049] Figure 10 This is a longitudinal section along C-C' of a positioning tube in a femoral fiber composite material positioning device, showing a 40% proportion of the cancellous bone pore unit area.

[0050] Figure 11 This is a flowchart illustrating a design method for a femoral fiber composite positioner according to an exemplary embodiment.

[0051] Figure 12 This is a topology diagram of a BP neural network in a design method for a femoral fiber composite material positioner, according to an exemplary embodiment. Detailed Implementation

[0052] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] In the description of this invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0054] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0055] Example 1

[0056] Figure 1This is an exemplary embodiment of a femoral fiber composite material positioner, comprising a positioning tube 1, which is connected at both ends to a positioning connection point 2 and a positioning clamp 3, as shown below. Figure 2-4 As shown, the positioning tube 1 includes a femoral shell, and a femoral support structure and a filling support structure are provided inside the femoral shell cavity. The porosity of the bone in the femoral support structure gradually decreases from the positioning clamp 3 to the positioning connection point 2. The porosity is smallest at the positioning connection point 2, and the torque it can withstand is the largest.

[0057] The simulated femoral support structure comprises a compact bone pore unit region and a cancellous bone pore unit region. The compact bone pore unit region has small pore sizes and thick pore walls; the cancellous bone pore unit region has large pore sizes and thin pore walls. The compact bone pore unit region, with its high density and dense structure, provides excellent support and stability, capable of withstanding the weight of the tube itself and external forces. The cancellous bone pore unit region effectively absorbs and disperses external impacts and pressures, thereby reducing the impact on the tube. The cancellous bone pore unit region constitutes a relatively small proportion of the tube. The compact bone pore unit region is adjacent to the femoral shell. The pore diameter of the compact bone pore unit region is smaller than that of the cancellous bone pore unit region. Both the compact and cancellous bone pore unit regions have closed pores, making it difficult for moisture and water to penetrate, reducing maintenance and inspection costs and extending the service life of the support structure. The femoral shell is formed by heating and pressurizing a fiber resin layer, while the filling support structure is formed by secondary processing and curing of carbon fiber bundles and resin. The varying proportions of compact and cancellous bone porous unit areas and the bone porosity reduction coefficient θ in the femoral support structure affect the mass and maximum deformation that the positioning tube can withstand. With a fixed bone porosity reduction coefficient, increasing the proportion of the compact bone porous unit area in the femoral support structure increases the mass of the positioning tube and its maximum deformation capacity; conversely, increasing the proportion of the cancellous bone porous unit area decreases the mass and maximum deformation capacity. At a fixed proportion of the compact and cancellous bone porous unit areas, an increase in the bone porosity reduction coefficient θ leads to a larger change in the density of the positioning tube body, resulting in a smaller maximum deformation capacity but a smaller overall mass; conversely, a smaller change in the density of the positioning tube body increases the maximum deformation capacity but an overall mass. Determining the optimal proportion of the compact and cancellous bone porous unit areas and the optimal bone porosity reduction coefficient θ under specified operating conditions is crucial for lightweighting the positioning device, extending its service life, and reducing costs. Figure 2-4 This is a positioning tube that occupies 80% of the cancellous bone pore unit area. For example... Figure 5-7 A positioning tube that occupies 50% of the cancellous bone pore unit area, such as... Figure 8-10 The positioning tube accounts for 40% of the cancellous bone perforation unit area.

[0058] Example 2

[0059] Figure 11 This is a design method for a femoral fiber composite positioner according to an exemplary embodiment, comprising:

[0060] Step 101: Obtain the design length and maximum design working load of the positioning tube. Determine the value range of the design parameters of the femoral fiber composite positioning tube based on the positioning tube length and maximum working load. The specific steps are as follows:

[0061] Before obtaining the design length of the positioning tube and the maximum design working load, the following steps are also included:

[0062] Using the Monte Carlo sampling method, we obtained an existing dataset of multiple sets of femoral fiber composite positioning tube design parameters and corresponding multiple positioning tube lengths. The femoral fiber composite positioning tube design parameters include: femoral shell radius, number of femoral shell carbon fiber layers, proportion of cancellous bone pore unit area, and bone pore reduction coefficient.

[0063] A finite element model of a femoral fiber composite positioning tube was established using UG. Based on existing datasets of multiple sets of femoral fiber composite positioning tube design parameters and corresponding multiple positioning tube lengths, a database of femoral fiber composite positioning tube design parameters and deformation was constructed using Python secondary development ABAQUS simulation.

