Abnormality detection method, electronic equipment and storage medium

Abstract

The invention discloses an anomaly detection method, electronic equipment and a storage medium, and the method comprises the steps: obtaining first docking position deviations between a first automatic guided vehicle and a plurality of target docking objects, and obtaining second docking position deviations between a second automatic guided vehicle and the plurality of target docking objects, both the first automatic guided vehicle and the second automatic guided vehicle are provided with positioning devices for navigation; determining the positioning similarity between the first automated guided vehicle and the second automated guided vehicle based on the first docking pose deviations and the second docking pose deviations; and determining whether the positioning device of the first automated guided vehicle is abnormal or not by using the positioning similarity. According to the scheme, whether the positioning device is abnormal or not can be accurately detected.

Description

Technical Field

[0001] The present application relates to the technical field of automatic guided vehicles, and in particular to an abnormality detection method, electronic equipment, and storage medium. Background Art

[0002] Before an Automated Guided Vehicle (AGV) docks with a docking object, it needs to use a positioning device to locate and plan a route to the vicinity of the docking object before docking with the docking object. However, the installation position of the positioning device on the AGV may slowly shift over time, causing errors in the calibration results of the positioning device. Therefore, the positioning, navigation and other information obtained by the positioning device must also have errors. If the error is large, the AGV will not be able to operate normally, and even destroy the overall operating status of the docking site. Therefore, how to detect abnormal positioning devices is an urgent problem to be solved. Summary of the invention

[0003] The present application at least provides an abnormality detection method, electronic equipment, and storage medium, which can accurately detect whether there is an abnormality in the positioning device.

[0004] A first aspect of the present application provides an abnormality detection method, the method comprising: obtaining first docking position deviations between a first automated guided vehicle and a plurality of target docking objects, and obtaining second docking position deviations between a second automated guided vehicle and a plurality of target docking objects, wherein both the first automated guided vehicle and the second automated guided vehicle are provided with a positioning device for navigation; based on each first docking position deviation and each second docking position deviation, determining the positioning similarity between the first automated guided vehicle and the second automated guided vehicle; and using the positioning similarity, determining whether there is an abnormality in the positioning device of the first automated guided vehicle.

[0005] Among them, based on each first docking posture deviation and each second docking posture deviation, determining the positioning similarity between the first automated guided vehicle and the second automated guided vehicle, including: determining the first representative posture deviation of the first automated guided vehicle based on each first docking posture deviation, and determining the second representative posture deviation corresponding to the second automated guided vehicle based on each second docking posture deviation; obtaining the first posture deviation similarity between the first representative posture deviation and the second representative posture deviation as the positioning similarity; or, based on each first docking posture deviation and each second docking posture deviation, obtaining the second posture deviation similarity of each target docking object between the first automated guided vehicle and the second automated guided vehicle; and taking the central trend representation value of the second posture deviation similarity corresponding to each target docking object as the positioning similarity.

[0006] Among them, based on each first docking posture deviation and each second docking posture deviation, determining the positioning similarity between the first automated guided vehicle and the second automated guided vehicle, includes: obtaining the total number of target docking objects; in response to the total number being less than a first value, determining the first representative posture deviation of the first automated guided vehicle based on each first docking posture deviation, and determining the second representative posture deviation corresponding to the second automated guided vehicle based on each second docking posture deviation; obtaining the first posture deviation similarity between the first representative posture deviation and the second representative posture deviation as the positioning similarity; in response to the total number being greater than or equal to the first value, obtaining the second posture deviation similarity between the first automated guided vehicle and the second automated guided vehicle corresponding to each representative docking object based on each first docking posture deviation and each second docking posture deviation; and taking the central trend representation value of the second posture deviation similarity corresponding to each representative docking object as the positioning similarity.

[0007] Among them, before obtaining the second posture deviation similarity of the first automated guided vehicle and the second automated guided vehicle corresponding to each representative docking object based on each first docking posture deviation and each second docking posture deviation, it also includes: when the total number is between the first value and the second value, taking several target docking objects as representative docking objects, and the first value is less than the second value; when the total number is greater than the second value, selecting a second value of target docking objects from several target docking objects as representative docking objects.

[0008] Among them, based on each first docking posture deviation, determining the first representative posture deviation of the first automatic guided vehicle, including: clustering each first docking posture deviation to obtain a first cluster center, and using the first cluster center as the first representative posture deviation; based on each second docking posture deviation, determining the second representative posture deviation corresponding to the second automatic guided vehicle, including: clustering each first docking posture deviation to obtain a second cluster center, and using the second cluster center as the second representative posture deviation of the second automatic guided vehicle.

[0009] The target docking objects are all docking objects that the first automated guided vehicle and the second automated guided vehicle have docked with.

[0010] Among them, obtaining the first docking posture deviations between the first automatic guided vehicle and several target docking objects respectively includes: for each target docking object, obtaining the first historical docking posture deviation between the first automatic guided vehicle and the target docking object at at least one first historical moment, and obtaining the central trend characterization value corresponding to each first historical docking posture deviation as the first docking posture deviation; obtaining the second docking position deviations between the second automatic guided vehicle and several target docking objects respectively, including: for each target docking object, obtaining the second historical docking posture deviation between the second automatic guided vehicle and the target docking object at at least one second historical moment, and obtaining the central trend characterization value corresponding to each second historical docking posture deviation as the second docking posture deviation.

[0011] Among them, the step of obtaining the first historical docking posture deviation or the second historical docking posture deviation includes: obtaining the target docking posture when the current automated guided vehicle moves to the target position point according to the planned path, wherein the current automated guided vehicle is the first automated guided vehicle corresponding to the first historical docking posture deviation or the second automated guided vehicle corresponding to the second historical docking posture deviation, the planned path is generated based on the collected data of the positioning device on the current automated guided vehicle, and the target position point represents the position of the current automated guided vehicle docking with the target docking object; and, using the image acquisition device on the current automated guided vehicle, determining the actual docking posture required for the current automated guided vehicle to dock with the target docking object; obtaining the difference between the actual docking posture and the target docking posture. or, the step of acquiring the first historical docking posture deviation or the second historical docking posture deviation includes: when the current automated guided vehicle moves to the target position point according to the planned path, using an image acquisition device to shoot the target docking object to obtain a target docking image, wherein the current automated guided vehicle is the first automated guided vehicle corresponding to the first historical docking posture deviation or the second automated guided vehicle corresponding to the second historical docking posture deviation, the planned path is generated based on the acquisition data of the positioning device on the current automated guided vehicle, and the target position point represents the position of the current automated guided vehicle docking with the target docking object; the target docking image is analyzed to obtain the docking posture deviation between the current automated guided vehicle and the target docking object.

