Automobile door cover mounting control method and device, computer device, medium and product

Abstract

The application relates to an automobile door cover installation control method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: in response to an installation instruction for a to-be-installed door cover, determining door cover information of the to-be-installed door cover; determining a target installation and adjustment model matched with the door cover information from a plurality of trained candidate installation and adjustment models; determining a first gap surface difference of the to-be-installed door cover at an initial installation and adjustment position under the condition that the to-be-installed door cover is controlled to move to the initial installation and adjustment position; inputting the first gap surface difference into the target installation and adjustment model, and determining an updated installation and adjustment position of the to-be-installed door cover based on the output of the target installation and adjustment model; and if a second gap surface difference of the to-be-installed door cover at the updated installation and adjustment position meets an installation condition, installing the to-be-installed door cover at the updated installation and adjustment position. The method can improve the installation and adjustment accuracy and success rate in the door cover installation process, thereby improving the door cover installation efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automobile production, in particular to an automobile door cover installation control method and device, computer equipment, a storage medium and a computer program product. BACKGROUND

[0002] With the rapid development of production automation and intelligent technology, using artificial intelligence and visual algorithms to guide robots to manufacture and assemble parts has become a general trend in the industry. Compared with traditional manual manufacturing, automated and intelligent automobile manufacturing technology has the advantages of high efficiency and low cost. The automatic assembly function of installing vehicle body accessories such as doors and hoods on the vehicle body also has such advantages.

[0003] When using the automatic assembly function, it is found that if only a fixed trajectory is used to install the door cover, the gaps and surface differences between the door cover and the vehicle body are uneven, affecting the overall appearance of the vehicle. In order to further eliminate process errors, a 321 adjustment algorithm is usually used to automatically adjust the automatic assembly function to achieve the requirement that the installed door cover meets the requirements of aesthetics and balance in terms of gaps and surface differences at all places on the vehicle body.

[0004] Although the 321 adjustment algorithm has the advantages of fast rhythm and high efficiency, since the technical principle of the algorithm is to first determine the parallelism of the door cover and the vehicle body through three measuring points, then control the rotation and high-low gap of the door cover on the parallel plane through two measuring points, and finally determine the left-right gap of the door cover through one measuring point, it means that when moving the latter steps, the adjusted gap and surface difference of the measuring points referred to in the former steps may be affected, so that the entire door cover is still deviated from the ideal position, resulting in a low success rate of installation and adjustment, and still requiring manual intervention. Therefore, the current commonly used automatic adjustment scheme has the problems of low adjustment accuracy and low success rate of installation and adjustment. SUMMARY

[0005] Therefore, it is necessary to provide an automobile door cover installation control method, device, computer equipment, computer readable storage medium and computer program product capable of improving the installation accuracy and success rate of installation and adjustment during the installation of the door cover, and thereby improving the installation efficiency of the door cover.

[0006] In a first aspect, the present application provides an automobile door cover installation control method, which comprises:

[0007] In response to an installation instruction for a door cover to be installed, determining door cover information of the door cover to be installed;

[0008] From a plurality of trained candidate installation and adjustment models, determining a target installation and adjustment model adapted to the door cover information;

[0009] In a case that the door cover to be installed is controlled to move to an initial installation position, a first gap face difference of the door cover to be installed in the initial installation position is determined; a training sample of the target installation model includes sample gap face differences respectively corresponding to sample installation positions of a sample door cover represented by the door cover information; and the sample installation positions are determined based on the initial installation position.

[0010] The first gap face difference is input into the target installation model, and an updated installation position of the door cover to be installed is determined based on an output of the target installation model.

[0011] If a second gap face difference of the door cover to be installed in the updated installation position satisfies an installation condition, the door cover to be installed is installed in the updated installation position.

[0012] In one of the embodiments, the door cover information includes a vehicle type to which the door cover to be installed belongs and a door cover position identifier of the door cover to be installed in the vehicle type.

[0013] The target installation model adapted to the door cover information is determined from the plurality of candidate installation models, including:

[0014] A plurality of selected installation models adapted to the vehicle type are determined from the plurality of candidate installation models according to the vehicle type to which the door cover to be installed belongs.

[0015] A target installation model adapted to the door cover position identifier is determined from the plurality of selected installation models based on the door cover position identifier of the door cover to be installed in the vehicle type.

[0016] In one of the embodiments, the method further includes:

[0017] For each sample door cover, an initial installation position and an initial installation model of the sample door cover are obtained; and the initial installation position is an installation position of the sample door cover in an ideal installation state.

[0018] A plurality of sample installation positions are determined according to the initial installation position and a preset training sample expansion manner.

[0019] The sample door cover is respectively controlled to move to each of the sample installation positions, and a sample gap face difference of the sample door cover in each of the sample installation positions is determined.

[0020] A training sample of the sample door cover is determined based on each of the sample gap face differences and each of the sample installation positions.

[0021] The initial installation model is subjected to model training based on the training sample, and a target installation model after training is obtained.

[0022] In one of the embodiments, the determining of the plurality of sample adjustment positions according to the initial adjustment position and the preset sample expansion manner comprises:

[0023] dividing a sample generation space for the sample door cover according to a preset sample expansion range with the initial adjustment position as a center position;

[0024] establishing a plurality of anchor points on each spatial axis of the sample generation space based on a preset anchor point establishment manner;

[0025] determining each of the anchor points and the initial adjustment position as a sample generation reference point of the sample door cover;

[0026] for each of the sample generation reference points, generating a plurality of sample adjustment positions for the sample generation reference point based on a Gaussian distribution scheme with the sample generation reference point as a center point.

[0027] In one of the embodiments, the method further comprises:

[0028] if the second gap surface difference of the to-be-installed door cover does not satisfy the installation condition, updating the first gap surface difference using the second gap surface difference to obtain an updated first gap surface difference;

[0029] returning to the step of inputting the first gap surface difference into the target adjustment model based on the updated first gap surface difference.

[0030] In one of the embodiments, the method further comprises:

[0031] in response to an installation completion event of the to-be-installed door cover, determining an installation surface difference of an installed door cover and a flat suction point surface difference of the installed door cover in a flat suction state;

[0032] determining a correction adjustment position of the installed door cover according to difference information of the installation surface difference and the flat suction point surface difference and a correction rotation dimension of the installed door cover;

[0033] performing adjustment correction on the installed door cover based on the correction adjustment position;

[0034] determining an adjustment re-inspection result of the installed door cover according to a third gap surface difference of the installed door cover after the adjustment correction is completed.

[0035] In a second aspect, the application further provides an automobile door cover installation control device, the device comprising:

[0036] an instruction response module configured to determine door cover information of a to-be-installed door cover in response to an installation instruction of the to-be-installed door cover;

[0037] The model determination module is used to determine the target assembly model that matches the door cover information from multiple trained candidate assembly models.

[0038] The gap surface difference determination module is used to determine the first gap surface difference of the door cover to be installed at the initial assembly position when controlling the door cover to be installed to move to the initial assembly position; the training samples of the target assembly model include the sample gap surface differences of the sample door cover represented by the door cover information at multiple sample assembly positions respectively; the sample assembly position is determined based on the initial assembly position.

