Device and method for optimizing constraint device specifications using influence analysis based on particle swarm optimization algorithm

Abstract

The present invention relates to an apparatus and method for optimizing the specifications of a restraint device using an impact analysis based on a particle swarm optimization algorithm. Provided are an apparatus and method for optimizing the specifications of a restraint device using an impact analysis between the specifications of the restraint device and passenger injuries based on a particle swarm optimization algorithm. By sequentially inputting the values ​​of the specifications of each design variable included in the injury result data into the particle swarm optimization algorithm, a passenger injury prediction value is generated according to the degree of injury to the passenger. The optimization algorithm is implemented by forming a plurality of calculation layers for each specific configuration to search for the optimal specifications of the restraint device. Information on the optimal specifications of the restraint device is provided by calculating a comprehensive result index using the passenger injury prediction value. Therefore, information on the optimal specifications of the restraint device can be calculated without performing repeated evaluation tests.

Description

Technical Field

[0001] The present invention relates to a technology capable of determining the specific specifications of a restraint device optimized to minimize passenger injury. More specifically, the present invention relates to an apparatus and method for optimizing the specifications of a restraint device based on a particle swarm optimization algorithm using an impact analysis between the restraint device specifications and passenger injury. Background Art

[0002] In order to obtain the optimal specifications for the design of an automobile restraint, multiple collision tests are usually repeated multiple times to collect the degree of injury to each body part through the test results. Such tests are performed in a manner that changes the specific specifications of the restraint.

[0003] Here, as the collision test, a New Car Assessment Program (NCAP) evaluation method using an actual vehicle collision is mainly used, and this test is generally used as a vehicle safety evaluation.

[0004] According to the NCAP assessment method, a complete vehicle is tested during vehicle development, from the prototype to the pre-production stage. The specific specifications of the adjusted restraints are applied to the complete vehicle. A test dummy is repeatedly installed in the vehicle according to each specific specification. The vehicle collides with a fixed wall or obstacle at a speed of 56 km / h. This method generates test dummy data.

[0005] In this paper, NCAP performance can be calculated by calculating the injury criterion (HIC), neck injury criterion (Nij), chest displacement (CD), and femur injury for each part of the test dummy in the driver and passenger seats from the acquired data. Manual factor analysis is performed using the calculated results, and the optimal restraint specifications can be selected based on the analysis results.

[0006] However, the related art uses a method that manually selects optimal restraint device specifications based on engineers' analysis and experience, raising concerns about unpredictable specifications. Therefore, subjective interpretations based on experience are often included. This also raises the issue of excessive labor and testing costs due to repeated testing caused by a lack of correlation. Summary of the Invention

[0007] Therefore, the present invention has been developed in consideration of the aforementioned problems encountered in the prior art. An object of the present invention is to provide a technique for generating a passenger injury prediction value based on the passenger's injury severity by sequentially inputting the values ​​corresponding to the specifications of each design variable included in injury result data into a particle swarm optimization algorithm. The particle swarm optimization algorithm is implemented by forming multiple computational layers for each specific configuration to search for the optimal specifications of the restraint device. Information regarding the optimal specifications of the restraint device is calculated using the passenger injury prediction value.

[0008] According to an embodiment of the present invention, a device for optimizing the specifications of a restraint device based on a particle swarm optimization algorithm utilizes an impact analysis between the specifications of the restraint device and passenger injuries. The device utilizes a particle swarm optimization algorithm including multiple computing layers, and the device includes: a restraint device specification input unit configured to utilize each specific configuration of the restraint device as each design variable, and receive passenger injury data for each part according to each design variable specification. Thus, an injury result data table for each specification is generated. The device further includes: a passenger injury prediction unit configured to generate a passenger injury prediction value by inputting the injury result data table for each design variable specification into a prediction model for passenger injury prediction results for each design variable specification and each part. The passenger injury prediction unit is further configured to calculate a comprehensive result index using the passenger injury prediction value. The device further includes an optimal specification selection unit configured to select the specific configuration of the restraint device included in the data group with the highest comprehensive result index as the optimal value.

[0009] According to an embodiment of the present invention, the passenger injury prediction unit may include an injury prediction result generator. The injury prediction result generator may be configured to generate a prediction model based on a particle swarm optimization algorithm for each design variable specification and the passenger injury prediction result of each part using a third-order Hermit interpolation polynomial for each interval. The injury prediction result generator may be further configured to input the injury result data table of each design variable specification into the prediction model to generate a passenger injury prediction value. The passenger injury prediction unit may further include a comprehensive result calculator configured to calculate a comprehensive result index according to a preset standard using the passenger injury prediction result of each part.