[0064] The design parameters and deformation data of the femoral fiber composite positioning tube were preprocessed, and training and testing sets were constructed. A BP neural network model for deformation calculation was constructed, and the optimal BP neural network model was obtained by training and testing the training and testing sets.

[0065] The above BP neural network model, such as Figure 12 As shown, it includes: an input layer neural network, a hidden layer neural network, and an output layer neural network. The input layer neural network includes: a femoral shell radius R input neural unit, a femoral shell carbon fiber layer number A input neural unit, a cancellous bone pore unit area ratio X input neural unit, and a bone pore reduction coefficient θ input neural unit. The hidden layer neural network includes: 10 hidden neural units and a hidden layer bias bj1. The 10 hidden neural units are represented by Zj1, Zj2, Zj3, ..., Zj10. The output layer neural network includes: a deformation output neural unit Y and an output layer bias bj2.

[0066] The above-mentioned training and testing of the BP neural network model using the training and testing sets to obtain the optimal BP neural network model includes:

[0067] The design parameters and deformation of the femoral fiber composite positioning tube in the training set are weighted and summed with the connection weights of each hidden neural unit. The values ​​of each hidden neural unit are obtained through the activation function of the input layer. The values ​​of each hidden neural unit are weighted and summed with the connection weights of the corresponding output neural units. The first output deformation is obtained through the activation function of the output layer. The preferred activation function for the input layer is the logsig function, and the preferred activation function for the output layer is the sigmoid function with an introduced slope factor λ. During backpropagation, there is a flat region on the error surface, indicating that the neuron output has entered the saturation region of the activation function. The slope factor λ is introduced to make it exit the saturation region of the activation function. The initial values ​​of the connection weights of the hidden neural units and the connection weights of the output neural units in the Bone-BP neural network are generated by random numbers, with the initial values ​​ranging from (0, 1).

[0068] The first error value is obtained by subtracting the first output deformation from the deformation in the training set. This first error value is then used for backpropagation through a BP neural network model to update the connection weights of each hidden neuron and the output neuron. The backpropagation uses the LM learning algorithm, which is a combination of gradient descent and Gauss-Newton's method. It utilizes approximate second-order derivative information, effectively addressing redundant parameter issues and significantly reducing the chance of the algebraic function getting trapped in local minima. The specific steps are as follows:

[0069] The value obtained by the first error value LM learning algorithm is multiplied by the learning rate η to obtain the adjustment value of the connection weight of each hidden neuron and the adjustment amount of the connection weight of the output neuron, respectively.

[0070] The adjusted connection weights of each hidden neural unit and the adjusted connection weights of the output neural unit are added to the current connection weights of the hidden neural unit and the output neural unit, respectively, to obtain the updated connection weights of the hidden neural unit and the output neural unit.

[0071] The learning rate η mentioned above uses an adaptive method. The initial value of the learning rate η is 1. If the current iteration error is greater than the previous iteration error learning rate η, the value of the learning rate η is decreased, and the decrease range is (0, 1). If the current iteration error is less than the previous error or the current learning rate η is less than the set minimum value, the learning rate η is increased, and the increase value is greater than 1. By optimizing the learning rate algorithm, the iteration speed can be increased, and the training of the BP neural network can be optimized. For example, if the learning rate is initially 1, after one decay, the learning rate becomes 0.5. If the gain is set to 5, the learning rate becomes 2.5.

[0072] The updated connection weights of each hidden neural unit and the connection weights of the output neural unit are forward propagated again to obtain the second output deformation. The difference between the second output deformation and the deformation in the training set is used to obtain the second error value. It is then determined whether the second error value is greater than the target error value.

[0073] Yes, the second error value is backpropagated through the BP neural network model again to update the connection weights of each hidden neural unit and the connection weights of the output neural unit. The process is iterated until the error value is less than or equal to the target error value, and then the next step is executed.

[0074] No, proceed to the next step;

[0075] Save the connection weights of each hidden layer neural unit and the connection weights of the output neural unit to complete the training of the neural network and obtain the optimal BP neural network model. Then, test the performance of the optimal BP neural network model through a test set.

[0076] Based on the above-mentioned optimal BP neural network model and the database of femoral fiber composite positioning tube design parameters and deformation, the following describes the range of values ​​for determining the design parameters of the imitation femoral fiber composite positioning tube.