[0039] The installation and adjustment position update module is used to input the first gap surface difference into the target installation and adjustment model, and determine the updated installation and adjustment position of the door cover to be installed based on the output of the target installation and adjustment model;

[0040] A door cover installation module is used to install the door cover to be installed at the update and adjustment position if the second gap surface difference of the door cover to be installed at the update and adjustment position meets the installation conditions.

[0041] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0042] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0043] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method.

[0044] The aforementioned automotive door cover installation control method, device, computer equipment, storage medium, and computer program product, when performing automated installation control of the door cover to be installed, can first determine a target assembly and adjustment model that matches the door cover information from multiple trained candidate assembly and adjustment models based on the door cover information to be installed. Since the training samples of the target assembly and adjustment model include the sample gap surface differences corresponding to the sample door cover at multiple sample assembly and adjustment positions represented by the door cover information, and the sample assembly and adjustment positions are determined based on the initial assembly and adjustment positions, the door cover to be installed can be controlled to move to the initial assembly and adjustment position, the first gap surface difference of the door cover to be installed at the initial assembly and adjustment position can be determined, and the first gap surface difference can be input into the target assembly and adjustment model. The target assembly and adjustment model can directly determine the assembly and adjustment information required for door cover assembly and adjustment based on the first gap surface difference. According to the output of the target assembly and adjustment model, the updated assembly and adjustment position of the door cover to be installed can be determined. If the second gap surface difference of the door cover to be installed at the updated assembly and adjustment position meets the installation conditions, the door cover to be installed is installed at the updated assembly and adjustment position. By pre-training a target assembly and adjustment model that is compatible with the door cover to be installed, the gap and surface difference of all test points can be considered and balanced during automatic adjustment. The updated assembly and adjustment position of the door cover to be installed can be directly determined without the need for staged operations. This effectively improves the accuracy and success rate of assembly and adjustment during the installation process, reduces the configuration workload and complexity of the assembly and adjustment process, and thus improves the efficiency of door cover installation. Attached Figure Description

[0045] Figure 1 This is an application environment diagram of a car door cover installation control method in one embodiment;

[0046] Figure 2 This is a flowchart illustrating a method for controlling the installation of a car door cover in one embodiment;

[0047] Figure 3 This is a flowchart illustrating the automotive door cover installation control method in another embodiment;

[0048] Figure 4 This is a flowchart illustrating the process of determining the assembly and adjustment positions of multiple samples based on the initial assembly and adjustment position and the preset training sample expansion method in one embodiment.

[0049] Figure 5 This is a flowchart illustrating the automotive door cover installation control method in another embodiment;

[0050] Figure 6 This is a flowchart illustrating the automotive door cover installation control method in another embodiment;

[0051] Figure 7 This is a structural block diagram of a car door cover mounting control device in one embodiment;

[0052] Figure 8This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0054] The automotive door cover installation control method provided in this application embodiment can be applied to, for example... Figure 1 In the application environment shown, the control platform 102 communicates with the server 104 and the control terminal 106 via a network. A data storage system can store the data that the control platform 102 needs to process. The data storage system can be integrated into the control platform 102 or placed in the cloud or on another network server. When the door cover to be installed needs to be installed, the control terminal 106 can trigger an installation command for the door cover. In response to the installation command, the control platform 102 determines the door cover information, identifies a target assembly model that matches the door cover information from multiple trained candidate assembly models, and then controls the corresponding control terminal 106 to move the door cover to the initial assembly position. The first gap surface difference of the door cover at the initial assembly position is determined. The training samples of the target assembly model include the sample gap surface differences corresponding to multiple sample assembly positions represented by the door cover information. The sample assembly positions are determined based on the initial assembly position. The first gap surface difference is input into the target assembly and adjustment model. Based on the output of the target assembly and adjustment model, the updated assembly and adjustment position of the door cover to be installed is determined. If the second gap surface difference of the door cover to be installed at the updated assembly and adjustment position meets the installation conditions, the corresponding control terminal 106 is controlled to install the door cover to be installed at the updated assembly and adjustment position.

[0055] The control platform 102 is a software platform used to perform control logic operations, generate and issue control commands. The control platform 102 can perform control logic operations or processing based on collected or acquired information, generating corresponding control commands and issuing them to the corresponding control terminals 106 for processing. It is understood that the control platform 102 can be integrated into the user terminal of technicians or into the server 104. The user terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented using a standalone server or a server cluster composed of multiple servers.

[0056] The control terminal 106 can be a terminal device on the production line controlled by the control platform 102, actually performing assembly and installation operations. For example, on a door installation production line, the control terminal 106 can include, but is not limited to, mobile control devices and information acquisition devices. The mobile control device is a device capable of grasping and moving the door cover, such as a robot or intelligent robotic arm. The information acquisition device is a device that collects information about the position of the door cover relative to the vehicle body during installation; for example, the information acquisition device can include a camera or a distance sensor. The control platform 102 can generate corresponding control commands based on the actual control situation and send them to the corresponding control terminal 106, controlling the control terminal 106 based on the control commands.

[0057] In one embodiment, such as Figure 2 As shown, a method for controlling the installation of an automotive door cover is provided, which is applied to... Figure 1 Taking the control platform 102 as an example, the following steps are included:

[0058] S202, in response to an installation command for a door cover to be installed, determines the door cover information for the door cover to be installed.

[0059] The door cover to be installed refers to the door cover component that needs to be installed on the vehicle body, such as a car door, hood, or trunk lid. The installation command for the door cover to be installed is a signal used by the control platform to instruct the installation of the door cover onto the vehicle body. This command can be triggered by a technician via a user terminal or by a custom event. For example, when the door cover to be installed is transported to the installation station, the technician can select the door cover via the user terminal and click the installation control to trigger the generation of the installation command. Alternatively, the technician can pre-set installation events for various vehicle body components in the control platform's program based on production task information. These installation events can be time-based events, such as specifying the exact time the door cover to be installed, thus triggering the installation command for that door cover.

[0060] The door cover information is information data used to characterize the configuration of the door cover to be installed. For example, the door cover information may include, but is not limited to, the vehicle model to which the door cover to be installed belongs, the door cover location identifier in the vehicle model, and the door cover specification parameters.

[0061] Specifically, the control platform responds to the installation command for the door cover to be installed and determines the door cover information of the door cover to be installed.

[0062] In some embodiments, the installation instructions carry door cover information of the door cover to be installed, and the control platform can directly obtain the door cover information of the door cover to be installed according to the installation instructions.

[0063] In some embodiments, the installation instructions carry a door cover identifier for the door cover to be installed, and the control platform can obtain the door cover information of the door cover to be installed from the server based on the door cover identifier.

[0064] S204. From the multiple trained candidate assembly and adjustment models, determine the target assembly and adjustment model that is compatible with the door cover information.

[0065] The candidate assembly / adjustment model is a model obtained after training on sample doors corresponding to different door cover information. Designers can pre-train on sample doors corresponding to different door cover information, and then train the initial assembly / adjustment model based on the training samples collected during the assembly / adjustment training to obtain candidate assembly / adjustment models corresponding to the door cover information. Understandably, the specific number of candidate assembly / adjustment models can be determined based on the number of door cover information types that the control platform can support installing. The same door cover information can be classified into the same category, and the initial state model can be any deep learning model.

