An intelligent testing method and system for electric beds based on visual analysis

Abstract

The present invention discloses an intelligent testing method and system for electric beds based on visual analysis. The method comprises the following steps: in a first test cycle, using a machine vision module to perform spatial positioning and displacement trajectory analysis on multiple parts of the electric bed, generate displacement monitoring data, and evaluate deviation nodes; synchronously collect motor operating parameters, and construct a multidimensional data item set in combination with the displacement data. The FP-Growth association algorithm is used to mine the implicit association rules between displacement anomalies and motor parameters. When it is determined that a specific displacement deviation is associated with a motor state parameter, an abnormal association node is generated and an abnormal displacement-motor parameter group is defined. Finally, the subsequent test plan is dynamically optimized based on the abnormal characteristic parameter group, forming a "detection-diagnosis-adaptation" closed-loop test process. The present invention realizes the adaptive adjustment of the test plan, significantly improving the intelligence level of electric bed quality inspection and the efficiency of abnormal factor tracing.

Description

Technical Field

[0001] The present invention relates to the field of intelligent analysis of electric beds, and more specifically, to an intelligent testing method and system for electric beds based on visual analysis. Background Art

[0002] Traditional electric bed testing methods rely on manual operation and subjective judgment, resulting in low efficiency, poor accuracy, and limited coverage scenarios. Electric beds are prone to control deviations and inaccurate lifting and lowering in diverse scenarios and complex operating conditions. Traditional testing processes struggle to effectively assess abnormal conditions and regulate production, lacking the ability to leverage machine vision for efficient deviation analysis and test control. Furthermore, they lack correlation analysis between electric bed control deviations and motor operation, making it impossible to effectively identify potential abnormal control parameters.

[0003] Therefore, there is an urgent need for an intelligent testing method for electric beds based on visual analysis. Summary of the Invention

[0004] The present invention overcomes the defects of the prior art and proposes an intelligent testing method and system for an electric bed based on visual analysis.

[0005] A first aspect of the present invention provides an intelligent testing method for an electric bed based on visual analysis, comprising:

[0006] S1: In the first test cycle, multiple test nodes are set, and the target electric bed is tested based on a preset plan. Based on the machine vision module, multiple parts of the target electric bed are located and displacement analysis is performed, and displacement monitoring data is obtained;

[0007] S2: Evaluate displacement deviation based on displacement monitoring data and set abnormal nodes;

[0008] S3: Obtain motor monitoring data of the target electric bed through the motor module, set a data item according to each parameter, generate multiple data items according to the displacement monitoring data and the motor monitoring data, analyze the correlation between the data items through the FP-Growth correlation algorithm, and obtain the correlated data items;

[0009] S4: If the associated data item includes displacement and motor parameters, and the corresponding test node is an abnormal node, the corresponding node is marked as an abnormal associated node, and an abnormal displacement-motor parameter group is dynamically set based on the abnormal associated node;

[0010] S5: Dynamically set the test plan according to the abnormal displacement-motor parameter group, and generate a dynamic test plan by combining the abnormal related nodes.

[0011] In this solution, S1 is specifically:

[0012] In the first test cycle, multiple test nodes are set to ensure that the intervals between each test node are consistent;

[0013] The target electric bed is tested based on a preset plan. During the test, the video data of the target electric bed is obtained through the machine vision module;

[0014] Extract key frames from video data and perform smoothing, noise reduction and standardization preprocessing on key frame images;

[0015] The key frame images include multi-angle image frames of the target electric bed;

[0016] Based on the YOLOv7 model, feature extraction and target recognition are performed on key frame images. Identification and positioning of various parts of the target electric bed and target movement monitoring are performed. Parameters of multiple test nodes are recorded based on the movement speed, movement amplitude, rotation angle, and movement trajectory of each part to obtain displacement monitoring data.

[0017] In this solution, S2 is specifically:

[0018] The displacement monitoring data is compared with the expected displacement parameters of each part of the bed, and the displacement parameters are set with preset weights to calculate the displacement weighted deviation rate of multiple test nodes;

[0019] Determine whether the displacement weighted deviation rate is greater than the preset deviation rate. If so, mark the corresponding test node as an abnormal node.

[0020] In this solution, S3 is specifically:

[0021] The motor parameters are monitored and recorded in the first test cycle by the motor module to obtain motor monitoring data;

[0022] Motor parameters include current effective value, current harmonic components, voltage, speed, and vibration frequency;

[0023] In the motor monitoring data, each motor parameter is used as a data item, and in the displacement monitoring data, each displacement parameter is used as a data item to generate a data item set;

[0024] Based on the FP-Growth association algorithm, each data item in the data item set is scanned and its support is calculated, and frequent items are screened out based on the minimum support;

[0025] Construct the initial FP tree structure, combine the frequent items of adjacent nodes to construct the transaction item set, sort the transaction item set in descending order based on support, and insert the sorted transaction item set into the FP tree in sequence to complete the FP tree construction;

[0026] In a recursive form, frequent item sets are mined on the FP tree, and the frequent item sets obtained by each screening are marked as associated data items until all frequent item sets are mined and multiple associated data items are obtained.

