Immune mechanism fault diagnosis method and device for intelligent robot and medium

Abstract

According to the immune mechanism fault diagnosis method and device for the intelligent robot and the medium, the memory cell set with the fault recognition capability is generated by collecting the fault data sample of the intelligent robot, and the robot fault diagnosis model based on the immune mechanism is constructed; and performing fault early warning and diagnosis on robot equipment by using the robot fault diagnosis model. According to the method, starting from existing fault data characteristics of the robot, small sample data are utilized, a biological immune mechanism is used for reference, an autologous and allosome recognition mechanism of a biological immune system is simulated, a memory cell set is generated through a memory cell optimization strategy, and recognition, diagnosis and early warning of robot faults are achieved. In the fault early warning and diagnosis process of the intelligent robot, the situation that an early warning and diagnosis model is excessively sensitive and ineffective in recognition is avoided, and the defect that a traditional training model needs to depend on a large amount of fault sample information and artificial experience is overcome.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis, and in particular to an immune mechanism fault diagnosis method, device and storage medium for an intelligent robot. Background Art

[0002] With the development of 5G-based Internet of Things technology and the popularization of service robots, robots are the result of the deep integration of AI and high-end manufacturing. The intelligent upgrade of multiple industries will rely on the expansion of the robot field. Using new technologies to realize robot fault diagnosis has become an important research field. As an IoT device terminal, intelligent robots use 5G communication technology to report data and issue tasks. Downlink data supports robot business scenarios, while uplink data can be used as a warning and diagnosis basis for robot operation failures.

[0003] An effective fault warning and diagnosis mechanism plays a vital role in comprehensively ensuring the continuous and stable operation of robots, improving their work efficiency and extending their service life. However, the low recognition rate and overly sensitive fault warning and diagnosis models have seriously restricted the development and application of robot fault warning and diagnosis technology. At the same time, relying on the research results of artificial intelligence in other fields, a large number of model training methods for robot equipment fault warning and diagnosis have emerged, such as those based on neural networks, decision trees, support vector machines, K nearest neighbors, and deep learning. However, most of these methods belong to batch learning methods, which use a limited number of learning and training methods and are based on sufficient fault sample information, resulting in that they can only identify and warn more frequent faults during the diagnosis process. Moreover, these traditional fault diagnosis methods are overly dependent on manual experience and expert knowledge, and have problems such as low diagnostic efficiency and high misdiagnosis rate.

[0004] Robots have a variety of service scenarios, types, and brands. How to train a fault diagnosis model that is applicable to robots of multiple types, scenarios, and brands has become an important research direction.

[0005] Since the environments and scenarios in which robots are deployed and used are different, and the environment is unpredictable due to human activities, it is impossible to enumerate the occurrence of robot accidents. It is impossible for us to simulate the data generated by each accident to train the recognition model. Collecting data in this way is extremely expensive and unrealistic, which makes it extremely difficult to build a large-scale, highly reliable fault data set.

[0006] Glossary The Internet of Things (IoT) is a network that connects things through information carriers such as the Internet and traditional communication networks.

[0007] Robot: A robot that integrates multiple sensors, actuators, control systems and artificial intelligence algorithms and can perform tasks autonomously in complex environments.

[0008] Immunity-based learning: It is a learning algorithm inspired by the biological immune system. It simulates the diversity, adaptability, memory and learning ability of the immune system.

[0009] Fault diagnosis: It is a technical process of identifying problems and faults in a system or equipment. Summary of the invention

[0010] The present invention provides an immune mechanism fault diagnosis method, device and storage medium for intelligent robots, which can solve at least one of the technical problems in the background technology.

[0011] To achieve the above object, the present invention adopts the following technical solutions: An immune mechanism fault diagnosis method for intelligent robots, which generates a memory cell set with fault identification capability by collecting fault data samples of intelligent robots, and constructs a robot fault diagnosis model based on immune mechanism; The robot fault diagnosis model is used to provide fault warning and diagnosis for robot equipment.