[0077] Based on the design length and maximum design working load of the positioning tube, it is determined that the positioning tube is only subjected to unidirectional tensile load during operation. The strength design must satisfy σmax≤[σ]. The stress calculation formula for the positioning tube is as follows:

[0078] σ=F / S (1)

[0079] S=πR 2 (2)

[0080] σ max =σ / n (3)

[0081] Where: σ is the tensile strength, F is the working load, S is the area, R is the radius of the femoral shell, and n is the safety factor, preferably 10. max This represents the maximum strength.

[0082] Based on the tensile strength of the femoral fiber composite locator, the working stress should not exceed 65% of its tensile strength. The tensile strength σ of the femoral fiber composite locator... w Using 410 MPa, based on the above formulas (1)-(3), the range of values ​​for R is R≥R min =6.0; Select the value range of R as [6,8], select the value range of A as [6,8] according to the positioner structure parameter - deformation database, select the value range of X as [30%, 50%], and select the value range of θ as [1,2].

[0083] Step 102: Normalize the value range of the design parameters of the imitation femoral fiber composite positioning tube to obtain the value range of the design parameters of the imitation femoral fiber composite positioning tube after preprocessing.

[0084] Step 103: Arrange and combine the value range of the preprocessed femoral fiber composite positioning tube design parameters to obtain multiple sets of femoral fiber composite positioning tube design parameters, and input the multiple sets of femoral fiber composite positioning tube design parameters into the trained optimal BP neural network model.

[0085] Step 104: Use the optimal BP neural network model to obtain the optimal design parameter set for the femoral fiber composite positioning tube, and model the positioning tube according to the optimal design parameter set for the femoral fiber composite positioning tube and the positioning tube length to obtain the positioning tube model.

[0086] The above-mentioned optimal BP neural network model is used to obtain the optimal design parameter set for the femoral fiber composite positioning tube. The specific steps are as follows:

[0087] Multiple deformation values ​​are obtained using the optimal BP neural network model, and the minimum deformation data is obtained by comparing the multiple deformation values.

[0088] Based on the minimum deformation data, multiple sets of femoral fiber composite positioning tube design parameter sets are traversed to obtain the corresponding femoral fiber composite positioning tube design parameter set, which is the optimal femoral fiber composite positioning tube design parameter set.

[0089] Step 105: Design the positioning connection point model and the positioning clamp model according to the positioning tube model, and assemble the positioning tube model, the positioning connection point model, and the positioning clamp model to obtain the locator model. Although the embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. It can be applied to various fields suitable for the present invention. Other modifications can be easily implemented by those skilled in the art. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and examples shown and described herein.

Claims

1. A femoral fiber composite material positioner, characterized in that, It includes a positioning tube (1), the two ends of which are connected to a positioning connection point (2) and a positioning clamp (3) respectively. The positioning tube (1) includes a femoral shell. The femoral shell cavity is provided with a femoral support structure and a filling support structure. The bone porosity of the femoral support structure gradually decreases from the positioning clamp (3) to the positioning connection point (2). The femoral support structure includes a compact bone pore unit region and a cancellous bone pore unit region. The compact bone pore unit region is adjacent to the femoral shell. The pore diameter of the compact bone pore unit region is smaller than that of the cancellous bone pore unit region. Both the pores of the compact bone pore unit region and the pores of the cancellous bone pore unit region are closed pores. The femoral shell is formed by heating and pressurizing a fiber resin layer, and the filling support structure is formed by secondary processing and curing of carbon fiber bundles and resin.

2. A design method for a femoral fiber composite material positioner, applied to the femoral fiber composite material positioner described in claim 1, characterized in that, include: Obtain the design length and maximum design working load of the positioning tube, and determine the value range of the design parameters of the femoral fiber composite positioning tube based on the design length and maximum design working load. The value range of the design parameters of the imitation femoral fiber composite positioning tube is normalized to obtain the value range of the design parameters of the imitation femoral fiber composite positioning tube after preprocessing. The value ranges of the pre-processed femoral fiber composite positioning tube design parameters are arranged and combined to obtain multiple sets of femoral fiber composite positioning tube design parameters. The multiple sets of femoral fiber composite positioning tube design parameters are then fed into the trained optimal BP neural network model. The optimal design parameter set for the femoral fiber composite positioning tube is obtained using the optimal BP neural network model. The positioning tube model is then obtained by modeling based on the optimal design parameter set for the femoral fiber composite positioning tube and the positioning tube length. Based on the positioning tube model, design the positioning connection point model and the positioning clamp model respectively, and assemble the positioning tube model, the positioning connection point model and the positioning clamp model to obtain the locator model; The design parameters of the femoral fiber composite positioning tube include: femoral shell radius, number of femoral shell carbon fiber layers, proportion of cancellous bone pore unit area, and bone pore reduction coefficient. The BP neural network model includes: an input layer neural network, a hidden layer neural network, and an output layer neural network. The input layer neural network includes: a femoral shell radius input neural unit, a femoral shell carbon fiber layer number input neural unit, a cancellous bone pore unit area proportion input neural unit, and a bone pore reduction coefficient input neural unit. The hidden layer neural network includes: 10 hidden neural units and a hidden layer bias. The output layer neural network includes: a deformation output neural unit and an output layer bias.