[0010] According to an embodiment of the present invention, the comprehensive result calculator may calculate a comprehensive result index based on the passenger injury prediction result of each part according to a New Car Assessment Program (NCAP) evaluation method.

[0011] According to an embodiment of the present invention, the restraint device specification input unit may utilize the criteria of vent hole diameter, upper tether length, lower tether length, and load limiter force of a specific configuration of the restraint device as each design variable.

[0012] According to an embodiment of the present invention, the restraint device specification input unit can analyze the degree of influence by sequentially inputting the numerical value of each design variable specification included in the injury result data into each calculation layer of the particle swarm optimization algorithm. The particle swarm optimization algorithm can be implemented by forming a plurality of calculation layers for each design variable to calculate the weight of each variable by performing component analysis on each normalized variable.

[0013] According to an embodiment of the present invention, the restraint device specification input unit may be configured to enable multiple calculation layers to calculate the degree of influence using each of vent hole diameter, upper tether length, lower tether length and load limiter force as a variable.

[0014] According to an embodiment of the present invention, the restraint device specification input unit may be configured such that a plurality of calculation layers are arranged in the order of a vent diameter layer, an upper tether length layer, a lower tether length layer, and a load limiter force layer. The restraint device specification input unit may also be configured such that a result value calculated in each layer serves as an input value for a layer located subsequent to the layer.

[0015] According to an embodiment of the present invention, the restraint device specification input unit may normalize each variable before inputting the standard of the specific configuration of the restraint device as each variable. The restraint device specification input unit may further perform a weight analysis on each normalized variable to calculate a weight for each variable.

[0016] According to an embodiment of the present invention, the restraint device specification input unit may calculate a passenger injury prediction value according to the passenger's injury degree by reflecting the weight of each variable.

[0017] According to an embodiment of the present invention, the injury prediction result generating unit can calculate the passenger injury prediction results of the head injury standard (HIC), chest dislocation (CD), and neck injury standard (Nij) as the passenger injury prediction results of each part.

[0018] According to an embodiment of the present invention, a method optimizes the specifications of a restraint device based on a particle swarm optimization algorithm using an impact analysis between the specifications of the restraint device and the passenger injury. The method utilizes a particle swarm optimization algorithm including a plurality of calculation layers, and the method includes: utilizing each specific configuration of the restraint device as each design variable, and receiving the passenger injury data of each part according to each design variable specification. Thus, an injury result data table for each specification is generated. The method further includes: generating a passenger injury prediction value by inputting the injury result data table for each design variable specification into a prediction model for the passenger injury prediction result for each design variable specification and each part. The method further includes: calculating a comprehensive result index by utilizing the passenger injury prediction value, and selecting the specific configuration of the restraint device included in the data group with the highest comprehensive result index as the optimal value.

[0019] According to an embodiment of the present invention, calculating a comprehensive result index may include generating a prediction model based on a particle swarm optimization algorithm for each design variable specification and each location's passenger injury prediction result. The prediction model is generated using a third-order Hermit interpolation polynomial for each interval, and a passenger injury prediction value is generated by inputting a table of injury results for each design variable specification into the prediction model. The comprehensive result index may be calculated based on preset criteria using the passenger injury prediction results for each location.

[0020] According to an embodiment of the present invention, calculating the comprehensive result index may include: calculating the comprehensive result index according to the NCAP evaluation method based on the passenger injury prediction result of each part.

[0021] According to an embodiment of the present invention, generating an injury results data table may utilize the criteria of vent diameter, upper tether length, lower tether length, and load limiter force for a specific configuration of the restraint device as each design variable.

[0022] According to an embodiment of the present invention, generating the damage result data table may include analyzing the degree of influence by sequentially inputting a numerical value according to a specification of each design variable included in the damage result data into each calculation layer of a particle swarm optimization algorithm. The particle swarm optimization algorithm may be implemented by forming a plurality of calculation layers for each design variable to calculate a weight of each variable by performing component analysis on each normalized variable.

[0023] According to an embodiment of the present invention, the generation of the injury result data table may be configured so that multiple calculation layers can calculate the impact degree using each of the vent diameter, upper tether length, lower tether length, and load limiter force as a variable.