[0012] Furthermore, the model training process of the robot fault diagnosis model based on the immune mechanism includes: ① Antigen set classification calculation of individual recognition threshold The antigen set Ag = {Ag1, Ag2, …, Agn} containing n individuals is divided into m sub-antigen sets TAgm, each of which contains antigens of the same type, that is, TAg = {TAg1, TAg2, …, TAgm} (m = 1,2,3,...), and the individual recognition threshold RTim of each type of antigen is calculated by formula (1); (1) In the formula, the ƒd function represents the affinity between the mth antigen and other antigens. (2) Among them, δm is the cytokine concentration of the antibody with an affinity of 1 for the mth antigen; ② Vaccine antigen clone mutation proliferation to generate antibody set From the mth antigen set TAgm, the antigen that can promote the highest cytokine concentration is selected as the vaccine antigen Agv. (3) In the formula, the optimization rules for the two parameters σ and ɛ are as follows: (4) Differentiate the clones and mutations of the vaccine antigen Agv and generate targeted antibody sets Ab. (5) In the formula, κ is the isotype matrix of the vaccine antigen Agv, represents the mutation rate, and takes a random number in the range of (-1, 1), and the α parameter represents the length of cell mutation; ③. Screening of high affinity antibodies When the affinity function ƒt(Ab, Ag) is greater than the antibody recognition threshold, ƒt(Ab, Ag) = 1, otherwise ƒt(Ab, Ag) = 0, that is, ƒt(Ab, Ag) = 1 means that the antigen Ag is covered by the antibody cell Ab; the antibody cell (Ab, δ) that can cover the largest number of antigens is selected as the memory cell M through the affinity function, as shown in formula (6); (6) Randomly select one of the non-empty memory cell sets as the initial affinity-mature memory cell M. At this time, antigens that cannot be covered by memory cells will appear in TAgm, and these remaining antigens will form a new antigen set TAgmr; (7) ④. Generate memory cell sets The process of generating memory cell set TM is a cyclic process of repeated clonal mutation and proliferation of the remaining antigen set to generate affinity mature antibodies. On the other hand, the present invention further discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the above method.

[0013] On the other hand, the present invention further discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.

[0014] From the above technical solution, it can be seen that the immune mechanism fault diagnosis method for intelligent robots of the present invention starts from the existing fault data characteristics of the robot, uses small sample data, draws on the biological immune mechanism, simulates the self- and foreign-recognition mechanism of the biological immune system, and proposes an immune learning model training method for robot faults. This method generates a memory cell set by optimizing the memory cell strategy to achieve the recognition, diagnosis and early warning of robot faults.

[0015] In the process of fault warning and diagnosis of intelligent robots, the present invention avoids the oversensitivity and invalid recognition of the warning diagnosis model, and overcomes the disadvantage that the traditional training model needs to rely on a large amount of fault sample information and manual experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flowchart of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0018] Starting from the characteristics of existing robot fault data, this paper uses small sample data, draws on biological immune mechanisms, simulates the self- and foreign-recognition mechanisms of the biological immune system, and proposes an immune learning model training method for robot faults. This method generates a memory cell set by optimizing the memory cell strategy to achieve recognition, diagnosis and early warning of robot faults.

[0019] The following are specific instructions: The model training method is to generate a memory cell set with fault recognition ability through fault data samples (antigens) to build a robot fault diagnosis model based on immune mechanism. Figure 1 As shown in the figure, the model training process includes: ① Antigen set classification calculation of individual recognition threshold The antigen set Ag = {Ag1, Ag2, …, Agn} containing n individuals is divided into m sub-antigen sets TAgm, each of which contains antigens of the same type, i.e., TAg = {TAg1, TAg2, …, TAgm} (m = 1,2,3,...), and the individual recognition threshold RTim of each type of antigen is calculated by formula (1).

[0020] (1) In the formula, the ƒd function represents the affinity between the mth antigen and other antigens. (2) Wherein, δm is the cytokine concentration of the antibody with affinity of 1 for the mth antigen.