3. The design method of a femoral fiber composite material positioner according to claim 2, characterized in that, Before obtaining the design length of the positioning tube and the maximum design working load, the following steps are also included: Obtain existing datasets of multiple sets of femoral fiber composite positioning tube design parameters and corresponding multiple positioning tube lengths; A finite element model of a femoral fiber composite positioning tube is established. Based on the existing dataset of multiple sets of design parameters of the femoral fiber composite positioning tube and the corresponding multiple positioning tube lengths, a database of design parameters and deformation of the femoral fiber composite positioning tube is constructed by ABAQUS simulation using Python secondary development. The design parameters and deformation database of the femoral fiber composite positioning tube are preprocessed and training and testing sets are constructed. A BP neural network model for deformation calculation is constructed, and the optimal BP neural network model is obtained by training and testing the BP neural network model using the training set and test set.

4. The design method of a femoral fiber composite material positioner according to claim 3, characterized in that, The process of training and testing the BP neural network model using the training and test sets to obtain the optimal BP neural network model includes: The design parameters and deformation of the femoral fiber composite positioning tube in the training set are weighted and summed with the connection weights of each hidden neural unit, and the values ​​of each hidden neural unit are obtained through the activation function of the input layer. The values ​​of each hidden neural unit are weighted and summed with the connection weights of the corresponding output neural units, and the first output deformation is obtained through the activation function of the output layer. The first error value is obtained by subtracting the first output deformation from the deformation in the training set. The first error value is then backpropagated through a BP neural network model to update the connection weights of each hidden neural unit and the connection weights of the output neural unit. The updated connection weights of each hidden neural unit and the connection weights of the output neural unit are forward propagated again to obtain the second output deformation. The difference between the second output deformation and the deformation in the training set is used to obtain the second error value. It is then determined whether the second error value is greater than the target error value. Yes, the second error value is backpropagated through the BP neural network model again to update the connection weights of each hidden neural unit and the connection weights of the output neural unit. The process is iterated until the error value is less than or equal to the target error value, and then the next step is executed. No, proceed to the next step; Save the connection weights of each hidden layer neural unit and the connection weights of the output neural unit to complete the training of the neural network and obtain the optimal BP neural network model. Then, test the performance of the optimal BP neural network model through a test set.

5. The design method of a femoral fiber composite material positioner according to claim 4, characterized in that, The activation function of the input layer is the logsig function, and the activation function of the output layer is the sigmoid function.

6. The design method of a femoral fiber composite material positioner according to claim 5, characterized in that, The step of updating the connection weights of each hidden neuron and the connection weights of the output neuron by backpropagating the first error value through a BP neural network model includes: The first error value obtained by the LM learning algorithm is multiplied by the learning rate to obtain the adjustment value of the connection weight of each hidden neuron and the adjustment amount of the connection weight of the output neuron, respectively. The adjusted connection weights of each hidden neural unit and the adjusted connection weights of the output neural unit are added to the current connection weights of the hidden neural unit and the output neural unit, respectively, to obtain the updated connection weights of the hidden neural unit and the output neural unit.

7. The design method of a femoral fiber composite material positioner according to claim 2, characterized in that, The optimal design parameter set for the femoral fiber composite positioning tube, obtained using the optimal BP neural network model, includes: Multiple deformation values ​​are obtained using the optimal BP neural network model, and the minimum deformation data is obtained by comparing the multiple deformation values. Based on the minimum deformation data, multiple sets of femoral fiber composite positioning tube design parameter sets are traversed to obtain the corresponding femoral fiber composite positioning tube design parameter set, and the corresponding femoral fiber composite positioning tube design parameter set is the optimal femoral fiber composite positioning tube design parameter set.

Images