[0012] Among them, there are multiple second automated guided vehicles; using positioning similarity, determining whether there is an abnormality in the positioning device of the first automated guided vehicle, including: screening several positioning similarities that meet the similarity requirements from the positioning similarities corresponding to the multiple second automated guided vehicles; counting the multiple positioning similarities to obtain representative positioning similarities between the first automated guided vehicle and the multiple second automated guided vehicles; based on the representative positioning similarities, determining whether there is an abnormality in the positioning device of the first automated guided vehicle.

[0013] Among them, the similarity requirement is that the positioning similarity is greater than a first similarity threshold; and / or, a number of positioning similarities are counted to obtain a representative positioning similarity between the first automatic guided vehicle and multiple second automatic guided vehicles, including: obtaining the sum of a number of positioning similarities, and taking the ratio of the sum of the positioning similarities to the number of second automatic guided vehicles as the representative positioning similarity; and / or, based on the representative positioning similarity, determining whether there is an abnormality in the positioning device of the first automatic guided vehicle, including: in response to the representative positioning similarity being less than the second similarity threshold, determining that there is an abnormality in the positioning device of the first automatic guided vehicle.

[0014] A second aspect of the present application provides an electronic device, comprising a memory and a processor coupled to each other, wherein the processor is used to execute program instructions stored in the memory to implement the anomaly detection method in the first aspect.

[0015] A third aspect of the present application provides a computer-readable storage medium having program instructions stored thereon, and when the program instructions are executed by a processor, the anomaly detection method in the first aspect described above is implemented.

[0016] The above scheme obtains the first docking position deviations between the first automated guided vehicle and several target docking objects, and obtains the second docking position deviations between the second automated guided vehicle and several target docking objects, and then uses each first docking position deviation and each second docking position deviation to perform calculations to obtain the positioning similarity between the first automated guided vehicle and the second automated guided vehicle. Through the positioning similarity, it can be determined whether there is an abnormality in the positioning device of the first automated guided vehicle, thereby being able to accurately detect whether there is an abnormality in the positioning device.

[0017] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and are used together with the specification to illustrate the technical solution of the present application.

[0019] Figure 1 It is a flowchart of an embodiment of an abnormality detection method of the present application;

[0020] Figure 2 is a flowchart of another embodiment of the abnormality detection method of the present application;

[0021] Figure 3 It is a schematic diagram of the framework of an embodiment of an abnormality detection device of the present application;

[0022] Figure 4 It is a schematic diagram of the framework of an embodiment of the electronic device of the present application;

[0023] Figure 5 It is a schematic diagram of a framework of an embodiment of a computer-readable storage medium of the present application. DETAILED DESCRIPTION

[0024] The scheme of the embodiment of the present application is described in detail below in conjunction with the drawings of the specification.

[0025] In the following description, for the purpose of explanation rather than limitation, specific details such as specific system structures, interfaces, and technologies are provided to facilitate a thorough understanding of the present application.

[0026] The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there may be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the previously associated objects are in an "or" relationship. In addition, "many" in this article means two or more than two. In addition, the term "at least one" in this article means any combination of at least two of any one or more of a plurality of, for example, including at least one of A, B, and C, can mean including any one or more elements selected from the set consisting of A, B and C.

[0027] See also Figure 1 , Figure 1 : is a flowchart of the training method of the image detection model of the embodiment of the present application. Specifically, it may include the following steps:

[0028] Step S110: obtaining first docking posture deviations between the first automated guided vehicle and several target docking objects, and obtaining second docking position deviations between the second automated guided vehicle and several target docking objects, wherein both the first automated guided vehicle and the second automated guided vehicle are provided with positioning devices for navigation.

[0029] This application is mainly used in the field of automated guided vehicles. By calculating the similarity between the first docking posture deviation of several target docking objects that the first automated guided vehicle has docked with and the second docking position deviation of several target docking objects that other second automated guided vehicles have docked with, it is possible to confirm whether an abnormality has occurred in the positioning device installed on the first automated guided vehicle. The docking object may be a shelf, an operating handle, a socket, etc., which are not specifically limited here. Therefore, after the automated guided vehicle completes docking with the docking object, it can carry the docking object, transmit data, and other operations.

[0030] Among them, the target docking objects are all docking objects that the first automated guided vehicle and the second automated guided vehicle have docked with, that is, the target docking objects are the intersection of the docking objects that the first automated guided vehicle and the second automated guided vehicle have docked with. Furthermore, in the case of multiple second automated guided vehicles, the number of target docking objects corresponding to different second automated guided vehicles may be the same or different. Therefore, the positioning similarity between the first automated guided vehicle and different second automated guided vehicles is also different.

[0031] For example, the first automated guided vehicle has docked with docking objects numbered 1, 2, 3, 4, and 5, the second automated guided vehicle has docked with docking objects numbered 2, 3, 4, 5, and 6, and the third automated guided vehicle has docked with docking objects numbered 1, 4, 5, and 7. Therefore, for the first automated guided vehicle and the second automated guided vehicle, the target docking objects are docking objects numbered 2, 3, 4, and 5; for the first automated guided vehicle and the third automated guided vehicle, the target docking objects are docking objects numbered 1, 4, and 5; and for the second automated guided vehicle and the third automated guided vehicle, the target docking objects are docking objects numbered 4 and 5.

[0032] In some embodiments, to obtain the first docking posture deviation, the first historical docking posture deviation between the first automated guided vehicle and the target docking object at at least one first historical moment may be obtained for each target docking object, and the central tendency characterization value corresponding to each first historical docking posture deviation may be obtained as the first docking posture deviation. The central tendency characterization value may be the median, mean, mode, variance, etc. selected from each first historical docking posture deviation, and is not specifically limited here.

[0033] Specifically, before obtaining the first historical docking posture deviation of the automatic guided vehicle and the target docking object at at least one first historical moment, it is necessary to save and package the historical docking posture deviation of each automatic guided vehicle and the docking object at each historical moment, and upload the packaged file to local storage or cloud storage. Therefore, when obtaining the first historical docking posture deviation, according to the label of the target docking object and the first automatic guided vehicle, the corresponding storage file of the first automatic guided vehicle can be queried in the local storage or cloud storage, and then the historical docking posture deviation corresponding to the target docking object is queried in the storage file as the first historical docking posture deviation. After statistically summing up the first historical docking posture deviations, the sum value obtained is divided by the total number of the first historical docking posture deviations to obtain the central tendency characterization value corresponding to each first historical docking posture deviation, and use it as the first docking posture deviation. Among them, the label of the target docking object and the first automatic guided vehicle can be a number or other unique data representation, which is not specifically limited here.