[0024] According to an embodiment of the present invention, the injury result data table may be configured such that the plurality of calculation layers are arranged in the order of the vent diameter layer, the upper tether length layer, the lower tether length layer, and the load limiter force layer. The injury result data table may also be configured such that the result value calculated in each layer serves as the input value for the layer located subsequent thereto.

[0025] According to an embodiment of the present invention, generating the injury result data table may include normalizing each variable before inputting the specific configuration criteria of the restraint device as each variable. Generating the injury result data table may further include performing a weight analysis on each normalized variable to calculate a weight for each variable.

[0026] According to an embodiment of the present invention, generating the injury result data table may include calculating a passenger injury prediction value according to the passenger's injury degree by reflecting the weight of each variable.

[0027] According to an embodiment of the present invention, generating the passenger injury prediction value may include calculating the passenger injury prediction results of HIC, CD, and Nij as the passenger injury prediction results for each part.

[0028] According to the present invention, even without repeating evaluation tests for each specific configuration of the restraint system, a passenger injury prediction value based on the passenger injury level can be generated for each specific configuration of the restraint system. This generation of passenger injury prediction values ​​can be performed using multiple computational layers implemented using a particle swarm optimization algorithm, utilizing injury results for standard configurations based on previously performed evaluation tests. Therefore, information regarding the optimal configuration of the restraint system can be calculated without repeating evaluation tests. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The above and other objects, features and other advantages of the present invention will become more clearly understood from the following detailed description when taken in conjunction with the accompanying drawings, in which:

[0030] Figure 1 A block diagram illustrating an apparatus for optimizing specifications of a constraint device based on a particle swarm optimization algorithm according to an embodiment of the present invention;

[0031] Figure 2 To show Figure 1 Detailed configuration diagram of the passenger injury prediction unit disclosed in.

[0032] Figure 3 A schematic diagram illustrating a process of inputting data into a device for optimizing specifications of a constraint device based on a particle swarm optimization algorithm according to an embodiment of the present invention;

[0033] Figure 4A schematic diagram illustrating a table of injury result data for each specification of a generated design variable according to an embodiment of the present invention;

[0034] Figure 5 A schematic diagram illustrating calculating the weight of each variable by normalizing each variable according to an embodiment of the present invention;

[0035] Figure 6 A schematic diagram illustrating a method for analyzing an impact by sequentially inputting a numerical value according to a specification of each design variable included in damage result data into each calculation layer to implement a particle swarm optimization algorithm using a plurality of calculation layers according to an embodiment of the present invention;

[0036] Figure 7 A schematic diagram illustrating a comprehensive result indicator result table generated according to an embodiment of the present invention;

[0037] Figure 8 Tables and graphs comparing predicted results with actual analysis results according to embodiments of the present invention; and

[0038] Figure 9 The flowchart is a method for optimizing the specification of a constraint device based on a particle swarm optimization algorithm according to an embodiment of the present invention. DETAILED DESCRIPTION

[0039] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Therefore, those skilled in the art can easily implement the present invention. Those skilled in the art will appreciate that various modifications may be made to the described embodiments without departing from the spirit or scope of the present invention.

[0040] In order to simplify the description, parts not related to the description are omitted in the drawings. Throughout the specification, the same or similar reference numerals represent the same or equivalent parts.

[0041] Throughout the specification, when a section “includes” a certain component, it means that the section does not exclude other components, but may further include other components, unless specifically stated otherwise.

[0042] Hereinafter, an apparatus and method for optimizing the specification of a constraint device based on a particle swarm optimization algorithm according to an embodiment of the present invention will be described with reference to the accompanying drawings.

[0043] Figure 1 1 is a block diagram illustrating an apparatus for optimizing specifications of a constraint device based on a particle swarm optimization algorithm according to an embodiment of the present invention.

[0044] refer to Figure 1According to an embodiment of the present invention, a device for optimizing restraint device specifications based on a particle swarm optimization algorithm can utilize a particle swarm optimization algorithm comprising multiple computational layers. The device may include a restraint device specification input unit 100, a passenger injury prediction unit 200, and an optimal specification selection unit 300.

[0045] According to an embodiment of the present invention, the particle swarm optimization algorithm is an algorithm that is inspired by the social behavior and dynamics of fish and birds. The particle swarm optimization algorithm can be referred to as an algorithm that can search for an optimal solution (specification) based on randomly given specification information through the information exchange and mobility characteristics of entities.

[0046] According to the above embodiment, each entity can determine a new position by its previous position, its own optimal position, and the optimal position of the entire cluster.