[0027] In this solution, S4 is specifically:

[0028] For an associated data item, if there are displacement parameters and motor parameters at the same time, the corresponding multiple associated parameters are extracted, and the test nodes corresponding to the multiple associated parameters are marked as associated nodes;

[0029] If the associated nodes are the same node or adjacent nodes, and all the associated nodes are abnormal nodes, the associated nodes are marked as abnormal associated nodes;

[0030] Taking the abnormal associated node as the reference point, based on the preset time span, the displacement parameters and motor parameters within the preset time range are obtained, and the parameters are associated and stored to obtain the abnormal displacement-motor parameter group.

[0031] In this solution, S5 is specifically:

[0032] For an abnormal displacement-motor parameter group, dynamically set the test content, combine the proportion of abnormal related nodes to the test nodes, set the test time period, and generate a dynamic test plan;

[0033] Based on multiple abnormal displacement-motor parameter groups, multiple dynamic test scenarios are generated.

[0034] A second aspect of the present invention further provides a visual analysis-based intelligent testing system for an electric bed, the system comprising: a memory and a processor, wherein the memory includes a visual analysis-based intelligent testing program for an electric bed, and when the visual analysis-based intelligent testing program for an electric bed is executed by the processor, the following steps are implemented:

[0035] S1: In the first test cycle, multiple test nodes are set, and the target electric bed is tested based on a preset plan. Based on the machine vision module, multiple parts of the target electric bed are located and displacement analysis is performed, and displacement monitoring data is obtained;

[0036] S2: Evaluate displacement deviation based on displacement monitoring data and set abnormal nodes;

[0037] S3: Obtain motor monitoring data of the target electric bed through the motor module, set a data item according to each parameter, generate multiple data items according to the displacement monitoring data and the motor monitoring data, analyze the correlation between the data items through the FP-Growth correlation algorithm, and obtain the correlated data items;

[0038] S4: If the associated data item includes displacement and motor parameters, and the corresponding test node is an abnormal node, the corresponding node is marked as an abnormal associated node, and an abnormal displacement-motor parameter group is dynamically set based on the abnormal associated node;

[0039] S5: Dynamically set the test plan according to the abnormal displacement-motor parameter group, and generate a dynamic test plan by combining the abnormal related nodes.

[0040] The third aspect of the present invention also provides a computer-readable storage medium, which includes an electric bed intelligent testing program based on visual analysis. When the electric bed intelligent testing program based on visual analysis is executed by a processor, the steps of the electric bed intelligent testing method based on visual analysis as described in any one of the above are implemented.

[0041] The present invention discloses an intelligent testing method and system for electric beds based on visual analysis. The method comprises the following steps: in a first test cycle, using a machine vision module to perform spatial positioning and displacement trajectory analysis on multiple parts of the electric bed, generate displacement monitoring data, and evaluate deviation nodes; synchronously collect motor operating parameters, and construct a multidimensional data item set in combination with the displacement data. The FP-Growth association algorithm is used to mine the implicit association rules between displacement anomalies and motor parameters. When it is determined that a specific displacement deviation is associated with a motor state parameter, an abnormal association node is generated and an abnormal displacement-motor parameter group is defined. Finally, the subsequent test plan is dynamically optimized based on the abnormal characteristic parameter group, forming a closed-loop test process of "detection-diagnosis-adaptation". The present invention realizes the adaptive adjustment of the test plan, significantly improving the intelligence level of electric bed quality inspection and the efficiency of abnormal factor tracing. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 A flow chart of an intelligent testing method for an electric bed based on visual analysis according to the present invention is shown;

[0043] Figure 2 The block diagram of the intelligent testing system for electric bed based on visual analysis of the present invention is shown. DETAILED DESCRIPTION

[0044] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present application and the features therein can be combined with each other.

[0045] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0046] Figure 1 The flowchart of the intelligent testing method of an electric bed based on visual analysis of the present invention is shown.