[0034] Similarly, to obtain the second docking posture deviation, the second historical docking posture deviation between the second automated guided vehicle and the target docking object at at least one second historical moment can be obtained for each target docking object, and the central tendency characterization value corresponding to each second historical docking posture deviation can be obtained as the second docking posture deviation. Among them, the central tendency characterization value can be the median, mean, mode, variance, etc. selected from each second historical docking posture deviation, which is not specifically limited here.

[0035] Specifically, before obtaining the second historical docking posture deviation between the second automated guided vehicle and the target docking object at at least one second historical moment, it is necessary to save and package the historical docking posture deviation generated by the second automated guided vehicle during the docking process at each historical moment in the past, and store the packaged file in the storage space under the name of the second automated guided vehicle. Therefore, when the second docking posture deviation is subsequently obtained, the storage space corresponding to the second automated guided vehicle is queried according to the label of the target docking object and the second automated guided vehicle, and then the historical docking posture deviation corresponding to the target docking object is queried in the storage space as the second historical docking posture deviation. Then, the second historical docking posture deviation with the largest number of occurrences is selected from each second historical docking posture deviation as the second docking posture deviation. Among them, the label of the target docking object and the second automated guided vehicle can be a number or other unique data representation, which is not specifically limited here.

[0036] It can be understood that the first docking posture deviation and the second docking posture deviation may be obtained by the same acquisition method or by different acquisition methods, which are not specifically limited here.

[0037] From the above content, it can be seen that the first docking posture deviation is obtained by the first historical docking posture deviations, and the second docking posture deviation is obtained by the second historical docking posture deviations. Therefore, in order to accurately obtain the first docking posture deviation and the second docking posture deviation, the generation of the first historical docking posture deviation and the second historical docking posture deviation is very important.

[0038] In a specific embodiment, in order to obtain the first historical docking posture deviation or the second historical docking posture deviation, the target docking posture when the current automated guided vehicle moves to the target position point according to the planned path can be obtained first, wherein the current automated guided vehicle is the first automated guided vehicle corresponding to the first historical docking posture deviation or the second automated guided vehicle corresponding to the second historical docking posture deviation, and the planned path is generated based on the collected data of the positioning device on the current automated guided vehicle, and the target position point represents the position of the current automated guided vehicle docking with the target docking object. Afterwards, the image acquisition device on the current automated guided vehicle is used to determine the actual docking posture required for the current automated guided vehicle to dock with the target docking object. Finally, the posture deviation between the actual docking posture and the target docking posture is obtained, and the posture deviation is used as the first historical docking posture deviation or the second historical docking posture deviation.

[0039] For example, the positioning device can be a laser radar, which uses the laser radar to scan the surrounding environment and the target docking object to obtain point cloud data. The current automatic guided vehicle generates the center coordinates of the target docking object based on the point cloud data, and plans the planned path from the current position of the current automatic guided vehicle to the target docking object based on the center coordinates of the target docking object, and obtains the target docking posture when the current automatic guided vehicle moves to the target position point according to the planned path, and tracks the planned path. When the current automatic guided vehicle is about to reach the target position point of the planned path, the image acquisition device on the current automatic guided vehicle is turned on, and the image of the target docking object is acquired, and the target position point of the planned path is modified according to the acquired image, and the path is re-planned and the new planned path is tracked to obtain the actual docking posture of the current automatic guided vehicle reaching the target position point of the new planned path. Finally, the actual docking posture is compared with the target docking posture to obtain the posture deviation, and the posture deviation is used as the first historical docking posture deviation or the second historical docking posture deviation.

[0040] In another specific embodiment, in order to obtain the first historical docking posture deviation or the second historical docking posture deviation, when the current automatic guided vehicle moves to the target position point according to the planned path, the target docking object can be photographed using an image acquisition device to obtain a target docking image, wherein the current automatic guided vehicle is the first automatic guided vehicle corresponding to the first historical docking posture deviation or the second automatic guided vehicle corresponding to the second historical docking posture deviation, and the planned path is generated based on the acquisition data of the positioning device on the current automatic guided vehicle, and the target position point represents the position of the current automatic guided vehicle docking with the target docking object. Afterwards, the target docking image is analyzed to obtain the docking posture deviation between the current automatic guided vehicle and the target docking object, and the posture deviation is used as the first historical docking posture deviation or the second historical docking posture deviation.

[0041] For example, the planned path is the default path, and when the current automated guided vehicle moves to the target position point according to the planned path, the target docking object is photographed by the image acquisition device to obtain the target docking image. After that, the current automated guided vehicle analyzes the target docking image to obtain the center coordinates of the target docking object, and compares the coordinate position of the current automated guided vehicle at this time with the center coordinates of the target docking object, thereby obtaining the docking posture deviation between the current automated guided vehicle and the target docking object, and uses the posture deviation as the first historical docking posture deviation or the second historical docking posture deviation.

[0042] After the first docking posture deviation and each second docking posture deviation are obtained, the positioning similarity between the first automated guided vehicle and the second automated guided vehicle may be obtained according to the first docking posture deviation and each second docking posture deviation.

[0043] Step S120: Determine the positioning similarity between the first automated guided vehicle and the second automated guided vehicle based on each first docking posture deviation and each second docking posture deviation.

[0044] In some embodiments, a first representative posture deviation of the first automated guided vehicle may be determined based on each first docking posture deviation, and a second representative posture deviation corresponding to the second automated guided vehicle may be determined based on each second docking posture deviation. Afterwards, the first posture deviation similarity between the first representative posture deviation and the second representative posture deviation is obtained as the positioning similarity.

[0045] For example, after obtaining each first docking posture deviation and each second docking posture deviation, the variance of each first docking posture deviation is calculated to obtain the first representative posture deviation of the first automated guided vehicle; the variance of each second docking posture deviation is calculated to obtain the second representative posture deviation corresponding to the second automated guided vehicle. Afterwards, the similarity between the first representative posture deviation and the second representative posture deviation is calculated using the Euclidean distance method to obtain the first posture deviation similarity, and the first posture deviation similarity is used as the positioning similarity.

[0046] For another example, after obtaining each first docking posture deviation and each second docking posture deviation, each first docking posture deviation can be clustered to obtain a first cluster center, and the first cluster center is used as the first representative posture deviation of the first automatic guided vehicle, and then each second docking posture deviation is clustered to obtain a second cluster center, and the second cluster center is used as the second representative posture deviation corresponding to the second automatic guided vehicle. Afterwards, the cosine similarity calculation is performed on the first representative posture deviation and the second representative posture deviation to obtain the first posture deviation similarity, and the first posture deviation similarity is used as the positioning similarity.