[0047] The restraint device specification input unit 100 may receive passenger injury data of each portion according to the specification of each design variable by using each specific configuration of the restraint device as each design variable, and generate an injury result data table for each specification.

[0048] According to an embodiment of the present invention, the specific configuration of the restraint device may include: a vent in the driver airbag (DAB), an upper tether, and a lower tether. The specific configuration of the restraint device may also include, but is not limited to, a load limiter in a seat belt. The specific configuration of the restraint device may be used without limitation, as long as it is included in the restraint device that may affect passenger injury due to specification changes.

[0049] According to embodiments of the present invention, criteria for vent diameter, upper tether length, lower tether length, and load limiter force may be set as each design variable for a specific configuration of the restraint device.

[0050] According to an embodiment of the present invention, before the standard of the specific configuration of the restraint device is input as each variable, each variable may be normalized and component analysis may be performed on each normalized variable to calculate the weight of each variable.

[0051] According to an embodiment of the present invention, the restraint device specification input unit 100 may set each specific configuration as a design variable. The restraint device specification input unit 100 may receive passenger injury data for each position based on the change in the specification of each design variable. Based on the received passenger injury data, the restraint device specification input unit 100 may generate an injury result data table for each specification.

[0052] According to an embodiment of the present invention, to calculate the weight of each variable by performing component analysis on each normalized variable, a particle swarm optimization algorithm can be used to analyze the degree of influence. The particle swarm optimization algorithm can be implemented by forming multiple calculation layers for each design variable. The degree of influence can be analyzed by sequentially inputting numerical values ​​according to the specifications of each design variable included in the injury result data into each calculation layer.

[0053] According to an embodiment of the present invention, the weight of each variable is calculated in Figure 5 Shown in detail.

[0054] According to an embodiment of the present invention, a plurality of calculation layers may calculate the degree of influence by using each of the vent diameter, the upper tether length, the lower tether length, and the load limiter force as a variable.

[0055] According to an embodiment of the present invention, multiple calculation layers are arranged in the order of vent diameter layer, upper tether length layer, lower tether length layer, and load limiter force layer. Therefore, the result value calculated in each layer can be used as the input value of the layer located after that layer.

[0056] The damage result data table for each specification generated in this article can be expressed as Figure 4 shown.

[0057] According to an embodiment of the present invention, the restraint device specification input unit 100 may normalize each variable before inputting a standard of a specific configuration of the restraint device as each variable.

[0058] According to an embodiment of the present invention, a passenger injury prediction value may be calculated according to the passenger's injury degree by reflecting the weight of each variable.

[0059] The passenger injury prediction unit 200 generates a passenger injury prediction value by inputting the injury result data table of each design variable specification into a prediction model of passenger injury prediction results for each design variable specification and each part. The passenger injury prediction unit 200 uses the passenger injury prediction value to calculate a comprehensive result index.

[0060] According to an embodiment of the present invention, a prediction model for passenger injury prediction results at each location can be implemented based on a particle swarm optimization algorithm. The prediction model can be trained using output values ​​obtained by inputting an injury result data table.

[0061] According to an embodiment of the present invention, training may be performed by reflecting the weight calculated by the restraint device specification input unit 100 in the prediction model.

[0062] According to an embodiment of the present invention, when each specification of the design variables is input, the prediction model may apply numerical analysis techniques to search for a function with the highest accuracy of passenger injury prediction value.

[0063] refer to Figure 2 The passenger injury prediction unit 200 is described in more detail.

[0064] The optimal specification selection unit 300 may select the specific configuration of the restraint device included in the data group having the highest comprehensive result index as the optimal value.

[0065] According to an embodiment of the present invention, a passenger injury prediction value is calculated for each combination of the specifications of each design variable. When the calculated injury prediction value is low, a higher-scoring comprehensive outcome index is calculated. Therefore, the combination of specifications of each design variable with the highest comprehensive outcome index can be selected as the optimal one.

[0066] According to an embodiment of the present invention, a comprehensive result index may be calculated based on the damage prediction result of each part according to the New Car Assessment Program (NCAP) evaluation method.

[0067] In this article, the NCAP assessment method applies the adjusted specific specifications of the restraint device to the entire vehicle. The NCAP assessment method repeatedly installs the test dummy on the vehicle according to each specific specification and causes the vehicle to collide with a fixed wall or obstacle at a speed of 56kph. Therefore, the NCAP assessment method obtains test dummy data and thus obtains passenger injury data. In the injury data, the NCAP performance can be calculated by calculating the injury of each part of the test dummy on the driver's seat / passenger seat from the acquired data, namely, the head injury criterion (HIC), neck injury criterion (Nij), chest displacement (CD) and femur injury. The calculation results can be used to select the optimal restraint device specifications.