[0047] like Figure 1 As shown, the first aspect of the present invention provides an intelligent testing method for an electric bed based on visual analysis, comprising:

[0048] S1: In the first test cycle, multiple test nodes are set, and the target electric bed is tested based on a preset plan. Based on the machine vision module, multiple parts of the target electric bed are located and displacement analysis is performed, and displacement monitoring data is obtained;

[0049] S2: Evaluate displacement deviation based on displacement monitoring data and set abnormal nodes;

[0050] S3: Obtain motor monitoring data of the target electric bed through the motor module, set a data item according to each parameter, generate multiple data items according to the displacement monitoring data and the motor monitoring data, analyze the correlation between the data items through the FP-Growth correlation algorithm, and obtain the correlated data items;

[0051] S4: If the associated data item includes displacement and motor parameters, and the corresponding test node is an abnormal node, the corresponding node is marked as an abnormal associated node, and an abnormal displacement-motor parameter group is dynamically set based on the abnormal associated node;

[0052] S5: Dynamically set the test plan according to the abnormal displacement-motor parameter group, and generate a dynamic test plan by combining the abnormal related nodes.

[0053] According to an embodiment of the present invention, S1 specifically includes:

[0054] In the first test cycle, multiple test nodes are set to ensure that the intervals between each test node are consistent;

[0055] The target electric bed is tested based on a preset plan. During the test, the video data of the target electric bed is obtained through the machine vision module;

[0056] Extract key frames from video data and perform smoothing, noise reduction and standardization preprocessing on key frame images;

[0057] The key frame images include multi-angle image frames of the target electric bed;

[0058] Based on the YOLOv7 model, feature extraction and target recognition are performed on key frame images. Identification and positioning of various parts of the target electric bed and target movement monitoring are performed. Parameters of multiple test nodes are recorded based on the movement speed, movement amplitude, rotation angle, and movement trajectory of each part to obtain displacement monitoring data.

[0059] It should be noted that the test node is a time node, which is used to set multiple time points within the test cycle to analyze the displacement deviation of the electric bed and the abnormal changes in the electrical parameters of the motor. The machine vision module includes multiple camera devices, which obtain video streams corresponding to different angles, and perform corresponding displacement parameter analysis and calculation based on the image features extracted at different angles. The displacement parameters include movement speed, movement amplitude, rotation angle, and movement trajectory. Here, each part can be set according to the movement structure of the bed. For example, three parts can be set, corresponding to the head, waist, and legs that are in contact with the human body, and the corresponding bed parts can be moved to meet the user's control requirements. In the present invention, target detection is performed based on the YOLOv7 model to improve detection efficiency. According to research needs, other detection algorithms can be applied for target recognition detection, such as R-CNN, Fast R-CNN and other image algorithm networks. Before applying target detection, initial training can be performed based on the relevant preset image features of the target electric bed.

[0060] The preset plan is a general test plan, including movement and lifting tests, rotation tests, bed pressure tests, load tests, etc. of multiple parts of the bed.

[0061] According to an embodiment of the present invention, S2 specifically includes:

[0062] The displacement monitoring data is compared with the expected displacement parameters of each part of the bed, and the displacement parameters are set with preset weights to calculate the displacement weighted deviation rate of multiple test nodes;

[0063] Determine whether the displacement weighted deviation rate is greater than the preset deviation rate. If so, mark the corresponding test node as an abnormal node.

[0064] It should be noted that there are multiple displacement parameters. Within the preset weights, weights for different displacement parameters are set based on the bed's production and control requirements. This helps control the system's deviation analysis of the electric bed. Expected data, such as expected displacement parameters, are idealized parameter values ​​set based on the test plan and used for comparative analysis of control deviations.

[0065] It is worth mentioning here that the preset deviation rate can be set to a lower range to screen out nodes that may have deviations and abnormal risks, so that a larger number of abnormal nodes can be screened out. Based on analysis requirements, the number of abnormal nodes obtained can be greater than 50% of all nodes. Further, subsequent correlation analysis is used to perform correlation anomaly assessment.

[0066] According to an embodiment of the present invention, S3 specifically includes:

[0067] The motor parameters are monitored and recorded in the first test cycle by the motor module to obtain motor monitoring data;

[0068] Motor parameters include current effective value, current harmonic components, voltage, speed, and vibration frequency;

[0069] In the motor monitoring data, each motor parameter is used as a data item, and in the displacement monitoring data, each displacement parameter is used as a data item to generate a data item set;

[0070] Based on the FP-Growth association algorithm, each data item in the data item set is scanned and its support is calculated, and frequent items are screened out based on the minimum support;

[0071] Construct the initial FP tree structure, combine the frequent items of adjacent nodes to construct the transaction item set, sort the transaction item set in descending order based on support, and insert the sorted transaction item set into the FP tree in sequence to complete the FP tree construction;

[0072] In a recursive form, frequent item sets are mined on the FP tree, and the frequent item sets obtained by each screening are marked as associated data items until all frequent item sets are mined and multiple associated data items are obtained.