[0047] It is understandable that, in addition to the above-mentioned cosine similarity and Euclidean distance calculation methods, the positioning similarity calculation method may also use methods such as Jaccard similarity coefficient and Chebyshev distance, which are not specifically limited here.

[0048] In other embodiments, based on each first docking posture deviation and each second docking posture deviation, the similarity of the second posture deviation of each target docking object corresponding to the first automated guided vehicle and the second automated guided vehicle can be obtained. Then, the central tendency representation value of the similarity of the second posture deviation corresponding to each target docking object is used as the positioning similarity.

[0049] For example, the first automated guided vehicle and the second automated guided vehicle have both docked with target docking objects numbered 1 and 2. Therefore, in order to obtain the positioning similarity between the first automated guided vehicle and the second automated guided vehicle with respect to the target docking objects numbered 1 and 2, the first docking posture deviation and the second docking posture deviation of the first automated guided vehicle and the second automated guided vehicle corresponding to the target docking object numbered 1 can be obtained respectively, and then the cosine similarity of the first docking posture deviation and the second docking posture deviation can be calculated to obtain the second posture deviation similarity of the target docking object numbered 1. Similarly, the second posture deviation similarity of the target docking object numbered 2 can be calculated. Then the average value of the second posture deviation similarity of the target docking objects numbered 1 and 2 is obtained to obtain a central tendency characterization value, and the central tendency characterization value is used as the positioning similarity.

[0050] In some embodiments, in order to more accurately obtain the positioning similarity between the first automated guided vehicle and the second automated guided vehicle, the total number of target docking objects for docking between the first automated guided vehicle and the second automated guided vehicle can be subdivided, for example, different total numbers use different similarity calculation methods. Specifically, the total number of target docking objects is first obtained, and then the total number of target docking objects is subdivided; if the total number is less than the first value, the first representative posture deviation of the first automated guided vehicle is determined based on each first docking posture deviation, and the second representative posture deviation corresponding to the second automated guided vehicle is determined based on each second docking posture deviation, and then the first posture deviation similarity between the first representative posture deviation and the second representative posture deviation is obtained as the positioning similarity; if the total number is greater than or equal to the first value, the second posture deviation similarity of each representative docking object corresponding to the first automated guided vehicle and the second automated guided vehicle is obtained based on each first docking posture deviation and each second docking posture deviation, and then the central trend representation value of the second posture deviation similarity corresponding to each representative docking object is used as the positioning similarity. The central tendency characterization value of the second posture deviation similarities corresponding to each representative docking object is the average value of the second posture deviation similarities corresponding to each representative docking object.

[0051] In addition, when the total number is between the first value and the second value, several target docking objects are all used as representative docking objects, and the first value is less than the second value; when the total number is greater than the second value, the second value of target docking objects among the several target docking objects are selected as representative docking objects.

[0052] For example, the total number of target docking objects for the first automated guided vehicle to dock with the second automated guided vehicle is obtained to be n, the first value is N1, the second value is N2, N1 and N2 are default values ​​and N1<N2, n is subdivided, if n<N1, all docking objects that the first automated guided vehicle has docked with are taken as representative docking objects, and then the first docking posture deviation of each representative docking object is obtained, and then each first docking posture deviation is clustered to obtain the first clustering center center1, and the first clustering center center1 is used as the first representative posture deviation A, similarly, all docking objects that the second automated guided vehicle has docked with are taken as representative docking objects, and then the second docking posture deviation of each representative docking object is obtained, and then each second docking posture deviation is clustered to obtain the second clustering center center2, and the second clustering center center2 is used as the second representative posture deviation B, and then the cosine similarity between the first representative posture deviation A and the second representative posture deviation B is calculated, wherein the calculation formula of the cosine similarity is:

[0053]

[0054] If N1≤n≤N2, then several target docking objects are all taken as representative docking objects, and the first docking posture deviations and the second docking posture deviations corresponding to the representative docking objects are obtained. Then, according to the representative docking objects, the first docking posture deviations and the second docking posture deviations are matched one by one, and the similarity of the second posture deviations is calculated. For example, the first docking posture deviations and the second docking posture deviations corresponding to the representative docking object numbered 1 are matched one by one, and the similarity of the second posture deviations between the first docking posture deviations and the second docking posture deviations corresponding to the representative docking object numbered 1 is calculated; the first docking posture deviations and the second docking posture deviations corresponding to the representative docking object numbered 2 are matched one by one, and the similarity of the second posture deviations between the first docking posture deviations and the second docking posture deviations corresponding to the representative docking object numbered 2 is calculated. Then, the average value of the similarities of the second posture deviations corresponding to the representative docking objects is calculated, and the average value is the central tendency representation value, and the average value is used as the positioning similarity.

[0055] If n>N2, then select the second number of target docking objects from the target docking objects as representative docking objects, that is, randomly select N2 target docking objects as representative docking objects, and then obtain the first docking posture deviations and the second docking posture deviations corresponding to the representative docking objects, and then according to the representative docking objects, make one-to-one correspondence between the first docking posture deviations and the second docking posture deviations, and calculate the similarity of the second posture deviations, and then calculate the average value of the similarities of the second posture deviations corresponding to the representative docking objects, which is the central tendency characterization value, and the average value is used as the positioning similarity.

[0056] Step S130: using the positioning similarity, determine whether there is an abnormality in the positioning device of the first automated guided vehicle.

[0057] In some embodiments, after the positioning similarity is obtained, the positioning similarity can be compared with a preset normal value. If it is equal to the preset normal value, it means that the positioning device of the first automatic guided vehicle is normal; otherwise, it means that there is an abnormality in the positioning device of the first automatic guided vehicle.

[0058] In other embodiments, in order to make the final result more accurate and stable, the first automated guided vehicle and multiple second automated guided vehicles may be similarly calculated to obtain multiple positioning similarities, and then the positioning device of the first automated guided vehicle may be judged to be abnormal based on the multiple positioning similarities. Specifically, refer to steps S131 to S133.

[0059] Step S131: Filtering a number of positioning similarities that meet the similarity requirement from the positioning similarities corresponding to the plurality of second automated guided vehicles.