[0068] According to an embodiment of the present invention, HIC, CD and Nij can be selected as predictive injury sites. The fixed variable knee load (Fz) of 2.8 kN cannot be used as a predictive variable.

[0069] Figure 2 To show Figure 1 Detailed configuration diagram of the passenger injury prediction unit 200 disclosed in .

[0070] refer to Figure 2 , Figure 1The passenger injury prediction unit 200 disclosed in may include an injury prediction result generator 210 and a comprehensive result calculator 220 .

[0071] Injury prediction result generator 210 generates a prediction model based on a particle swarm optimization algorithm for each design variable specification and each component's passenger injury prediction result. The prediction model is generated using a third-order Hermit interpolation polynomial for each interval. Injury prediction result generator 210 generates each design in the prediction model. Injury prediction result generator 210 generates passenger injury prediction values ​​by inputting the injury result data table for each design variable specification into the prediction model.

[0072] According to an embodiment of the present invention, a third-order Hermit interpolation polynomial for each interval can be used as a numerical analysis technique. However, the present invention is not limited thereto. Any function whose optimal value can be selected by numerical analysis can be used without restriction.

[0073] Here, the third-order Hermit interpolation polynomial for each interval may represent a comparison of the interpolation results generated in the spline and pchip for two different functions. The third-order Hermit interpolation polynomial for each interval can form a vector consisting of the x value, the function value y at that point, and the query point xq.

[0074] Additionally, an operation may be performed of calculating an interpolated value at a query point using spline and pchip and plotting the interpolated function value at the query point for comparison.

[0075] The comprehensive result calculator 220 may calculate a comprehensive result index according to a preset standard by using the injury prediction result of each part.

[0076] According to an embodiment of the present invention, a passenger injury prediction value for each location is calculated based on each design variable specification. The passenger injury prediction value for each location can be HIC, CD, and Nij. A comprehensive result index can be calculated based on the calculated passenger injury prediction value for each location.

[0077] According to an embodiment of the present invention, a comprehensive result index may be calculated by reflecting the weight of each variable calculated by the restraint device specification input unit 100 in the calculated passenger injury prediction value of each part.

[0078] Figure 3 The diagram is a schematic diagram illustrating a process of inputting data into a device for optimizing the specification of a constraint device based on a particle swarm optimization algorithm according to an embodiment of the present invention.

[0079] refer to Figure 3According to an embodiment of the present invention, passenger injury data can be obtained by evaluating passenger injury results according to each design variable specification. Passenger injury data can be obtained by using the specific configuration included in the restraint device as a design variable.

[0080] According to the above embodiment, the acquired passenger injury data for each part can be normalized and a weight can be calculated according to each design variable. A prediction model can also be generated and trained, which predicts the passenger injury data according to the specifications of each design variable that reflects the passenger injury data.

[0081] In addition, the trained prediction model can be used to generate passenger injury prediction values ​​for various combinations of specific specifications of each design variable and a comprehensive result indicator based on the injury prediction values.

[0082] Additionally, the specific configuration of the restraint device included in the data set having the highest comprehensive result index may be selected as optimal.

[0083] Figure 4 Schematic diagram showing a table of damage result data for each specification according to generated design variables according to an embodiment of the present invention.

[0084] refer to Figure 4 , a table of injury results for each specification is shown based on the generated design variables. The injury result data table can be generated by changing the passenger injury data in a data table format according to the changes in the specifications of each design variable. According to an embodiment of the present invention, the specifications of each design variable can be a vent, an upper tether, a lower tether, and a load limiter.

[0085] Figure 5 FIG. 1 is a schematic diagram illustrating calculating the weight of each variable by normalizing each variable according to an embodiment of the present invention.

[0086] refer to Figure 5 According to an embodiment of the present invention, a formula for calculating the weight of each variable by normalizing each variable is disclosed. The weight calculated as described above can be used to train a prediction model. When calculating a comprehensive evaluation index, the calculation can be performed by reflecting the weight in each design variable.

[0087] Figure 6 A schematic diagram illustrating a method for analyzing impact by sequentially inputting numerical values ​​according to specifications of each design variable included in damage result data into each computational layer. According to an embodiment of the present invention, the analysis is performed using a particle swarm optimization algorithm implemented using multiple computational layers.