[0073] It should be noted that one data item corresponds to one parameter information of a test node. The minimum support can be set to 0.3. The FP tree is a compressed data structure for efficiently mining frequent item sets. It can be constructed by scanning the data set twice. Frequent item sets are mined from the FP tree. Generally, a conditional FP tree is constructed in a recursive form and a frequent item set generation method is applied to mine related items. During the mining process, a frequent item in the FP tree is selected as a suffix, and its conditional pattern base is constructed. Based on the conditional pattern base, the mining process is recursively performed on the conditional FP tree to generate all frequent item sets. The frequent item sets obtained by recursive mining are integrated to obtain complete frequent item set data. In the present invention, the parameter information of the relevant test nodes is used as the data item, and the data items of the adjacent nodes are used as the transaction items. The adjacent nodes are the frequent items of the adjacent test nodes. The frequent items are a special data item corresponding to a parameter. In the combination process, two or more data items are usually combined to form a transaction item. For example, transaction A [movement amplitude = 0.5, current effective value = 10]. Furthermore, based on the FP-Growth association algorithm, associated data items are mined from frequent items, and frequent item sets are associated data items.

[0074] Based on the test requirements, the motor parameters are increased or decreased accordingly for analysis.

[0075] According to an embodiment of the present invention, the S4 is specifically:

[0076] For an associated data item, if there are displacement parameters and motor parameters at the same time, the corresponding multiple associated parameters are extracted, and the test nodes corresponding to the multiple associated parameters are marked as associated nodes;

[0077] If the associated nodes are the same node or adjacent nodes, and all the associated nodes are abnormal nodes, the associated nodes are marked as abnormal associated nodes;

[0078] Taking the abnormal associated node as the reference point, based on the preset time span, the displacement parameters and motor parameters within the preset time range are obtained, and the parameters are associated and stored to obtain the abnormal displacement-motor parameter group.

[0079] It should be noted that there are multiple test nodes corresponding to an associated data item, and multiple associated nodes can be marked. The associated data item includes at least two data items, that is, corresponding to two parameters. The preset time span is set by the user, and the user obtains the associated parameter group within a certain time range. For example, two test nodes can be set as the preset time span, and then an abnormal associated node is used as the reference point. All displacements and motor parameters of the two test nodes forward and backward on the time axis are extracted to obtain an abnormal displacement-motor parameter group, in which the displacement and motor parameters under the same node are mapped, associated and stored in the parameter group. The extraction process is obtained through the displacement monitoring data and motor monitoring data in the first test cycle.

[0080] According to an embodiment of the present invention, the S5 is specifically:

[0081] For an abnormal displacement-motor parameter group, dynamically set the test content, combine the proportion of abnormal related nodes to the test nodes, set the test time period, and generate a dynamic test plan;

[0082] Based on multiple abnormal displacement-motor parameter groups, multiple dynamic test scenarios are generated.

[0083] It should be noted that the higher the proportion of test nodes, the longer the test period. The proportion is calculated based on the total number of abnormally associated nodes and the total number of test nodes. Each dynamic test plan dynamically sets the associated parameters for each associated abnormal node. The dynamic test plan setting can effectively explore potential abnormal factors and related abnormal operating conditions of the electric bed.

[0084] It is worth noting here that the traditional electric bed production and testing process often relies on manual operation and subjective judgment, and the testing scheme is single. Electric beds are prone to control deviation and inaccurate lifting and lowering in multiple scenarios and complex operating conditions. The traditional testing process is difficult to effectively evaluate abnormal conditions and regulate production, and there is a lack of correlation analysis between the control deviation of the electric bed and the motor operation process, and a lack of dynamic test analysis process, which makes it impossible to effectively explore potential abnormal control factors.

[0085] Based on this, the present invention performs preliminary testing and monitoring on the target electric bed through a preset test scheme, obtains multi-parameter displacement monitoring data and motor monitoring data based on the machine vision module and the motor module, screens preliminary abnormal nodes through precise analysis of multiple test nodes, introduces the FP-Growth association algorithm, determines the nodes where the motor and displacement parameters are associated under multiple nodes, and screens out abnormal associated nodes. The abnormal associated nodes are time nodes where abnormal working conditions exist and there are potential control anomalies in the two parameters. After mining the nodes, the preset time range is further dynamically expanded to extract and associate the relevant displacement and motor parameters to form a parameter group, and a dynamic test scheme is constructed for the parameter group and production control is carried out during the test process. Through the present invention, a test scheme suitable for the target electric bed can be dynamically set, and potential abnormal control factors can be accurately mined. It can also dynamically set effective complex test content to achieve collaborative analysis and fault prediction of multiple parameters of the electric bed.