[0060] In some embodiments, the similarity requirement is that the positioning similarity is greater than a first similarity threshold. Specifically, a logic function may be used for screening, and the logic function formula is:

[0061]

[0062] Wherein, T is the first similarity threshold, t is the positioning similarity; when the positioning similarity is greater than or equal to the first similarity threshold, the logic function R(t, T) outputs 1; when the positioning similarity is less than the first similarity threshold, the logic function R(t, T) outputs 0.

[0063] Step S132: Count a number of positioning similarities to obtain representative positioning similarities between the first automated guided vehicle and the plurality of second automated guided vehicles.

[0064] For the plurality of positioning similarities obtained after screening, the sum of the plurality of positioning similarities is obtained, and the ratio of the sum of the positioning similarities to the number of second automated guided vehicles is taken as the representative positioning similarity.

[0065] Step S133: Based on the representative positioning similarity, determine whether there is an abnormality in the positioning device of the first automated guided vehicle.

[0066] In some embodiments, the representative positioning similarity is judged using the second similarity threshold to determine whether the positioning device of the first automated guided vehicle has an abnormality. If the representative positioning similarity is less than the second similarity threshold, it is determined that the positioning device of the first automated guided vehicle has an abnormality. The second similarity threshold can be a value such as 0.3, 0.5, Π, etc., which is not specifically limited here.

[0067] In a specific implementation scenario, the positioning device is a laser radar, the image acquisition device is a camera, and the docking object is a shelf. And the docking process of the automatic guided vehicle can be converted into a digital space, that is, in the docking process of the automatic guided vehicle numbered a j The automatic guided vehicle sends the docking number s k When the docking task is to dock the shelf, the docking task is saved in the first input space X, and then the docking posture deviation is obtained through the corresponding first output space Y. The first input space is X = {(a1, s1), ..., (a1, s N ),…,(a j ,s k ), …(a M , s1),…,(a M ,s N )}, the first output space is M is the number of automated guided vehicles, N is the number of shelves, and 1≤j≤M, 1≤k≤N. Input x=(a j ,s k )∈X represents the number a j AGV docking number s k The output y∈Y represents the position deviation between the actual docking position and the target docking position when the automatic guided vehicle docks the shelf. From the first input space to the first output space, there is the following function:

[0068] y=f(x) (3).

[0069] With number a j The automatic guided vehicle docking number is s k Take the shelf of a for example, the specific process of collecting docking task data is as follows: First, at a certain historical moment, the shelf numbered a j The automated guided vehicle receives the docking number s k After the task of the shelf, numbered a j The automated guided vehicle uses its laser radar to detect the surrounding environment and the k The shelf is scanned to obtain point cloud data, numbered a jThe automated guided vehicle generates a number based on the point cloud data k The center coordinates of the shelf, and according to the number s k The center coordinates of the shelf are planned from number a j The current position of the automated guided vehicle to the number s k The planning path of the shelf and the number a j The target docking posture when the AGV moves to the target location point according to the planned path and tracks the planned path. j When the AGV is about to reach the target position of the planned path, the j The camera on the automated guided vehicle is k The image of the shelf is collected, and the target position point of the planned path is modified according to the collected image, and the path is re-planned and the new planned path is tracked to obtain a numbered a j The actual docking posture of the automated guided vehicle when it reaches the target position of the new planned path. Finally, the function y obtains the posture deviation bias according to the actual docking posture and the target docking posture. sample Among them, the posture deviation bias sample As f((a j ,s k )) of the historical docking posture deviation. In order to ensure the stability of the model output results, the data about f((a j ,s k )) of several historical docking posture deviations. Then obtain the number a j Automated guided vehicle about number s k The average value of the shelf is used as the docking posture deviation, and the formula is: Where L is f((a j ,s k ))’s number of historical docking pose deviations.

[0070] The above can obtain the docking posture deviation of each automated guided vehicle. When obtaining the positioning similarity between different automated guided vehicles, the process can also be converted into a digital space, that is, the positioning similarity between different automated guided vehicles is obtained by using the second input space C and the second output space Z, where the second input space is C = {(a1, a1), ..., (a1, a M ),…,(a j , a k ), …(a M , a1),…,(a M , a M )}, the second output space is Z = [0, 1] ∈ R, where 1 ≤ j ≤ M, 1 ≤ k ≤ M, and the input c = (a j , a k )∈C represents the number aj The automated guided vehicle and number a k The output z∈Z represents the positioning similarity between the automatic guided vehicles. From the second input space to the second output space, there is the following function:

[0071] z=g(c) (4).

[0072] where g((a j , a k ))=g((a k , a j )), g(a j , a j )=1.

[0073] To solve the number a j The first automated guided vehicle and numbered a k Taking the positioning similarity between the second automated guided vehicle as an example, the detailed process is as follows:

[0074] (a) Filter out the items with number a j The first automated guided vehicle and numbered a k The second automated guided vehicle has docked with the shelf Right now and The number of samples is not zero, where i=1,...,n, where n is the total number of target shelves.

[0075] (b) Different processing is performed according to the total number n of the target shelves, where the first value N1 and the second value N2 are the minimum number of shelves and the maximum number of shelves set manually, respectively:

[0076] i. If n < N1, obtain all the shelves that the first automated guided vehicle has docked with as representative shelves, then obtain the first docking posture deviation of each representative shelf, and then cluster the first docking posture deviations to obtain the first cluster center The first cluster center As the first representative posture deviation A; similarly, all the shelves that the second automated guided vehicle has docked with are obtained as representative shelves, and then the second docking posture deviations of each representative shelf are obtained, and then the second docking posture deviations are clustered to obtain the second cluster center The first cluster center As the second representative posture deviation B. Then, the cosine similarity between the first representative posture deviation A and the second representative posture deviation B is obtained to obtain the first posture deviation similarity as the positioning similarity, where the calculation formula of the cosine similarity is the above formula 1,

[0077] ii. If N1≤n≤N2, then take n shelves as representative shelves and find the first AGV docking shelf The first pose deviation data The same operation is performed to obtain the second posture deviation data corresponding to the second automated guided vehicle Obtain and The cosine similarity between them is used to obtain the second posture deviation similarity. Then the second posture deviation similarities corresponding to each representative shelf are averaged to obtain the positioning similarity.

[0078] iii. If n>N2, randomly select N2 shelves from the n shelves and then perform the operation in (ii).