[0088] refer to Figure 6By sequentially inputting the numerical values ​​of each design variable specification included in the damage result data into each calculation layer, the degree of damage impact on each part can be analyzed based on the specification. This analysis is performed using a particle swarm optimization algorithm, which utilizes multiple calculation layers formed for each design variable to perform component analysis on each normalized variable. This allows the weight of each variable to be calculated.

[0089] Figure 7 It is a schematic diagram showing a comprehensive result indicator result table generated according to an embodiment of the present invention.

[0090] refer to Figure 7 , showing a comprehensive evaluation index result table generated according to an embodiment of the present invention. The comprehensive evaluation index is calculated by using each damage prediction value in the specification combination of each design variable. The comprehensive evaluation index selection with the maximum value can be selected as the optimal specification.

[0091] Figure 8 Tables and graphs comparing predicted results to actual analysis results are shown according to embodiments of the present invention.

[0092] refer to Figure 8 According to an embodiment of the present invention, the results predicted by the particle swarm optimization algorithm showed no difference or close difference within the error range from the actual analysis results. The actual analysis utilized Madymo software, which is used for occupant safety system analysis in the automotive and transportation industries. The results showed a significant effect in saving time and money compared to performing a large number of evaluation tests for practical understanding.

[0093] Figure 9 The flowchart is a method for optimizing the specification of a constraint device based on a particle swarm optimization algorithm according to an embodiment of the present invention.

[0094] By using each specific configuration of the restraint device as each design variable, the passenger injury data of each part according to the specification of each design variable is received, and thus, an injury result data table of each specification can be generated (S10).

[0095] According to an embodiment of the present invention, each specific configuration of the restraint device is used as each design variable, and the passenger injury data of each part according to the specification of each design variable can be input, thereby generating an injury result data table for each specification.

[0096] According to an embodiment of the present invention, the specific configuration of the restraint device is configured such that the DAB may include a vent, an upper tether (UPR tether), and a lower tether (LWR tether), and the seat belt may include a load limiter. However, the present invention is not limited thereto. Any component included in the restraint device that can affect passenger injury according to the changes in specifications may be utilized without restriction.

[0097] According to embodiments of the present invention, criteria for vent diameter, upper tether length, lower tether length, and load limiter force for a specific configuration of the restraint device may be set as each design variable.

[0098] According to an embodiment of the present invention, before inputting the standard of the specific configuration of the restraint device as each variable, each variable is normalized. The weight of each variable can be calculated by performing component analysis on each normalized variable.

[0099] According to an embodiment of the present invention, each specific configuration can be set as a design variable. Passenger injury data for each part of the vehicle according to the specification change of each design variable is received. An injury result data table can be generated for each specification based on the received passenger injury data.

[0100] According to an embodiment of the present invention, the degree of influence can be analyzed by sequentially inputting the values ​​of each design variable included in the injury result data into each calculation layer. This analysis is performed using a particle swarm optimization algorithm that utilizes multiple calculation layers for each design variable. Therefore, the weight of each variable is calculated by performing component analysis on each normalized variable.

[0101] According to an embodiment of the present invention, a plurality of calculation layers may calculate each degree of influence by using each of the vent diameter, the upper tether length, the lower tether length, and the load limiter force as a variable.

[0102] According to an embodiment of the present invention, multiple calculation layers are arranged in the order of vent diameter layer, upper tether length layer, lower tether length layer, and load limiter force layer. Therefore, the result value calculated in each layer can be used as the input value of the layer located after that layer.

[0103] According to an embodiment of the present invention, before inputting the criteria of the specific configuration of the restraint device as a variable, each variable may be normalized.

[0104] According to an embodiment of the present invention, a passenger injury prediction value may be calculated according to the passenger's injury degree by reflecting the weight of each variable.

[0105] The injury result data table of each design variable specification is input into the prediction model of the passenger injury prediction result of each design variable specification and each part. Thus, the passenger injury prediction value is generated (S20).

[0106] According to an embodiment of the present invention, the injury result data table of each design variable specification is input into the passenger injury prediction result for each part and the prediction model of each design variable specification, thereby generating a passenger injury prediction value.

[0107] According to an embodiment of the present invention, a prediction model for passenger injury prediction results at each location can be implemented based on a particle swarm optimization algorithm. The prediction model can be trained using output values ​​obtained by inputting an injury result data table.

[0108] According to an embodiment of the present invention, training may be performed by reflecting the calculated weights in a prediction model.