[0066] In some embodiments, the candidate assembly model can be trained by the designer on the control platform.

[0067] In some embodiments, to improve the versatility of the control platform, candidate assembly and adjustment models can also be obtained by technicians uploading them to the control platform via user terminals. When a user needs to use the control platform to control the installation of a door cover corresponding to newly added door cover information, the technician can upload the candidate assembly and adjustment model corresponding to the newly added door cover information to the control platform for the platform to call upon at any time.

[0068] Specifically, after obtaining the door cover information to be installed, the control platform can determine the target assembly model that matches the door cover information from multiple trained candidate assembly models based on the door cover information.

[0069] In some embodiments, the door cover information may include the door cover identifier of the door cover to be installed. The control platform is pre-configured with the correspondence between each door cover identifier and each candidate assembly and adjustment model. After obtaining the door cover information of the door cover to be installed, the correspondence between each door cover identifier and each candidate assembly and adjustment model can be found based on the door cover identifier in the door cover information, and the target assembly and adjustment model that matches the door cover identifier can be determined from multiple candidate assembly and adjustment models.

[0070] S206, while controlling the door cover to be installed to move to the initial installation position, determine the first gap surface difference of the door cover to be installed at the initial installation position.

[0071] The training samples for the target assembly and adjustment model include the sample gap surface differences corresponding to multiple sample assembly and adjustment positions, represented by the door cover information. The sample assembly and adjustment positions are determined based on the initial assembly and adjustment positions.

[0072] The sample door cover can be understood as the standard door cover corresponding to the door cover information. In actual production, due to differences in processes, even if the door cover information is exactly the same, there will be some slight differences between the door cover to be installed and the standard door cover. Therefore, even if a fixed installation trajectory is obtained by simulating the sample door cover, when the door cover to be installed is controlled by the fixed installation trajectory, the gap and surface difference between the door cover and the vehicle body will still be uneven, which will affect the overall aesthetics of the vehicle.

[0073] Gap and surface difference are information data used to characterize the gap and surface difference between the hood and the vehicle body. Gap refers to the deviation of the minimum distance between the hood and the vehicle body, while surface difference refers to the difference in height between the hood and the vehicle body. Gap and surface difference determine the flatness and aesthetics of the vehicle body's appearance. Sample gap and surface difference is the gap and surface difference determined when the sample hood is moved to the sample assembly position. The first gap and surface difference is the gap and surface difference determined when the hood to be installed is moved to the initial assembly position.

[0074] In some embodiments, the control platform can directly obtain the gap surface difference uploaded by the technician. For example, the technician can determine the gap surface difference of the door cover to be installed through the gap surface difference determination system, and then upload the determined gap surface difference to the control platform to provide data support for the control process of the control platform.

[0075] In some embodiments, the control platform can collect the gap surface difference of the door cover when it is in the assembly / adjustment position using a pre-configured information acquisition device. Specifically, after determining that the door cover has moved to the assembly / adjustment position, the control platform can control the information acquisition device to collect point cloud information of the door cover when it is in the assembly / adjustment position, and perform information analysis and recognition based on the point cloud information to determine the gap surface difference of the door cover when it is in the assembly / adjustment position. It is understood that the door cover may include a door cover to be installed or a sample door cover, and the assembly / adjustment position may include an initial assembly / adjustment position, an updated assembly / adjustment position, or a sample assembly / adjustment position.

[0076] The "assembly and adjustment position" refers to the adjustment position used during the installation and control of the door cover. During the assembly and adjustment process, by determining the gap and surface difference of the door cover in the adjustment position, the relative position of the door cover and the vehicle body can be understood, thus providing a data basis for whether subsequent door cover installation is feasible. The sample assembly and adjustment position refers to the training assembly and adjustment positions determined during training, while the initial assembly and adjustment position is the assembly and adjustment position of the sample door cover in an ideal assembly and adjustment state. Understandably, the initial assembly and adjustment position can be determined by technicians based on experimental data and uploaded to the control platform. Each door cover has its own corresponding initial assembly and adjustment position.

[0077] In some embodiments, the installation position may refer to the position of the door cover to be installed during the installation process. Since the movement of the door cover to be installed is completed with the assistance of a corresponding mobile control terminal, such as a robot, during the automatic installation process, the initial installation position of the door cover to be installed corresponds one-to-one with the terminal position of the mobile control terminal. For example, the position coordinates of the initial installation position of the door cover to be installed correspond one-to-one with the robot coordinates of the gripping robot. If you want to control the door cover to be installed to move to the initial installation position, the control platform only needs to generate a control command based on the terminal position information of the mobile control terminal corresponding to the initial installation position to instruct the mobile control terminal to move, thereby moving the door cover to be installed to the corresponding initial installation position.

[0078] Specifically, the control platform can determine the initial installation position matching the door cover information based on the door cover information to be installed, and then instruct the corresponding control terminal to move the door cover to be installed to the initial installation position. When it is determined that the door cover to be installed has been moved to the initial installation position, the control platform can determine the first gap surface difference of the door cover to be installed at the initial installation position.

[0079] S208, input the first gap surface difference into the target assembly and adjustment model, and determine the update and adjustment position of the door cover to be installed based on the output of the target assembly and adjustment model.

[0080] The updated installation location refers to the theoretical installation location determined after the analysis of the target installation model. Installing the door cover to be installed at the updated installation location can theoretically meet the user's installation requirements, that is, meet the installation conditions.

[0081] Specifically, after obtaining the first gap surface difference, the controller can input it into the target assembly and adjustment model. The target assembly and adjustment model can then analyze the difference between the first gap surface difference and the ideal gap surface difference of the door cover to be installed, determining the adjustment information required to adjust the first gap surface difference to the theoretical gap surface difference. The control platform can acquire the output of the target assembly and adjustment model and, based on this output, determine the updated assembly and adjustment position of the door cover to be installed.

[0082] In some embodiments, the output of the target assembly model is the position information of the mobile control terminal corresponding to the updated assembly position of the door cover to be installed. For example, it can be the robot position coordinates corresponding to the updated assembly position.

[0083] In some embodiments, the output of the target assembly and adjustment model can be the position difference information between the initial assembly and adjustment position and the updated assembly and adjustment position. After obtaining the position difference information, the control platform can determine the updated assembly and adjustment position of the door cover to be installed based on the position difference information and the initial assembly and adjustment position.

[0084] S210, if the second gap surface difference of the door cover to be installed at the replacement and adjustment position meets the installation conditions, then the door cover to be installed is installed at the replacement and adjustment position.

[0085] The installation conditions are preset conditions used to determine whether the door cover to be installed can be installed at the replacement and adjustment position. If the second gap surface difference of the door cover to be installed at the replacement and adjustment position meets the installation conditions, it means that installing the door cover to be installed at the replacement and adjustment position can meet the user's installation requirements.

[0086] In some embodiments, the installation conditions can be that the second gap surface difference is the same as the theoretical gap surface difference.

[0087] In some embodiments, the installation conditions can be such that the difference between the second gap surface difference and the theoretical gap surface difference conforms to a preset allowable difference range. For example, the difference in gap surface difference is within ±0.5 mm.