[0079] After that, a certain frequency (such as half a day or a day) can be set to regularly calculate the positioning similarity between the automated guided vehicles, and then the lasers of the M automated guided vehicles are judged to be abnormal. j Take the automatic guided vehicle laser as an example to find out whether it is abnormal. j The similarity between the AGV and other AGVs (except itself) is obtained as [g((a j , a1)),...,g((a j , a i )),...,g((a j , a M ))], where i = 1, ..., M, and i ≠ j. Define the logic function And set the first similarity threshold Thre. j The similarity between the automated guided vehicle and other automated guided vehicles (except itself) is brought into the following function: Where i≠j. Set the second similarity threshold to 0.5. If result<0.5, the judgment number is a. j The laser radar of the automated guided vehicle is abnormal; if result ≥ 0.5, the judgment number is a j The laser radar of the automated guided vehicle is normal.

[0080] See also Figure 2 , Figure 2 This is a flow chart of another embodiment of the abnormality detection method of the present application. Specifically, it may include the following steps:

[0081] Step S210: Based on the first automated guided vehicle, the storage space is traversed to select other automated guided vehicles that meet the comparison conditions as the second automated guided vehicle.

[0082] In some embodiments, the comparison condition can be set to at least one docking object that other automated guided vehicles in the storage space have docked with is a docking object that the first automated guided vehicle has docked with. For example, if the first automated guided vehicle has docked with docking objects numbered 1, 2, 3, and 4, then a traversal is performed in the storage space to query the docking objects numbered 1, 3, and 5 that the automated guided vehicle numbered P has docked with, and the docking objects numbered 5, 6, and 7 that the automated guided vehicle numbered Q has docked with. The docking objects that the first automated guided vehicle has docked with are compared with the docking objects that the automated guided vehicles numbered P and Q have docked with, respectively, to obtain a comparison result. The comparison result is that the docking objects that the automated guided vehicle numbered Q has docked with do not include the docking objects that the first automated guided vehicle has docked with, and the automated guided vehicle numbered P and the first automated guided vehicle have both docked with docking objects numbered 1 and 3, so the automated guided vehicle numbered P can be used as the second automated guided vehicle. In this embodiment, other automated guided vehicles that have no intersection with the docking object that the first automated guided vehicle has docked with may be regarded as interference items and excluded, thereby reducing the amount of calculation.

[0083] Step S220: obtaining first historical docking posture deviations of the first automated guided vehicle and the plurality of target docking objects, and obtaining second historical docking position deviations of the second automated guided vehicle and the plurality of target docking objects, wherein both the first automated guided vehicle and the second automated guided vehicle are provided with a positioning device for navigation.

[0084] In some embodiments, after determining the second automated guided vehicle, the first historical docking posture deviations of the first automated guided vehicle and several target docking objects can be obtained in the storage space according to the label of the first automated guided vehicle and the label of the second automated guided vehicle, and the second historical docking position deviations of the second automated guided vehicle and several target docking objects can be obtained.

[0085] Step S230: Calculate the average value of each first historical docking posture deviation to obtain a first docking posture deviation, and calculate the average value of each second historical docking posture deviation to obtain a second docking posture deviation.

[0086] In some embodiments, since the first automated guided vehicle may have docked with the same target docking object multiple times, in order to reduce the amount of calculation, the first automated guided vehicle may first be averaged over multiple first historical docking posture deviations at the same target docking object to obtain the first docking posture deviation of the first automated guided vehicle at the target docking object. Similarly, the first docking posture deviation of the first automated guided vehicle at other target docking objects may also be obtained. Similarly, the first docking posture deviation of the second automated guided vehicle at each target docking object may also be obtained.

[0087] Step S240: Determine the positioning similarity between the first automated guided vehicle and the second automated guided vehicle based on each first docking posture deviation and each second docking posture deviation.

[0088] This step is the same as the above step S120 and will not be described again here.

[0089] Step S250: using the positioning similarity, determine whether there is an abnormality in the positioning device of the first automated guided vehicle.

[0090] This step is the same as the above step S130 and will not be described again here.

[0091] This solution evaluates the similarity between positioning devices by connecting the posture deviation data, and determines that the positioning device with low similarity with other positioning devices is abnormal. This solution can also feedback the positioning status of the automatic guided vehicle to the on-site maintenance personnel, which helps to timely troubleshoot problems on site.

[0092] In addition, this solution calculates the positioning similarity based on the docking posture deviation data. Positioning similarity is part of measuring the consistency of multiple vehicles. Applying this similarity to various business processes will help improve the consistency of multiple vehicles in various business processes of automated guided vehicles.

[0093] Those skilled in the art will appreciate that, in the above method of specific implementation, the order in which the steps are written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of the steps should be determined by their functions and possible internal logic.

[0094] See also Figure 3 , Figure 3 It is a schematic diagram of the framework of an embodiment of the abnormality detection device of the present application. The abnormality detection device 300 includes: an acquisition module 310, a calculation module 320 and a judgment module 330. The acquisition module 310 executes to obtain the first docking posture deviations between the first automated guided vehicle and several target docking objects, and obtains the second docking position deviations between the second automated guided vehicle and several target docking objects, respectively. The first automated guided vehicle and the second automated guided vehicle are both provided with a positioning device for navigation. The calculation module 320 determines the positioning similarity between the first automated guided vehicle and the second automated guided vehicle based on each first docking posture deviation and each second docking posture deviation. The judgment module 330 uses the positioning similarity to determine whether there is an abnormality in the positioning device of the first automated guided vehicle.

[0095] In some embodiments, the calculation module 320 determines the positioning similarity between the first automated guided vehicle and the second automated guided vehicle based on each first docking posture deviation and each second docking posture deviation, including: determining the first representative posture deviation of the first automated guided vehicle based on each first docking posture deviation, and determining the second representative posture deviation corresponding to the second automated guided vehicle based on each second docking posture deviation; obtaining the first posture deviation similarity between the first representative posture deviation and the second representative posture deviation as the positioning similarity; or, obtaining the second posture deviation similarity between the first automated guided vehicle and the second automated guided vehicle corresponding to each target docking object based on each first docking posture deviation and each second docking posture deviation; and taking the central trend representation value of the second posture deviation similarity corresponding to each target docking object as the positioning similarity.

[0096] In some embodiments, the calculation module 320 performs determination of the positioning similarity between the first automated guided vehicle and the second automated guided vehicle based on each first docking posture deviation and each second docking posture deviation, including: obtaining the total number of target docking objects; in response to the total number being less than a first value, determining the first representative posture deviation of the first automated guided vehicle based on each first docking posture deviation, and determining the second representative posture deviation corresponding to the second automated guided vehicle based on each second docking posture deviation; obtaining the first posture deviation similarity between the first representative posture deviation and the second representative posture deviation as the positioning similarity; in response to the total number being greater than or equal to the first value, obtaining the second posture deviation similarity between the first automated guided vehicle and the second automated guided vehicle corresponding to each representative docking object based on each first docking posture deviation and each second docking posture deviation; and taking the central trend representation value of the second posture deviation similarity corresponding to each representative docking object as the positioning similarity.