[0109] According to an embodiment of the present invention, when each specification of the design variables is input, the prediction model may apply numerical analysis techniques to search for a function with the highest accuracy of passenger injury prediction value.

[0110] According to an embodiment of the present invention, a prediction model is generated based on the particle swarm optimization algorithm for passenger injury prediction results for each design variable specification and each location. This prediction model is generated using a third-order Hermit interpolation polynomial for each interval. Passenger injury prediction values ​​can be generated by inputting a table of injury results for each design variable specification into the prediction model.

[0111] According to an embodiment of the present invention, a third-order Hermit interpolation polynomial for each interval can be used as a numerical analysis technique. However, the present invention is not limited thereto, and any function whose optimal value can be selected by numerical analysis can be used without limitation.

[0112] Here, the third-order Hermit interpolation polynomial for each interval may represent a comparison of the interpolation results generated in the spline and pchip for two different functions. The third-order Hermit interpolation polynomial for each interval can form a vector consisting of the x value, the function value y at that point, and the query point xq.

[0113] Additionally, an operation may be performed of calculating an interpolated value at a query point using spline and pchip and plotting the interpolated function value at the query point for comparison.

[0114] The passenger injury prediction value is used to calculate a comprehensive result index (S30).

[0115] According to an embodiment of the present invention, the passenger injury prediction value can be used to calculate the comprehensive result index. According to an embodiment, the injury prediction result of each part can be used to calculate the comprehensive result index according to a preset standard.

[0116] According to an embodiment of the present invention, a predicted passenger injury value for each location is calculated based on the specifications of each design variable. The predicted passenger injury value for each location can be HIC, CD, and Nij. A comprehensive result index can be calculated based on the calculated predicted passenger injury value for each location.

[0117] According to an embodiment of the present invention, a comprehensive result index may be calculated by reflecting the calculated weight of each variable in the passenger injury prediction value of each part.

[0118] The specific configuration of the restraint device included in the data group having the highest comprehensive result index is selected as the optimal value (S40).

[0119] The specific configuration of the restraint device included in the data set having the highest overall result index may be selected as the optimal value.

[0120] According to an embodiment of the present invention, a passenger injury prediction value is calculated for each combination of design variable specifications. The lower the calculated injury prediction value, the higher the comprehensive result index. Therefore, the combination of design variable specifications with the highest comprehensive result index can be selected as the optimal value.

[0121] According to an embodiment of the present invention, a comprehensive result index may be calculated based on the injury prediction result of each body part according to the NCAP evaluation method.

[0122] According to an embodiment of the present invention, HIC, CD and Nij can be selected as predictive injury sites.The fixed variable knee load (Fz) of 2.8 kN cannot be used as a predictive variable.

[0123] The embodiments of the present invention are not limited to the above-described devices and / or methods. The scope of the present invention is not limited thereto. Various modifications and improvements of the basic concept of the present invention as defined in the appended claims by those skilled in the art are also within the scope of the present invention.

Claims

1. A device for optimizing restraint device specifications based on a particle swarm optimization algorithm using an impact analysis between restraint device specifications and passenger injuries, the device utilizing a particle swarm optimization algorithm comprising multiple computational layers, the device comprising: a restraint device specification input unit configured to receive passenger injury data at each location according to each design variable specification using each specific configuration of the restraint device as each design variable to generate an injury result data table for each specification, the restraint device specification input unit being further configured to analyze the degree of influence by sequentially inputting a numerical value according to each design variable specification included in the injury result data into each calculation layer of a particle swarm optimization algorithm; a passenger injury prediction unit configured to generate a passenger injury prediction value by inputting an injury result data table for each design variable specification into a prediction model of passenger injury prediction results for each design variable specification and each portion, and configured to calculate a comprehensive result index using the passenger injury prediction value; as well as an optimal specification selection unit configured to select, as an optimal value, a specific configuration of the restraint device included in the data group having the highest comprehensive result index, Wherein, the passenger injury prediction unit includes: an injury prediction result generator configured to generate a prediction model based on a particle swarm optimization algorithm for each design variable specification and each part of the passenger injury prediction result using a third-order Hermit interpolation polynomial for each interval, and the injury prediction result generator is configured to input the injury result data table for each design variable specification into the prediction model to generate a passenger injury prediction value; and A comprehensive result calculator is configured to calculate a comprehensive result index according to a preset standard using the passenger injury prediction results of each part. In which, the restraint device specification input unit is configured so that multiple calculation layers are arranged in the order of vent diameter layer, upper tether length layer, lower tether length layer and load limiter force layer, and the restraint device specification input unit is configured so that the result value calculated in each layer becomes the input value of the layer located behind this layer.