[0088] Specifically, after determining the replacement and adjustment position of the door cover to be installed, the control platform can determine the second gap surface difference of the door cover to be installed at the replacement and adjustment position. Then, it calls the preset installation conditions to determine whether the second gap surface difference meets the installation conditions. If the second gap surface difference meets the installation conditions, the door cover to be installed is installed at the replacement and adjustment position.

[0089] In the above embodiments, when performing automated installation control on the door cover to be installed, a target installation model that matches the door cover information can be determined from multiple trained candidate installation models based on the door cover information. Since the training samples of the target installation model include the sample gap surface differences corresponding to the sample door cover at multiple sample installation positions as represented by the door cover information, and the sample installation positions are determined based on the initial installation positions, the door cover to be installed can be controlled to move to the initial installation positions, the first gap surface difference of the door cover to be installed at the initial installation positions can be determined, and the first gap surface difference can be input into the target installation model. The target installation model can directly determine the installation information required for door cover installation based on the first gap surface difference. According to the output of the target installation model, the updated installation position of the door cover to be installed can be determined. If the second gap surface difference of the door cover to be installed at the updated installation positions meets the installation conditions, the door cover to be installed is installed at the updated installation positions. By pre-training a target assembly and adjustment model that is compatible with the door cover to be installed, the gap and surface difference of all test points can be considered and balanced during automatic adjustment. The updated assembly and adjustment position of the door cover to be installed can be directly determined without the need for staged operations. This effectively improves the accuracy and success rate of assembly and adjustment during the installation process, reduces the configuration workload and complexity of the assembly and adjustment process, and thus improves the efficiency of door cover installation.

[0090] In other embodiments, the automotive door cover installation control method further includes: if the second gap surface difference of the door cover to be installed does not meet the installation conditions, then the first gap surface difference is updated using the second gap surface difference to obtain an updated first gap surface difference. Based on the updated first gap surface difference, the step of inputting the first gap surface difference into the target assembly model is returned to be executed.

[0091] If the second gap surface difference does not meet the installation conditions, it can be considered that installing the door cover to be installed at the replacement and adjustment position will not meet the user's installation needs. For example, there is a difference between the second gap surface difference and the theoretical gap surface difference, or the degree of difference between the second gap surface difference and the theoretical gap surface difference exceeds the allowable difference range.

[0092] Specifically, if the second gap surface difference of the door cover to be installed does not meet the installation conditions, it means that although the theoretically acceptable update and adjustment position that can meet the user's installation requirements has been determined, in the actual control process, due to control errors or other influencing factors, the current adjustment process has not achieved the expected results. In order to meet the user's installation requirements, the control platform can use the second gap surface difference to update the first gap surface difference to obtain the updated first gap surface difference. Then, based on the updated first gap surface difference, it returns to execute the step of inputting the first gap surface difference into the target adjustment model until the second gap surface difference meets the installation conditions.

[0093] In the above embodiments, if the installation and adjustment fails once, the control platform can perform installation and adjustment control on the door cover to be installed again by means of repeated installation and adjustment, until it is determined that the door cover to be installed can meet the user's installation requirements when installed in the updated installation and adjustment position, and then the door cover is installed. This can effectively reduce the situation where the installation and adjustment accuracy is reduced due to the influence of control deviation or other influencing factors.

[0094] In some embodiments, to avoid extreme situations, such as program or hardware failures causing the door cover to be installed to continuously fail to meet user needs in its update and adjustment position, designers can pre-set a threshold for the number of cycles. Before each installation and adjustment cycle is executed, the control platform first determines whether the number of installation and adjustment cycles for the door cover to be installed has reached the threshold. If it has, it is determined that there is an installation abnormality in the door cover to be installed, and an abnormality prompt message can be generated to notify technicians to investigate the abnormality.

[0095] In the automotive industry, different car models have different body contours, and the door cover installation requirements are also different. Even for the same car model, the door cover installation requirements may differ depending on the location of the door cover. For example, the installation requirements for the left and right doors are not necessarily the same as the installation requirements for the hood or trunk lid.

[0096] Based on this, in some embodiments, the door cover information includes the vehicle model to which the door cover to be installed belongs and the door cover position identifier within the vehicle model. S204, determining a target assembly / adjustment model adapted to the door cover information from a plurality of trained candidate assembly / adjustment models includes: determining a plurality of selected assembly / adjustment models adapted to the vehicle model from a plurality of candidate assembly / adjustment models based on the vehicle model to which the door cover to be installed belongs; and determining a target assembly / adjustment model adapted to the door cover position identifier from a plurality of selected assembly / adjustment models based on the door cover position identifier within the vehicle model.

[0097] The vehicle model to which the door cover to be installed belongs refers to the vehicle model information. As we understand it, vehicle models are categorized based on factors such as structure, purpose, function, and brand. Different vehicle models have different door cover installation requirements; each model has its own specific installation requirements. For example, regarding the surface difference of the trunk lid, some models have a 0mm installation requirement, while others require 0.5mm, and so on.

[0098] Door cover position markings are information data used to indicate the corresponding body mounting position of the door cover to be installed in the vehicle model. The body mounting position can include the left front door, right front door, left rear door, right rear door, hood, trunk lid, etc.

[0099] Among them, the selected assembly and adjustment model is the candidate assembly and adjustment model corresponding to the vehicle model. It can be understood that each vehicle model can have multiple selected assembly and adjustment models, and the number of these models is equal to the number of door covers that the vehicle model needs to install.

[0100] Specifically, the control platform can determine multiple selected assembly and adjustment models that are compatible with the vehicle model to which the door cover to be installed belongs from a number of candidate assembly and adjustment models. Then, based on the door cover position identifier in the vehicle model, the target assembly and adjustment model that matches the door cover position identifier is determined from the multiple selected assembly and adjustment models.

[0101] In some embodiments, the control platform pre-stores a first mapping relationship to characterize the correspondence between each vehicle model and each candidate assembly and adjustment model. The control platform can find the first mapping relationship according to the vehicle model to which the door cover to be installed belongs, thereby determining multiple selected assembly and adjustment models that are compatible with the vehicle model. Each selected assembly and adjustment model corresponds to a different door cover position identifier.

[0102] In some embodiments, the control platform pre-stores a second mapping relationship to characterize the correspondence between each door cover position identifier and each selected assembly and adjustment model. After determining multiple selected assembly and adjustment models, the control platform can search for the second mapping relationship based on the door cover position identifier of the door cover to be installed, thereby determining the target assembly and adjustment model that matches the door cover position identifier of the door cover to be installed.

[0103] In the above embodiments, by pre-training corresponding assembly and adjustment models for different vehicle models and door covers in different locations within different vehicle models, the universality of the control platform in door cover installation scenarios and the accuracy of door cover installation control for different door covers can be effectively improved.

[0104] In the entire scheme of automotive door cover installation control, the target assembly and adjustment model has a significant impact on the assembly and adjustment accuracy, and the training process of the target assembly and adjustment model will directly affect the actual use of the target assembly and adjustment model. Based on this, the training process of the target assembly and adjustment model will be described below through several examples.