[0097] In some embodiments, before the calculation module 320 executes to obtain the similarity of the second posture deviations of the first automated guided vehicle and the second automated guided vehicle corresponding to each representative docking object based on each first docking posture deviation and each second docking posture deviation, it also includes: when the total number is between the first value and the second value, taking several target docking objects as representative docking objects, and the first value is less than the second value; when the total number is greater than the second value, selecting a second value of target docking objects from the several target docking objects as representative docking objects.

[0098] In some embodiments, the calculation module 320 determines the first representative posture deviation of the first automatic guided vehicle based on each first docking posture deviation, including: clustering each first docking posture deviation to obtain a first cluster center, and using the first cluster center as the first representative posture deviation; based on each second docking posture deviation, determines the second representative posture deviation corresponding to the second automatic guided vehicle, including: clustering each first docking posture deviation to obtain a second cluster center, and using the second cluster center as the second representative posture deviation of the second automatic guided vehicle.

[0099] In some embodiments, the acquisition module 310 executes a plurality of target docking objects, which are all docking objects that both the first automated guided vehicle and the second automated guided vehicle have docked with.

[0100] In some embodiments, the acquisition module 310 executes the acquisition of the first docking posture deviations between the first automatic guided vehicle and several target docking objects, including: for each target docking object, acquiring the first historical docking posture deviation between the first automatic guided vehicle and the target docking object at at least one first historical moment, and acquiring the central trend characterization value corresponding to each first historical docking posture deviation as the first docking posture deviation; acquiring the second docking position deviations between the second automatic guided vehicle and several target docking objects, including: for each target docking object, acquiring the second historical docking posture deviation between the second automatic guided vehicle and the target docking object at at least one second historical moment, and acquiring the central trend characterization value corresponding to each second historical docking posture deviation as the second docking posture deviation.

[0101] In some embodiments, the acquisition module 310 performs the steps of acquiring the first historical docking posture deviation or the second historical docking posture deviation, including: acquiring the target docking posture when the current automated guided vehicle moves to the target position point according to the planned path, wherein the current automated guided vehicle is the first automated guided vehicle corresponding to the first historical docking posture deviation or the second automated guided vehicle corresponding to the second historical docking posture deviation, the planned path is generated based on the collected data of the positioning device on the current automated guided vehicle, and the target position point represents the position of the current automated guided vehicle docking with the target docking object; and, using the image acquisition device on the current automated guided vehicle, determining the actual docking posture required for the current automated guided vehicle to dock with the target docking object; acquiring the difference between the actual docking posture and the target The step of obtaining the first historical docking posture deviation or the second historical docking posture deviation includes: when the current automated guided vehicle moves to the target position point according to the planned path, using an image acquisition device to shoot the target docking object to obtain a target docking image, wherein the current automated guided vehicle is the first automated guided vehicle corresponding to the first historical docking posture deviation or the second automated guided vehicle corresponding to the second historical docking posture deviation, the planned path is generated based on the acquisition data of the positioning device on the current automated guided vehicle, and the target position point represents the position of the current automated guided vehicle docking with the target docking object; the target docking image is analyzed to obtain the docking posture deviation between the current automated guided vehicle and the target docking object.

[0102] In some embodiments, the judgment module 330 executes that there are multiple second automated guided vehicles; using the positioning similarity, determining whether there is an abnormality in the positioning device of the first automated guided vehicle, including: screening a number of positioning similarities that meet the similarity requirements from the positioning similarities corresponding to the multiple second automated guided vehicles; counting the multiple positioning similarities to obtain a representative positioning similarity between the first automated guided vehicle and the multiple second automated guided vehicles; based on the representative positioning similarity, determining whether there is an abnormality in the positioning device of the first automated guided vehicle.

[0103] In some embodiments, the judgment module 330 executes the similarity requirement that the positioning similarity is greater than a first similarity threshold; and / or, counts a number of positioning similarities to obtain a representative positioning similarity between the first automated guided vehicle and multiple second automated guided vehicles, including: obtaining the sum of the several positioning similarities, and taking the ratio of the sum of the positioning similarities to the number of second automated guided vehicles as the representative positioning similarity; and / or, based on the representative positioning similarity, determines whether there is an abnormality in the positioning device of the first automated guided vehicle, including: in response to the representative positioning similarity being less than the second similarity threshold, determines that there is an abnormality in the positioning device of the first automated guided vehicle.

[0104] See also Figure 4 , Figure 44 is a schematic diagram of a framework of an embodiment of an electronic device 40 of the present application. The electronic device 40 includes a memory 41 and a processor 42 coupled to each other, and the processor 42 is used to execute program instructions stored in the memory 41 to implement the steps in any of the above-mentioned abnormality detection method embodiments. In a specific implementation scenario, the electronic device 40 may include but is not limited to: a microcomputer, a server, and in addition, the electronic device 40 may also include a mobile device such as a laptop computer and a tablet computer, which is not limited here.

[0105] Specifically, the processor 42 is used to control itself and the memory 41 to implement the steps in any of the above-mentioned abnormality detection method embodiments. The processor 42 can also be called a CPU (Central Processing Unit). The processor 42 may be an integrated circuit chip with signal processing capabilities. The processor 42 can also be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field-programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. In addition, the processor 42 can be implemented by an integrated circuit chip.

[0106] See also Figure 5 , Figure 5 1 is a schematic diagram of a framework of an embodiment of a computer-readable storage medium 50 of the present application. The computer-readable storage medium 50 stores program instructions 501 that can be executed by a processor, and the program instructions 501 are used to implement the steps in any of the above-mentioned abnormality detection method embodiments.

[0107] In some embodiments, the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0108] The above description of various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced to each other, and for the sake of brevity, they will not be repeated herein.

[0109] In the several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation described above is only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation, such as units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0110] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0111] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor (processor) to perform all or part of the steps of each implementation method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

Claims

1. An anomaly detection method, characterized in that: include: Acquiring first docking posture deviations between the first automated guided vehicle and a plurality of target docking objects, and acquiring second docking position deviations between the second automated guided vehicle and the plurality of target docking objects, wherein the first automated guided vehicle and the second automated guided vehicle are both provided with a positioning device for navigation; Determining a positioning similarity between the first automated guided vehicle and the second automated guided vehicle based on each of the first docking posture deviations and each of the second docking posture deviations; By using the positioning similarity, it is determined whether there is an abnormality in the positioning device of the first automated guided vehicle.