[0086] At the same time, the present invention can improve the efficiency of capturing abnormal working conditions, accurately analyze and predict the operating status of the target electric bed, effectively evaluate and test the abnormal control deviation of the electric bed at different test time nodes, and achieve efficient guidance of electric bed production and finished product screening.

[0087] According to an embodiment of the present invention, the further embodiment includes:

[0088] Among the abnormal nodes, the associated abnormal nodes are marked as first nodes, and the non-associated abnormal nodes are marked as second nodes;

[0089] Based on the plurality of first nodes, calculating deviation rates of corresponding motor parameters to obtain a plurality of first deviation rates, and performing linear fitting based on the plurality of first deviation rates to obtain a first curve;

[0090] Perform motor parameter fitting based on the plurality of second nodes to obtain a second curve;

[0091] Performing weighted fusion on the first curve and the second curve, specifically performing weighted averaging on the fitting coefficients of the first curve and the second curve, and generating a motor operation attenuation curve;

[0092] During the target electric bed test, abnormal operating conditions are predicted based on the motor operation attenuation curve, and test warnings are set.

[0093] It should be noted that the deviation rate of the motor parameters can select one or more motor parameters and compare them with the expected parameter values ​​to analyze the motor deviation rate at a certain node. The first curve is the motor operation attenuation curve obtained based on the analysis of the first node. This curve can reflect the trend of the probability of abnormal operation at the first node, and the second curve is the motor operation attenuation curve based on the first node. The fitting process of the two is consistent. In the weighted average of the fitting coefficients, the weights of the first and second curves are the proportion of the first node to the abnormal node and the proportion of the second node to the abnormal node, respectively. The motor operation attenuation curve finally generated is a fitting curve for overall evaluation based on abnormal working conditions. The fitting curve can effectively predict the abnormal working condition trend of the motor operation, evaluate the normal operating state curve, and realize intelligent test warning, improve test safety and realize efficient production control of electric beds.

[0094] According to an embodiment of the present invention, the further embodiment includes:

[0095] Set up multiple continuous test cycles based on multiple dynamic test scenarios;

[0096] During multiple continuous test cycles, the electric bed is tested, monitored and analyzed, and the motor operating parameters of the electric bed are compensated and regulated based on the abnormal displacement-motor parameter group;

[0097] Determine the growth of the number of abnormal nodes and the number of abnormal-related nodes in each test cycle, and evaluate the effectiveness of compensation and control.

[0098] It should be noted that motor operating parameter compensation includes voltage and current compensation. Dynamic testing solutions can be applied during the second test cycle. Motor operating parameter compensation and control are used to analyze production control of electric beds during operation and adjust the production process of motors and related mechanical structures to improve product yield.

[0099] During the testing and research of a single electric bed, a dynamic plan can be generated based on the test process of the first test cycle, and the dynamic plan can be applied to the next test cycle. Based on the abnormal working conditions of the next test cycle, it is decided whether to perform intelligent evaluation and dynamic generation of the test plan.

[0100] Figure 2 The block diagram of the intelligent testing system for electric bed based on visual analysis of the present invention is shown.

[0101] A second aspect of the present invention further provides a visual analysis-based electric bed intelligent testing system 2, the system comprising: a memory 21 and a processor 22, wherein the memory 21 includes a visual analysis-based electric bed intelligent testing program, and when the visual analysis-based electric bed intelligent testing program is executed by the processor 22, the following steps are implemented:

[0102] S1: In the first test cycle, multiple test nodes are set, and the target electric bed is tested based on a preset plan. Based on the machine vision module, multiple parts of the target electric bed are located and displacement analysis is performed, and displacement monitoring data is obtained;

[0103] S2: Evaluate displacement deviation based on displacement monitoring data and set abnormal nodes;

[0104] S3: Obtain motor monitoring data of the target electric bed through the motor module, set a data item according to each parameter, generate multiple data items according to the displacement monitoring data and the motor monitoring data, analyze the correlation between the data items through the FP-Growth correlation algorithm, and obtain the correlated data items;

[0105] S4: If the associated data item includes displacement and motor parameters, and the corresponding test node is an abnormal node, the corresponding node is marked as an abnormal associated node, and an abnormal displacement-motor parameter group is dynamically set based on the abnormal associated node;

[0106] S5: Dynamically set the test plan according to the abnormal displacement-motor parameter group, and generate a dynamic test plan by combining the abnormal related nodes.