[0105] In some embodiments, such as Figure 3 As shown, the automotive door cover installation control method also includes:

[0106] S302, for each sample door cover, obtain the initial assembly position and initial assembly model of the sample door cover.

[0107] The initial assembly position refers to the assembly position of the sample door cover in its ideal assembly state. The initial assembly model is an untrained deep learning model, which can be pre-configured by the designer in the control platform.

[0108] Specifically, during model training, designers prepare corresponding sample door covers and corresponding sample vehicle bodies for each door cover location in each vehicle model. For each sample door cover, the control platform can obtain the initial assembly position and initial assembly model of the sample door cover.

[0109] S304. Based on the initial setup position and the preset training sample expansion method, determine multiple sample setup positions.

[0110] Among them, the training sample expansion method is a sample generation method used to generate multiple sample assembly positions based on the initial assembly position. It is understandable that designers can write training sample expansion methods according to actual sample generation needs and set the training sample expansion methods in the control platform so that the control platform can generate training samples based on an initial assembly position.

[0111] In some embodiments, the training sample expansion method can be random expansion, that is, taking the initial assembly position as the center, randomly selecting assembly positions from the surrounding areas as sample assembly positions.

[0112] In some embodiments, to reduce training complexity and resource consumption, the training sample expansion method may involve randomly selecting a first number of sample adjustment positions within a preset spatial range centered on the initial adjustment position, and randomly selecting a second number of sample adjustment positions outside the preset spatial range centered on the initial adjustment position, wherein the first number is greater than the second number. By generating more sample adjustment positions in regions closer to the initial adjustment position and fewer sample adjustment positions in regions farther from the initial adjustment position, it is possible to reduce training complexity and resource consumption during the training process while ensuring that fine adjustments around the initial adjustment position are trained and that training points farther from the initial adjustment position do not become discrete during adjustment.

[0113] S306 controls the sample door cover to move to each sample loading and unloading position, and determines the sample gap surface difference of the sample door cover at each sample loading and unloading position.

[0114] Specifically, the control platform can control the sample door cover to move to each sample assembly position. For each sample assembly position, the control platform will determine the sample gap surface difference of the sample door cover at the sample assembly position, and obtain the sample gap surface difference of the sample door cover at each sample assembly position.

[0115] Using the mobile control terminal as the robot, the sample assembly positions are {X1, X2, ..., X...} i For example, the control platform can control the robot to move to the corresponding robot coordinate position of each sample assembly position. Then, it controls the information acquisition device to collect the point cloud information of the sample door cover and the sample body, obtaining the point cloud information corresponding to each sample assembly position. Based on the point cloud information, the surface differences {Y1, Y2, ..., Y} between the sample door cover and the sample body are calculated. i}

[0116] S308, based on the surface difference between the gaps between each sample and the assembly position of each sample, determine the training samples for the sample door cover.

[0117] Specifically, the control platform can bind the surface difference between each sample gap to its corresponding sample assembly position to obtain multiple sets of samples, and then determine these multiple sets of samples as training samples for the sample door cover.

[0118] S310, The initial assembly and adjustment model is trained based on the training samples to obtain the target assembly and adjustment model after training.

[0119] Specifically, after obtaining the training samples of the sample door cover, the control platform can train the initial assembly model based on the training samples and the preset model training method, iteratively optimizing the model parameters in the initial assembly model to obtain the target assembly model after training. It is understandable that the preset model training method can be any deep learning model training method.

[0120] In some embodiments, the preset model training method can be stochastic gradient descent. By using stochastic gradient descent for model training, the influence of noise during training can be effectively suppressed, thereby improving the accuracy of model training.

[0121] In the above embodiments, by generating training samples for the sample door cover based on the initial assembly position and the preset training sample expansion method, the number of samples during model training can be effectively increased, thereby improving the training accuracy of the target assembly model and providing an accurate assembly tool for the subsequent actual installation process of the door cover.

[0122] Furthermore, in some embodiments, such as Figure 4 As shown in S304, based on the initial setup position and the preset training sample augmentation method, multiple sample setup positions are determined, including:

[0123] S402, with the initial assembly position as the center position, divides the sample generation space for the sample door cover according to the preset sample expansion range.

[0124] The preset sample expansion range is a range parameter used to limit the size of the sample generation space, which can be preset by the designer according to the actual training needs. The preset sample expansion range may include the axis length thresholds on each spatial axis with the initial assembly position as the center.

[0125] Specifically, the control platform can call a preset sample expansion range, take the initial assembly position as the center position, and divide the sample generation space for the sample door cover according to the preset sample expansion range.

[0126] In some embodiments, the sample generation space is a six-dimensional space, consisting of three translation dimensions and three selection dimensions.

[0127] S404, based on the preset anchor point setting method, sets multiple anchor points on each spatial axis of the sample generation space.

[0128] Anchor points are reference points used to generate the assembly and adjustment positions of each sample. Anchor points are related to the initial assembly and adjustment positions, and each sample assembly and adjustment position is related to each anchor point. That is, by setting anchor points, the assembly and adjustment positions of each sample can be associated with the initial assembly and adjustment positions, thus avoiding the sample assembly and adjustment positions from being discrete during assembly and adjustment.

[0129] The preset anchor point setting method is a preset method used to set anchor points in the sample generation space. The preset anchor point setting method can be generated by the designer according to the actual needs. For example, the preset anchor point setting method can be the axis evenly divided anchor point setting method. Taking the sample generation space as a six-dimensional space as an example, the control platform can set an anchor point at one-third and two-thirds of each spatial axis, for a total of 32 anchor points.

[0130] Specifically, the control platform can set up multiple anchor points on each spatial axis of the sample generation space based on a preset anchor point setting method.

[0131] S406, each anchor point and the initial assembly position are determined as the sample generation reference point for the sample door cover.

[0132] Specifically, the control platform can uniformly determine the identified anchor points and initial assembly positions as the sample generation reference points for the sample door cover.

[0133] For example, after the control platform sets an anchor point at one-third and two-thirds of each spatial axis, these 32 anchor points and the initial assembly position can be used as the sample generation reference points. Thus, the sample door cover includes 33 sample generation reference points.

[0134] S408 generates a reference point for each sample, and generates multiple sample assembly positions based on a Gaussian distribution scheme with the sample generation reference point as the center point.

[0135] The Gaussian distribution scheme refers to a point generation scheme that uses the Gaussian distribution to generate sample assembly locations. Its principle is based on the probability density function of the Gaussian distribution to generate a series of random points. The Gaussian distribution is a continuous probability distribution with a bell-shaped shape, determined by two parameters: the mean and the standard deviation. The mean represents the central location of the distribution, while the standard deviation represents the dispersion of the data.

[0136] Specifically, for each sample, a reference point is generated, and multiple sample assembly positions are randomly generated based on a Gaussian distribution scheme, with the reference point as the center point.

[0137] Taking the case where there are 33 sample generation reference points on the sample door cover as an example, the control platform can randomly use one of the sample generation reference points as the center and use a Gaussian distribution to generate a certain sample assembly position each time.