2. The method according to claim 1, characterized in that The determining, based on each of the first docking posture deviations and each of the second docking posture deviations, the positioning similarity between the first automated guided vehicle and the second automated guided vehicle comprises: Based on each of the first docking posture deviations, determine a first representative posture deviation of the first automated guided vehicle, and based on each of the second docking posture deviations, determine a second representative posture deviation corresponding to the second automated guided vehicle; obtain a first posture deviation similarity between the first representative posture deviation and the second representative posture deviation as the positioning similarity; or, Based on each of the first docking posture deviations and each of the second docking posture deviations, obtain the similarity of the second posture deviations of the first automated guided vehicle and the second automated guided vehicle corresponding to each of the target docking objects; and use the central trend representation value of the similarity of the second posture deviations corresponding to each of the target docking objects as the positioning similarity.

3. The method according to claim 1, characterized in that The determining, based on each of the first docking posture deviations and each of the second docking posture deviations, the positioning similarity between the first automated guided vehicle and the second automated guided vehicle comprises: Obtaining the total number of target docking objects; In response to the total being less than a first value, determining a first representative posture deviation of the first automated guided vehicle based on each of the first docking posture deviations, and determining a second representative posture deviation corresponding to the second automated guided vehicle based on each of the second docking posture deviations; obtaining a first posture deviation similarity between the first representative posture deviation and the second representative posture deviation as the positioning similarity; In response to the total being greater than or equal to the first value, based on each of the first docking posture deviations and each of the second docking posture deviations, obtain the similarity of the second posture deviations corresponding to each representative docking object between the first automated guided vehicle and the second automated guided vehicle; and use the central trend representation value of the second posture deviation similarities corresponding to each of the representative docking objects as the positioning similarity.

4. The method according to claim 3, characterized in that Before obtaining the second posture deviation similarity between the first automated guided vehicle and the second automated guided vehicle corresponding to each representative docking object based on each of the first docking posture deviations and each of the second docking posture deviations, the method further includes: In the case where the total number is between the first value and the second value, all of the target docking objects are used as the representative docking objects, and the first value is less than the second value; When the total number is greater than the second value, a second value of target docking objects among the plurality of target docking objects are selected as the representative docking objects.

5. The method according to claim 2 or 3, characterized in that: The determining, based on each of the first docking posture deviations, a first representative posture deviation of the first automated guided vehicle includes: Clustering each of the first docking posture deviations to obtain a first cluster center, and using the first cluster center as the first representative posture deviation; The determining, based on each of the second docking posture deviations, a second representative posture deviation corresponding to the second automated guided vehicle includes: Clustering is performed on each of the first docking posture deviations to obtain a second cluster center, and the second cluster center is used as a second representative posture deviation of the second automatic guided vehicle.

6. The method according to claim 1, characterized in that The plurality of target docking objects are all docking objects with which both the first automated guided vehicle and the second automated guided vehicle have docked.

7. The method according to claim 1, characterized in that The obtaining of first docking posture deviations between the first automated guided vehicle and a plurality of target docking objects respectively includes: For each of the target docking objects, obtaining first historical docking posture deviations between the first automated guided vehicle and the target docking object at at least one first historical moment, and obtaining a central tendency representation value corresponding to each of the first historical docking posture deviations as the first docking posture deviation; The obtaining of second docking position deviations between the second automated guided vehicle and the plurality of target docking objects respectively includes: For each target docking object, obtain the second historical docking posture deviation between the second automated guided vehicle and the target docking object at at least one second historical moment, and obtain the central trend characterization value corresponding to each of the second historical docking posture deviations as the second docking posture deviation.

8. The method according to claim 7, characterized in that The step of acquiring the first historical docking posture deviation or the second historical docking posture deviation comprises: Obtaining a target docking posture when the current automated guided vehicle moves to a target position point according to a planned path, wherein the current automated guided vehicle is a first automated guided vehicle corresponding to a first historical docking posture deviation or a second automated guided vehicle corresponding to a second historical docking posture deviation, the planned path is generated based on collected data of the positioning device on the current automated guided vehicle, and the target position point represents a position where the current automated guided vehicle docks with the target docking object; and Determine an actual docking posture required for the current automatic guided vehicle to dock with the target docking object by using an image acquisition device on the current automatic guided vehicle; Acquire a posture deviation between the actual docking posture and the target docking posture; Alternatively, the step of acquiring the first historical docking posture deviation or the second historical docking posture deviation includes: When the current automated guided vehicle moves to the target position point according to the planned path, the target docking object is photographed by using the image acquisition device to obtain a target docking image, wherein the current automated guided vehicle is the first automated guided vehicle corresponding to the first historical docking posture deviation or the second automated guided vehicle corresponding to the second historical docking posture deviation, the planned path is generated based on the acquisition data of the positioning device on the current automated guided vehicle, and the target position point represents the position of the current automated guided vehicle docking with the target docking object; The target docking image is analyzed to obtain a docking posture deviation between the current automatic guided vehicle and the target docking object.

9. The method according to claim 1, characterized in that: There are multiple second automated guided vehicles; and using the positioning similarity to determine whether the positioning device of the first automated guided vehicle is abnormal includes: Selecting a number of the positioning similarities that meet the similarity requirement from the positioning similarities corresponding to the plurality of the second automated guided vehicles; Counting a number of the positioning similarities to obtain representative positioning similarities between the first automated guided vehicle and a plurality of the second automated guided vehicles; Based on the representative positioning similarity, it is determined whether there is an abnormality in the positioning device of the first automated guided vehicle.

10. The method according to claim 9, characterized in that The similarity requirement is that the positioning similarity is greater than a first similarity threshold; And / or, the counting of a plurality of the positioning similarities to obtain representative positioning similarities between the first automated guided vehicle and a plurality of the second automated guided vehicles includes: Obtaining a sum of a number of the positioning similarities, and taking a ratio of the sum of the positioning similarities to the number of the second automated guided vehicles as the representative positioning similarity; And / or, determining whether there is an abnormality in the positioning device of the first automated guided vehicle based on the representative positioning similarity includes: In response to the representative positioning similarity being less than a second similarity threshold, it is determined that an abnormality exists in the positioning device of the first automated guided vehicle.

11. An electronic device, characterized in that: It comprises a memory and a processor coupled to each other, wherein the processor is used to execute program instructions stored in the memory to implement the abnormality detection method according to any one of claims 1 to 10.

12. A computer-readable storage medium having program instructions stored thereon, characterized in that: When the program instructions are executed by a processor, the abnormality detection method according to any one of claims 1 to 10 is implemented.

Images