[0107] According to an embodiment of the present invention, S1 specifically includes:

[0108] In the first test cycle, multiple test nodes are set to ensure that the intervals between each test node are consistent;

[0109] The target electric bed is tested based on a preset plan. During the test, the video data of the target electric bed is obtained through the machine vision module;

[0110] Extract key frames from video data and perform smoothing, noise reduction and standardization preprocessing on key frame images;

[0111] The key frame images include multi-angle image frames of the target electric bed;

[0112] Based on the YOLOv7 model, feature extraction and target recognition are performed on key frame images. Identification and positioning of various parts of the target electric bed and target movement monitoring are performed. Parameters of multiple test nodes are recorded based on the movement speed, movement amplitude, rotation angle, and movement trajectory of each part to obtain displacement monitoring data.

[0113] It should be noted that the test node is a time node, which is used to set multiple time points within the test cycle to analyze the displacement deviation of the electric bed and the abnormal changes in the electrical parameters of the motor. The machine vision module includes multiple camera devices, which obtain video streams corresponding to different angles, and perform corresponding displacement parameter analysis and calculation based on the image features extracted at different angles. The displacement parameters include movement speed, movement amplitude, rotation angle, and movement trajectory. Here, each part can be set according to the movement structure of the bed. For example, three parts can be set, corresponding to the head, waist, and legs that are in contact with the human body, and the corresponding bed parts can be moved to meet the user's control requirements. In the present invention, target detection is performed based on the YOLOv7 model to improve detection efficiency. According to research needs, other detection algorithms can be applied for target recognition detection, such as R-CNN, Fast R-CNN and other image algorithm networks. Before applying target detection, initial training can be performed based on the relevant preset image features of the target electric bed.

[0114] The preset plan is a general test plan, including movement and lifting tests, rotation tests, bed pressure tests, load tests, etc. of multiple parts of the bed.

[0115] According to an embodiment of the present invention, S2 specifically includes:

[0116] The displacement monitoring data is compared with the expected displacement parameters of each part of the bed, and the displacement parameters are set with preset weights to calculate the displacement weighted deviation rate of multiple test nodes;

[0117] Determine whether the displacement weighted deviation rate is greater than the preset deviation rate. If so, mark the corresponding test node as an abnormal node.

[0118] It should be noted that there are multiple displacement parameters. Within the preset weights, weights for different displacement parameters are set based on the bed's production and control requirements. This helps control the system's deviation analysis of the electric bed. Expected data, such as expected displacement parameters, are idealized parameter values ​​set based on the test plan and used for comparative analysis of control deviations.

[0119] It is worth mentioning here that the preset deviation rate can be set to a lower range to screen out nodes that may have deviations and abnormal risks, so that a larger number of abnormal nodes can be screened out. Based on analysis requirements, the number of abnormal nodes obtained can be greater than 50% of all nodes. Further, subsequent correlation analysis is used to perform correlation anomaly assessment.

[0120] According to an embodiment of the present invention, S3 specifically includes:

[0121] The motor parameters are monitored and recorded in the first test cycle by the motor module to obtain motor monitoring data;

[0122] Motor parameters include current effective value, current harmonic components, voltage, speed, and vibration frequency;

[0123] In the motor monitoring data, each motor parameter is used as a data item, and in the displacement monitoring data, each displacement parameter is used as a data item to generate a data item set;

[0124] Based on the FP-Growth association algorithm, each data item in the data item set is scanned and its support is calculated, and frequent items are screened out based on the minimum support;

[0125] Construct the initial FP tree structure, combine the frequent items of adjacent nodes to construct the transaction item set, sort the transaction item set in descending order based on support, and insert the sorted transaction item set into the FP tree in sequence to complete the FP tree construction;

[0126] In a recursive form, frequent item sets are mined on the FP tree, and the frequent item sets obtained by each screening are marked as associated data items until all frequent item sets are mined and multiple associated data items are obtained.

[0127] It should be noted that one data item corresponds to one parameter information of a test node. The minimum support can be set to 0.3. The FP tree is a compressed data structure for efficiently mining frequent item sets. It can be constructed by scanning the data set twice. Frequent item sets are mined from the FP tree. Generally, a conditional FP tree is constructed in a recursive form and a frequent item set generation method is applied to mine related items. During the mining process, a frequent item in the FP tree is selected as a suffix, and its conditional pattern base is constructed. Based on the conditional pattern base, the mining process is recursively performed on the conditional FP tree to generate all frequent item sets. The frequent item sets obtained by recursive mining are integrated to obtain complete frequent item set data. In the present invention, the parameter information of the relevant test nodes is used as the data item, and the data items of the adjacent nodes are used as the transaction items. The adjacent nodes are the frequent items of the adjacent test nodes. The frequent items are a special data item corresponding to a parameter. In the combination process, two or more data items are usually combined to form a transaction item. For example, transaction A [movement amplitude = 0.5, current effective value = 10]. Furthermore, based on the FP-Growth association algorithm, associated data items are mined from frequent items, and frequent item sets are associated data items.