[0138] In the above embodiments, multiple anchor points related to the initial assembly position are first set on each spatial axis in the sample generation space. Then, multiple sample assembly positions are generated based on each anchor point and the initial assembly position using a Gaussian distribution scheme. By generating multiple sample assembly positions based on the initial assembly position using this mixed Gaussian distribution scheme, it is ensured that the fine adjustments around the initial assembly position can be trained. It also ensures that the boundary points of the training space will not be discrete during assembly due to being too far away from the training points. This effectively improves the sample sampling rationality of the target assembly model and thus improves the training accuracy of the model.

[0139] In order to check the final installation effect, in some embodiments, such as Figure 5 As shown, the automotive door cover installation control method also includes the following steps:

[0140] S502, in response to an installation completion event for a door cover to be installed, determines the installation surface difference of the installed door cover and the flattening point surface difference of the installed door cover when it is in a flattening state.

[0141] The installation completion event is triggered after the door cover to be installed has been successfully installed. For example, after all the screws on the door cover to be installed are tightened, the technician can confirm that the door cover to be installed is complete and trigger the installation completion event based on the user terminal. It can be understood that once the door cover to be installed is complete, it can be called an installed door cover; that is, the installed door cover and the door cover to be installed are the same door cover in different states. The installation surface difference of the installed door cover refers to the surface difference information between the installed door cover and the vehicle body.

[0142] The "leveling state" refers to the surface difference between the installed door cover and the vehicle body during the leveling process. Understandably, after the door cover is installed, the corresponding mobile control terminal will release control of the door cover; for example, the robot will release its gripper. At this time, the door cover may differ from its original factory position due to changes in the vehicle door's position or its natural drooping without support. Directly re-measuring this discrepancy, i.e., checking the installation effect, can easily lead to measurement errors. Therefore, to ensure that the door cover position during re-measuring is consistent with or approximately consistent with its original factory position, the control platform needs to perform a leveling process on the installed door cover. This involves using a leveling device, such as a cylinder, to pull the door cover to its original factory position or a position close to it before measurement.

[0143] The leveling point is the measurement point corresponding to the ideal surface difference between the door hood and the body in their factory condition, as desired by the user. The leveling point surface difference represents the ideal surface difference between the door hood and the body in their factory condition.

[0144] Specifically, in response to an installation completion event for a door cover to be installed, the control platform can control the information acquisition device to collect information from the installed door cover, determine the installation surface difference of the installed door cover, and simultaneously acquire the leveling point surface difference of the installed door cover in the leveling state. Understandably, the leveling point surface difference can be pre-set by the designer according to the user's actual factory requirements and stored in the control platform.

[0145] S504. Based on the difference between the mounting surface difference and the leveling point surface difference, and the corrected rotation dimension of the installed door cover, determine the corrected installation position of the installed door cover.

[0146] The corrected rotation dimension of the installed door cover refers to the dimension that the installed door cover can rotate during the retesting process. After installation, the door cover cannot be rotated in any dimension as it was before installation; instead, it can only rotate within the corresponding rotatable dimension during normal use. For example, if the door cover is for one of the left or right vehicle doors, its rotatable dimension is the same as the rotation dimension when the door is normally opened and closed.

[0147] When correcting the installation position to compensate for discrepancies, the control platform needs to control the target installation position of the installed door cover in the corrected rotation dimension.

[0148] Specifically, after determining the installation surface difference and leveling point surface difference of the installed door cover, the control platform can first compare the differences between the installation surface difference and the leveling point surface difference to determine the difference information between the installation surface difference and the leveling point surface difference. Based on the difference information between the installation surface difference and the leveling point surface difference, as well as the correction rotation dimension of the installed door cover, the correction adjustment position of the installed door cover can be determined.

[0149] S506, Adjusting and correcting the installation of an installed door cover based on the corrected installation position.

[0150] Specifically, the control platform can control the rotation of the installed door cover to the corresponding correction rotation dimension, move to the correction and adjustment position, and complete the installation and adjustment correction of the installed door cover.

[0151] S508. Based on the third gap surface difference of the installed door cover after the installation and adjustment correction is completed, determine the installation and adjustment retest results of the installed door cover.

[0152] Specifically, after completing the installation and adjustment corrections of the installed door cover, the control platform can determine the third gap surface difference of the installed door cover after the installation and adjustment corrections are completed. The specific method for determining the gap surface difference has been explained in detail above and will not be repeated here. After obtaining the third gap surface difference, the control platform can determine the installation and adjustment retest results of the installed door cover based on the third gap surface difference.

[0153] In some embodiments, the control platform can directly determine the third gap surface difference as the installation and adjustment retest result of the installed door cover, providing a data basis for the subsequent installation and adjustment optimization process of the control platform.

[0154] In some embodiments, the control platform can determine the difference between the third gap surface difference and the flattening point surface difference, evaluate the effect of the door cover installation process based on the difference between the third gap surface difference and the flattening point surface difference, and obtain the installation and adjustment retest results of the installed door cover.

[0155] In the above embodiments, by retesting the installed door cover after installation, the installation effect of the installed door cover can be effectively fed back, providing a data basis for optimizing the subsequent installation process.

[0156] In some embodiments, such as Figure 6 As shown, a method for controlling the installation of an automotive door cover is provided. This method can be summarized into three parts: a model training part, an assembly and adjustment part, and a retesting part. The training part involves learning the relationship between the current gap surface difference and the required robot movement value using specific data to obtain an assembly and adjustment model, which only needs to be run once. The assembly and adjustment part involves using the trained assembly and adjustment model to install the door cover during daily production. The retesting part involves performing a functional check on the gap surface differences at various points after the door cover is installed but before placing it on the vehicle, thereby providing feedback on the final installation effect. The following will explain these three parts in detail:

[0157] The first part is the model training.

[0158] The machines required for model training include a controllable industrial assembly robot, a camera mounted on its mechanical structure for capturing gap and surface differences, and a connected industrial control computer, as well as a sample vehicle and a sample door cover for user training models.

[0159] During actual training, technicians can manually select a well-adjusted gap surface difference position as the teaching point X0, also known as the initial assembly and adjustment position. At this teaching point X0, the gap surface difference between the sample door cover and the sample vehicle at all points reaches the ideal value Y0.

[0160] Based on the teaching point X0 determined by the technicians, the control platform generates a certain number of random training points near the teaching point according to a preset Gaussian mixture distribution scheme. The Gaussian mixture distribution scheme refers to dividing a six-dimensional sample generation space centered on the teaching point X0 according to a preset sample expansion range. Anchor points are then established at one-third and two-thirds of each spatial axis within this space, resulting in a total of 2 × 6 = 32 anchor points. Adding the central teaching point X0, this creates a total of 33 sample generation reference points. The control platform can randomly generate a certain number of training points (i.e., multiple sample loading positions) centered on one of these reference points using a Gaussian distribution each time.

[0161] After obtaining multiple training points, the control platform will control the robot to grasp the sample door cover and move it sequentially to each training point {X1, X2, ..., X...}. i}, and take point cloud images at each training point, and calculate the surface difference values ​​{Y1, Y2, ..., Y} corresponding to each gap in the point cloud image. i It should be noted that X i It is a six-dimensional coordinate value that can be mapped to robot coordinates, while Y... i It is an N-dimensional vector, where N equals the number of all required gaps and surface differences, and i is any natural number. The specific value of N is related to the number of cameras mounted on the robot. For example, when the robot is equipped with one camera, the corresponding Y... i It's a two-dimensional vector that can represent a gap value and a surface difference value. When two cameras are mounted on the robot, the corresponding Y-axis value is... i It is a four-dimensional vector that can represent two gap values ​​and two surface difference values.