2. The device according to claim 1, wherein The comprehensive result calculator is further configured to calculate a comprehensive result index based on the passenger injury prediction results of each part according to the new car evaluation procedure evaluation method.

3. The device according to claim 1, wherein The restraint device specification input unit is further configured to utilize criteria for a vent diameter, upper tether length, lower tether length, and load limiter force of a specific configuration of the restraint device as each design variable.

4. The device according to claim 1, wherein The particle swarm optimization algorithm is implemented using a plurality of calculation layers formed for each design variable to calculate the weight of each variable by performing component analysis on each normalized variable.

5. The device according to claim 4, wherein The restraint device specification input unit is configured so that the plurality of calculation layers can calculate the degree of influence using each of the vent hole diameter, the upper tether length, the lower tether length, and the load limiter force as a variable.

6. The device according to claim 1, wherein The restraint device specification input unit is further configured to normalize each variable before inputting the standard of the specific configuration of the restraint device as each variable, and the restraint device specification input unit is further configured to perform weight analysis on each normalized variable to calculate the weight of each variable.

7. The device according to claim 6, wherein The restraint device specification input unit is further configured to calculate a passenger injury prediction value according to the passenger's injury level by reflecting the weight of each variable.

8. The device according to claim 1, wherein The injury prediction result generating unit is further configured to calculate passenger injury prediction results of head injury standard, chest dislocation, and neck injury standard as passenger injury prediction results for each part.

9. A method for optimizing restraint device specifications based on a particle swarm optimization algorithm using an impact analysis between restraint device specifications and passenger injuries, the method utilizing a particle swarm optimization algorithm comprising multiple computational layers, the method comprising: Using each specific configuration of the restraint device as each design variable, receiving passenger injury data at each location according to each design variable specification to generate an injury result data table for each specification; generating a passenger injury prediction value by inputting the injury result data table for each design variable specification into a prediction model of passenger injury prediction results for each design variable specification and each portion, and calculating a comprehensive result index by using the passenger injury prediction value; The specific configuration of the restraint device included in the data set with the highest comprehensive result index is selected as the optimal value, The generation of the damage result data table includes: The degree of impact is analyzed by sequentially inputting the numerical value according to the specification of each design variable included in the injury result data into each calculation layer of the particle swarm optimization algorithm; Among them, the calculation of comprehensive result indicators includes: Using the third-order Hermit interpolation polynomial for each interval, a prediction model is generated based on the particle swarm optimization algorithm for each design variable specification and the passenger injury prediction results for each location. The passenger injury prediction value is generated by inputting the injury result data table for each design variable specification into the prediction model. Using the passenger injury prediction results of each part, the comprehensive result index is calculated according to the preset standards. Among them, the generated injury result data table is set so that multiple calculation layers are arranged in the order of vent diameter layer, upper tether length layer, lower tether length layer and load limiter force layer, and the result value calculated in each layer becomes the input value of the layer behind this layer.

10. The method according to claim 9, wherein: Calculation of comprehensive outcome indicators includes: Based on the passenger injury prediction results of each part, the comprehensive result index is calculated according to the new car assessment procedure evaluation method.

11. The method according to claim 9, wherein The injury results data table was generated using the criteria for the vent diameter, upper tether length, lower tether length, and load limiter force for the specific configuration of the restraint as each design variable.

12. The method according to claim 9, wherein Generating the damage result data table further includes: The particle swarm optimization algorithm is implemented using a plurality of calculation layers formed for each design variable to calculate the weight of each variable by performing component analysis on each normalized variable.

13. The method according to claim 12, wherein: The generated injury result data table is configured to enable a plurality of calculation layers to calculate the degree of impact using each of the vent diameter, the upper tether length, the lower tether length, and the load limiter force as a variable.

14. The method according to claim 9, wherein Generating the damage result data table includes: Each variable is normalized before a criterion for a specific configuration of the restraint device is input as each variable, and a weight analysis is performed on each normalized variable to calculate a weight of each variable.

15. The method according to claim 14, wherein Generating the damage result data table includes: By reflecting the weight of each variable, the passenger injury prediction value is calculated according to the passenger's injury level.

16. The method according to claim 9, wherein Generating passenger injury predictions involves: The passenger injury prediction results of the head injury standard, chest dislocation, and neck injury standard are calculated as the passenger injury prediction results for each part.

Images