[0128] Based on the test requirements, the motor parameters are increased or decreased accordingly for analysis.

[0129] According to an embodiment of the present invention, the S4 is specifically:

[0130] For an associated data item, if there are displacement parameters and motor parameters at the same time, the corresponding multiple associated parameters are extracted, and the test nodes corresponding to the multiple associated parameters are marked as associated nodes;

[0131] If the associated nodes are the same node or adjacent nodes, and all the associated nodes are abnormal nodes, the associated nodes are marked as abnormal associated nodes;

[0132] Taking the abnormal associated node as the reference point, based on the preset time span, the displacement parameters and motor parameters within the preset time range are obtained, and the parameters are associated and stored to obtain the abnormal displacement-motor parameter group.

[0133] It should be noted that there are multiple test nodes corresponding to an associated data item, and multiple associated nodes can be marked. The associated data item includes at least two data items, that is, corresponding to two parameters. The preset time span is set by the user, and the user obtains the associated parameter group within a certain time range. For example, two test nodes can be set as the preset time span, and then an abnormal associated node is used as the reference point. All displacements and motor parameters of the two test nodes forward and backward on the time axis are extracted to obtain an abnormal displacement-motor parameter group, in which the displacement and motor parameters under the same node are mapped, associated and stored in the parameter group. The extraction process is obtained through the displacement monitoring data and motor monitoring data in the first test cycle.

[0134] According to an embodiment of the present invention, the S5 is specifically:

[0135] For an abnormal displacement-motor parameter group, dynamically set the test content, combine the proportion of abnormal related nodes to the test nodes, set the test time period, and generate a dynamic test plan;

[0136] Based on multiple abnormal displacement-motor parameter groups, multiple dynamic test scenarios are generated.

[0137] It should be noted that the higher the proportion of test nodes, the longer the test period. The proportion is calculated based on the total number of abnormally associated nodes and the total number of test nodes. Each dynamic test plan dynamically sets the associated parameters for each associated abnormal node. The dynamic test plan setting can effectively explore potential abnormal factors and related abnormal operating conditions of the electric bed.

[0138] The third aspect of the present invention also provides a computer-readable storage medium, which includes an electric bed intelligent testing program based on visual analysis. When the electric bed intelligent testing program based on visual analysis is executed by a processor, the steps of the electric bed intelligent testing method based on visual analysis as described in any one of the above are implemented.

[0139] The present invention discloses an intelligent testing method and system for electric beds based on visual analysis. The method comprises the following steps: in a first test cycle, using a machine vision module to perform spatial positioning and displacement trajectory analysis on multiple parts of the electric bed, generate displacement monitoring data, and evaluate deviation nodes; synchronously collect motor operating parameters, and construct a multidimensional data item set in combination with the displacement data. The FP-Growth association algorithm is used to mine the implicit association rules between displacement anomalies and motor parameters. When it is determined that a specific displacement deviation is associated with a motor state parameter, an abnormal association node is generated and an abnormal displacement-motor parameter group is defined. Finally, the subsequent test plan is dynamically optimized based on the abnormal characteristic parameter group, forming a closed-loop test process of "detection-diagnosis-adaptation". The present invention realizes the adaptive adjustment of the test plan, significantly improving the intelligence level of electric bed quality inspection and the efficiency of abnormal factor tracing.

[0140] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0141] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed across multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0142] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0143] Those skilled in the art will appreciate that all or part of the steps of the above-mentioned method embodiments may be implemented by hardware associated with program instructions, and the aforementioned program may be stored in a computer-readable storage medium. When the program is executed, the program executes the steps of the above-mentioned method embodiments. The aforementioned storage medium includes various media that can store program codes, such as mobile storage devices, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0144] Alternatively, if the integrated units described above are implemented as software modules and sold or used as standalone products, they can also be stored on a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product, stored on a storage medium, includes instructions for enabling a computer device (such as a personal computer, server, or network device) to execute all or part of the methods described in various embodiments of the present invention. The aforementioned storage media include various media capable of storing program code, such as removable storage devices, ROM, RAM, magnetic disks, or optical disks.

[0145] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be covered by the scope of protection of the present invention.