[0162] The control platform can generate training samples based on the obtained gap surface differences and training points. Based on the training samples and the stochastic gradient descent method, the servo model F is trained. The conversion relationship between the gap surface differences at each location and the gap surface differences at the teaching position and the robot displacement is obtained, i.e., X-X0=F(Y-Y0), thus obtaining the assembly and adjustment model corresponding to the sample door cover.

[0163] Next comes the assembly and adjustment process.

[0164] In response to the installation command for the door cover to be installed, the control platform determines the door cover information and selects the target assembly model that matches the door cover information from multiple trained candidate assembly models. Subsequently, the control platform can move the door cover to be installed to the teaching point X0, collect point cloud information at teaching point X0, and determine the first gap surface difference Y of the door cover to be installed based on the point cloud information. k Due to changes in the manufacturing process of the body and hood, the original teaching position X0 may not be the optimal matching position, but it is not far from the optimal matching position.

[0165] The control platform will control the first gap surface difference Y k The input is given to the target assembly and adjustment model, which can be based on the first gap surface difference Y. k Calculate the robot's adjustment displacement to obtain the updated adjustment position X. k+1 , that is, X k+1 =X k -F(Y k -Y0). The control platform controls the robot to move the door cover to be installed to the replacement and adjustment position X. k+1 Subsequently, point cloud information was collected to determine the second gap surface difference Y of the door cover to be installed. k+1 .

[0166] The control platform will control the second gap surface difference Y k+1 Comparing the two with the ideal gap surface difference Y0, if the difference is within a preset range, such as within ±0.5mm, then the second gap surface difference Y of the door cover to be installed at the replacement and adjustment position can be determined. k+1 The installation conditions are met, and the control platform is in the updated installation and adjustment position X. k+1 Install the door cover to be installed.

[0167] If the difference between the two is not within the preset difference range, then the second gap surface difference Y of the door cover to be installed at the replacement and adjustment position can be determined. k+1 If the installation conditions are not met, the control platform can use the second clearance surface difference Y. k+1 For the first gap surface difference Y k The update is performed to obtain the updated first gap surface difference Y. k Y at this time k =Y k+1 The control platform will be based on the updated first gap surface difference Y. k Return to execution and change the first gap surface difference Y k The steps for inputting data into the target assembly model continue until the second gap surface difference Y is reached. k+1 The installation conditions are met.

[0168] Finally, there is the retesting section.

[0169] The control platform can respond to the installation completion event of a door cover to be installed, determine the installation surface difference of the installed door cover, and the surface difference of the leveling point when the installed door cover is in the leveling state. Based on the difference between the installation surface difference and the leveling point surface difference, and the corrected rotation dimension of the installed door cover, the corrected installation position of the installed door cover is determined, and the installed door cover is adjusted and corrected based on the corrected installation position. Specifically, the control platform can run the installation and adjustment iterative process once. k+1 =X K -F(Y k-Y0), but only moved one rotational dimension to simulate the normal opening and closing of the door. Finally, based on the third gap surface difference of the installed door cover after the assembly and adjustment correction, the assembly and adjustment retest results of the installed door cover were determined.

[0170] The automotive door cover installation control method in this embodiment has the following advantages: First, it can simultaneously balance the balance of each measuring point to approximate the theoretical value, preventing the loss of some aspects due to the sequential relationship of the three-two-one method. Second, the entire process does not require camera calibration; the servo-trained model automatically learns the coefficient relationship between the gap surface difference value and the assembly and adjustment movement value, without manual intervention in the parameter values. Third, the process flow has a high degree of freedom; users can freely choose the assembly and adjustment process (assemble and adjust first, then embed; embed and then assemble and adjust; adjust the pose first, then embed, then adjust the translation, etc.), and the modular algorithm will execute the corresponding steps. Fourth, it can be used directly on different door covers by only modifying the number of gap surface difference measuring points required, and users can also freely choose the measuring points and movement degrees of freedom to be opened, facilitating use in multiple scenarios. Fifth, it has high robustness; multiple cameras can be mounted on the robot. If individual cameras malfunction, the parameters corresponding to those cameras will be automatically zeroed, allowing the production line to continue assembly and adjustment, enabling maintenance during production line downtime instead of requiring immediate line shutdown, although the effectiveness will be slightly affected. Sixth, the pace is fast, requiring only 2-3 moves to complete the installation and adjustment of the door cover.

[0171] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0172] Based on the same inventive concept, this application also provides an automotive door cover installation control device for implementing the aforementioned automotive door cover installation control method. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more embodiments of the automotive door cover installation control device provided below can be found in the limitations of the automotive door cover installation control method described above, and will not be repeated here.

[0173] In one embodiment, such as Figure 7As shown, an automotive door cover mounting control device 700 is provided, including: a command response module 701, a model determination module 702, a gap / surface difference determination module 703, an installation position update module 704, and a door cover mounting module 705, wherein:

[0174] The instruction response module 701 is used to respond to the installation instruction for the door cover to be installed and determine the door cover information of the door cover to be installed.

[0175] The model determination module 702 is used to determine the target assembly model that matches the door cover information from multiple trained candidate assembly models.

[0176] The gap surface difference determination module 703 is used to determine the first gap surface difference of the door cover to be installed at the initial assembly position while controlling the door cover to be installed to move to the initial assembly position; the training samples of the target assembly model include the sample gap surface differences of the sample door cover represented by the door cover information at multiple sample assembly positions respectively; the sample assembly position is determined based on the initial assembly position.

[0177] The installation position update module 704 is used to input the first gap surface difference into the target installation model and determine the updated installation position of the door cover to be installed based on the output of the target installation model.

[0178] The door cover mounting module 705 is used to install the door cover to be installed at the update and adjustment position if the second gap surface difference of the door cover to be installed at the update and adjustment position meets the installation conditions.

[0179] In some embodiments, the door cover information includes the vehicle model to which the door cover to be installed belongs and the door cover position identifier within the vehicle model. The model determination module 702 is configured to: determine, based on the vehicle model to which the door cover to be installed belongs, a plurality of selected assembly / adjustment models adapted to the vehicle model from a plurality of candidate assembly / adjustment models; and determine, based on the door cover position identifier within the vehicle model, a target assembly / adjustment model adapted to the door cover position identifier from the plurality of selected assembly / adjustment models.

[0180] In some embodiments, the vehicle door hood mounting control device further includes:

[0181] The initial information acquisition module is used to acquire the initial assembly position and initial assembly model of each sample door cover; the initial assembly position is the assembly position of the sample door cover in the ideal assembly state.

[0182] The sample expansion module is used to determine multiple sample assembly positions based on the initial assembly position and the preset training sample expansion method.

[0183] The sample gap surface difference determination module is used to control the sample door cover to move to each sample assembly position and determine the sample gap surface difference of the sample door cover at each sample assembly position.

[0184] The sample generation module is used to determine the training samples for the sample door cover based on the surface difference between each sample gap and the assembly position of each sample.

[0185] The model training module is used to train the initial assembly model based on training samples to obtain the target assembly model after training.

[0186] In some embodiments, the initial information acquisition module is used to: divide the sample generation space of the sample door cover into a sample generation space with the initial assembly position as the center position and according to the preset sample expansion range; set multiple anchor points on each spatial axis of the sample generation space based on the preset anchor point setting method; determine each anchor point and the initial assembly position as the sample generation reference point of the sample door cover; and for each sample generation reference point, generate multiple sample assembly positions with the sample generation reference point as the center point based on the Gaussian distribution scheme.

[0187] In some embodiments, the vehicle door hood mounting control device further includes:

[0188] The gap surface difference update module is used to update the first gap surface difference using the second gap surface difference if the second gap surface difference of the door cover to be installed does not meet the installation conditions, so as to obtain the updated first gap surface difference.

[0189] The loop module is used to return the updated first gap surface difference to the assembly position update module to perform the step of inputting the first gap surface difference into the target assembly model.

[0190] In some embodiments, the vehicle door hood mounting control device further includes:

[0191] The event response module is used to respond to the installation completion event of the door cover to be installed, determine the installation surface difference of the installed door cover, and the surface difference of the leveling point when the installed door cover is in the leveling state.

[0192] The corrected installation position determination module is used to determine the corrected installation position of the installed door cover based on the difference between the installation surface difference and the leveling point surface difference, as well as the corrected rotation dimension of the installed door cover.

[0193] The installation and adjustment correction module is used to perform installation and adjustment corrections on the installed door cover based on the correction installation and adjustment position.

[0194] The retest module is used to determine the retest results of the installed door cover based on the third gap surface difference of the installed door cover after the installation and adjustment correction is completed.

[0195] The various modules in the aforementioned automotive door cover mounting control device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0196] In one embodiment, a computer device is provided, which may be a terminal or server integrating a control platform, and its internal structure diagram may be as follows: Figure 8 As shown, the computer device includes a processor, memory, and network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores data such as door cover information, candidate assembly models, initial assembly positions, first gap surface differences, updated assembly positions, and second gap surface differences. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer program implements a method for controlling the installation of an automotive door cover.

[0197] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0198] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the specific steps of the above-described embodiment of the automobile door cover installation control method.

[0199] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the specific steps of the above-described embodiment of the automobile door cover installation control method.

[0200] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the specific steps of the above-described embodiment of the automobile door cover installation control method.

[0201] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the acquisition, storage, processing, and transmission of the data all comply with relevant laws and regulations.

[0202] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0203] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0204] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for controlling the installation of an automobile door cover, characterized in that, The method includes: In response to an installation command for a door cover to be installed, the door cover information of the door cover to be installed is determined; From the trained candidate assembly and adjustment models, determine the target assembly and adjustment model that is compatible with the door cover information; While controlling the door cover to be installed to move to the initial assembly position, the first gap surface difference of the door cover to be installed at the initial assembly position is determined; the training samples of the target assembly model include the sample gap surface differences of the sample door cover represented by the door cover information at multiple sample assembly positions respectively; the sample assembly position is determined based on the initial assembly position. The first gap surface difference is input into the target assembly and adjustment model, and the update and adjustment position of the door cover to be installed is determined based on the output of the target assembly and adjustment model. If the second gap surface difference of the door cover to be installed at the update and adjustment position meets the installation conditions, then the door cover to be installed is installed at the update and adjustment position.

2. The method according to claim 1, characterized in that, The door cover information includes the vehicle model to which the door cover to be installed belongs and the door cover position identifier of the door cover in the vehicle model; The step of determining the target assembly model that matches the door cover information from multiple trained candidate assembly models includes: Based on the vehicle model to which the door cover to be installed belongs, a number of selected assembly and adjustment models that are compatible with the vehicle model are determined from a number of candidate assembly and adjustment models; Based on the door cover position identifier in the vehicle model, a target assembly model that matches the door cover position identifier is determined from a plurality of selected assembly models.

3. The method according to claim 1, characterized in that, The method further includes: For each sample door cover, obtain the initial assembly and adjustment position and the initial assembly and adjustment model of the sample door cover; the initial assembly and adjustment position is the assembly and adjustment position of the sample door cover in the ideal assembly and adjustment state; Based on the initial setup position and the preset training sample expansion method, determine multiple sample setup positions; The sample door cover is moved to each of the sample loading and adjusting positions, and the sample gap surface difference of the sample door cover at each of the sample loading and adjusting positions is determined. Based on the surface difference between the gaps between the samples and the assembly and adjustment positions of the samples, the training samples for the sample door cover are determined; The initial assembly and adjustment model is trained based on the training samples to obtain the target assembly and adjustment model after training.

4. The method according to claim 3, characterized in that, The step of determining multiple sample assembly positions based on the initial assembly position and the preset training sample expansion method includes: Using the initial assembly position as the center, the sample generation space is divided for the sample door cover according to the preset sample expansion range; Based on the preset anchor point setting method, multiple anchor points are set on each spatial axis of the sample generation space; Each of the anchor points and the initial assembly position are determined as the sample generation reference point for the sample door cover; For each of the aforementioned sample generation reference points, multiple sample assembly and adjustment positions are generated based on a Gaussian distribution scheme, with the reference point as the center point.

5. The method according to any one of claims 1 to 4, characterized in that, The method further includes: If the second gap surface difference of the door cover to be installed does not meet the installation conditions, the first gap surface difference is updated using the second gap surface difference to obtain the updated first gap surface difference; Based on the updated first gap surface difference, return to the step of inputting the first gap surface difference into the target assembly model.

6. The method according to any one of claims 1 to 4, characterized in that, The method further includes: In response to the installation completion event of the door cover to be installed, the installation surface difference of the installed door cover and the flattening point surface difference of the installed door cover in the flattening state are determined; Based on the difference between the mounting surface difference and the leveling point surface difference, and the corrected rotation dimension of the installed door cover, the corrected installation position of the installed door cover is determined. The installed door cover is adjusted and corrected based on the corrected installation position; Based on the third gap surface difference of the installed door cover after the installation and adjustment correction is completed, the installation and adjustment retest result of the installed door cover is determined.

7. A car door cover mounting control device, characterized in that, The device includes: The instruction response module is used to determine the door cover information of the door cover to be installed in response to the installation instruction for the door cover to be installed; The model determination module is used to determine the target assembly model that matches the door cover information from multiple trained candidate assembly models. The gap surface difference determination module is used to determine the first gap surface difference of the door cover to be installed at the initial assembly position when controlling the door cover to be installed to move to the initial assembly position; the training samples of the target assembly model include the sample gap surface differences of the sample door cover represented by the door cover information at multiple sample assembly positions respectively; the sample assembly position is determined based on the initial assembly position. The installation and adjustment position update module is used to input the first gap surface difference into the target installation and adjustment model, and determine the updated installation and adjustment position of the door cover to be installed based on the output of the target installation and adjustment model; A door cover installation module is used to install the door cover to be installed at the update and adjustment position if the second gap surface difference of the door cover to be installed at the update and adjustment position meets the installation conditions.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images