Claims

1. An intelligent testing method for electric beds based on visual analysis, characterized in that: include: S1: In the first test cycle, multiple test nodes are set, and the target electric bed is tested based on a preset plan. Based on the machine vision module, multiple parts of the target electric bed are located and displacement analysis is performed. The displacement parameters of multiple test nodes are recorded based on the movement speed, movement amplitude, rotation angle, and movement trajectory of each part to obtain displacement monitoring data; S2: Evaluate displacement deviation based on displacement monitoring data and set abnormal nodes; S3: Using the motor module, monitor and record the motor parameters of the target electric bed during the first test cycle to obtain motor monitoring data; the motor parameters include effective current value, current harmonic components, voltage, speed, and vibration frequency; a data item is set for each motor parameter, multiple data items are generated based on the displacement monitoring data and the motor monitoring data, and the correlation between the data items is analyzed using the FP-Growth correlation algorithm to obtain the correlated data items; S4: If the associated data item includes displacement parameters and motor parameters, and the corresponding test node is an abnormal node, the corresponding node is marked as an abnormal associated node, and an abnormal displacement-motor parameter group is dynamically set based on the abnormal associated node; S5: Dynamically set the test plan based on the abnormal displacement-motor parameter group, and generate a dynamic test plan by combining the abnormal related nodes; The test node is a time node, which is used to set multiple time points within the test cycle to analyze the displacement deviation of the target electric bed and abnormal changes in motor parameters; The displacement parameters include movement speed, movement amplitude, rotation angle, and movement trajectory; the motor parameters include current effective value, current harmonic components, voltage, speed, and vibration frequency.

2. The intelligent testing method for an electric bed based on visual analysis according to claim 1 is characterized in that: The S1 is specifically: In the first test cycle, multiple test nodes are set to ensure that the intervals between each test node are consistent; The target electric bed is tested based on a preset plan. During the test, the video data of the target electric bed is obtained through the machine vision module; Extract key frames from video data and perform smoothing, noise reduction and standardization preprocessing on key frame images; The key frame images include multi-angle image frames of the target electric bed; Based on the YOLOv7 model, feature extraction and target recognition are performed on key frame images. Identification and positioning of various parts of the target electric bed and target movement monitoring are performed. Parameters of multiple test nodes are recorded based on the movement speed, movement amplitude, rotation angle, and movement trajectory of each part to obtain displacement monitoring data.

3. The intelligent testing method for an electric bed based on visual analysis according to claim 1 is characterized in that: The S2 is specifically: The displacement monitoring data is compared with the expected displacement parameters of each part of the bed, and the displacement parameters are set with preset weights to calculate the displacement weighted deviation rate of multiple test nodes; Determine whether the displacement weighted deviation rate is greater than the preset deviation rate. If so, mark the corresponding test node as an abnormal node.

4. The intelligent testing method for an electric bed based on visual analysis according to claim 1 is characterized in that: The S3 is specifically: In the motor monitoring data, each motor parameter is used as a data item, and in the displacement monitoring data, each displacement parameter is used as a data item to generate a data item set; Based on the FP-Growth association algorithm, each data item in the data item set is scanned and its support is calculated, and frequent items are screened out based on the minimum support; Construct the initial FP tree structure, combine the frequent items of adjacent nodes to construct the transaction item set, sort the transaction item set in descending order based on support, and insert the sorted transaction item set into the FP tree in sequence to complete the FP tree construction; In a recursive form, frequent item sets are mined on the FP tree, and the frequent item sets obtained by each screening are marked as associated data items until all frequent item sets are mined and multiple associated data items are obtained.

5. The intelligent testing method for an electric bed based on visual analysis according to claim 1 is characterized in that: The S4 is specifically: For an associated data item, if there are displacement parameters and motor parameters at the same time, the corresponding multiple associated parameters are extracted, and the test nodes corresponding to the multiple associated parameters are marked as associated nodes; If the associated nodes are the same node or adjacent nodes, and all the associated nodes are abnormal nodes, the associated nodes are marked as abnormal associated nodes; Taking the abnormal associated node as the reference point, based on the preset time span, the displacement parameters and motor parameters within the preset time range are obtained, and the displacement parameters and motor parameters are associated and stored to obtain the abnormal displacement-motor parameter group.

6. The intelligent testing method for an electric bed based on visual analysis according to claim 1 is characterized in that: The S5 is specifically: For an abnormal displacement-motor parameter group, dynamically set the test content, combine the proportion of abnormal related nodes to the test nodes, set the test time period, and generate a dynamic test plan; Based on multiple abnormal displacement-motor parameter groups, multiple dynamic test scenarios are generated.

7. An intelligent testing system for electric beds based on visual analysis, characterized in that: The system includes: a memory and a processor. The memory includes a visual analysis-based electric bed intelligent testing program. When the visual analysis-based electric bed intelligent testing program is executed by the processor, the steps of the visual analysis-based electric bed intelligent testing method as described in claim 1 are implemented.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a motor-assisted bed intelligent testing program based on visual analysis. When the motor-assisted bed intelligent testing program based on visual analysis is executed by a processor, the steps of the motor-assisted bed intelligent testing method based on visual analysis according to any one of claims 1 to 6 are implemented.

Images