[0021] ② Vaccine antigen clone mutation proliferation to generate antibody set From the mth antigen set TAgm, the antigen that can promote the highest cytokine concentration is selected as the vaccine antigen Agv. (3) In the formula, the optimization rules for the two parameters σ and ɛ are as follows: (4) Differentiate the clones and mutations of the vaccine antigen Agv and generate targeted antibody sets Ab. (5) Where κ is the isotype matrix of the vaccine antigen Agv, represents the mutation rate, and takes a random number in the range of (-1, 1), and the α parameter represents the length of cell mutation.

[0022] ③. Screening of high affinity antibodies When the affinity function ƒt(Ab, Ag) is greater than the antibody recognition threshold, ƒt(Ab, Ag) = 1, otherwise ƒt(Ab, Ag) = 0, that is, ƒt(Ab, Ag) = 1 indicates that the antigen Ag is covered by the antibody cell Ab. The antibody cell (Ab, δ) that can cover the largest number of antigens is selected as the memory cell M through the affinity function, as shown in formula (6).

[0023] (6) One of the non-empty memory cells is randomly selected as the initial affinity-mature memory cell M. At this time, antigens that cannot be covered by memory cells will appear in TAgm, and these remaining antigens will form a new antigen set TAgmr.

[0024] (7) ④. Generate memory cell sets The process of generating memory cell set TM is a cyclic process of repeated clonal mutation and proliferation of the remaining antigen set to generate affinity mature antibodies.

[0025] In summary, most of the existing methods for fault warning and diagnosis of robots are overly sensitive and have invalid recognition during the process of fault warning and diagnosis, and require a large amount of fault sample information when training the model.

[0026] The present invention uses the characteristics of robot fault data, small sample data, and biological immune mechanisms to propose an immune mechanism fault diagnosis method for intelligent robots. This method generates a memory cell set by optimizing the memory cell strategy, which can realize small sample model training and meet the necessary accuracy when performing robot fault identification, diagnosis, and early warning.

[0027] On the other hand, the present invention further discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the above method.

[0028] On the other hand, the present invention further discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.

[0029] In another embodiment provided in the present application, a computer program product including instructions is also provided, which, when executed on a computer, enables the computer to execute any of the immune mechanism fault diagnosis methods for intelligent robots in the above embodiments.

[0030] It is understandable that the system, device and storage medium provided in the embodiments of the present invention correspond to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant contents can refer to the corresponding parts in the above methods.

[0031] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive Solid State Disk (SSD)), etc.

[0032] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0033] Each embodiment in this specification is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0034] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for diagnosing immune mechanism faults of intelligent robots, characterized in that: By collecting fault data samples of intelligent robots to generate memory cell sets with fault recognition capabilities, a robot fault diagnosis model based on immune mechanism is constructed; The robot fault diagnosis model is used to provide fault warning and diagnosis for robot equipment.

2. The immune mechanism fault diagnosis method for intelligent robots according to claim 1, characterized in that: The training process of the robot fault diagnosis model based on immune mechanism includes: ① Antigen set classification calculation of individual recognition threshold The antigen set Ag = {Ag1, Ag2, …, Agn} containing n individuals is divided into m sub-antigen sets TAgm, each of which contains antigens of the same type, that is, TAg = {TAg1, TAg2, …, TAgm} (m = 1,2,3,...), and the individual recognition threshold RTim of each type of antigen is calculated by formula (1); (1) In the formula, the ƒd function represents the affinity between the m-th antigen and other antigens. (2) Among them, δm is the cytokine concentration of the antibody with an affinity of 1 for the mth antigen; ② Vaccine antigen clone mutation proliferation to generate antibody set From the mth antigen set TAgm, the antigen that can promote the highest cytokine concentration is selected as the vaccine antigen Agv. (3) In the formula, the optimization rules for the two parameters σ and ɛ are as follows: (4) Differentiate the clones and mutations of the vaccine antigen Agv and generate targeted antibody sets Ab. (5) In the formula, κ is the isotype matrix of the vaccine antigen Agv, represents the mutation rate, and takes a random number in the range of (-1, 1), and the α parameter represents the length of cell mutation; ③. Screening of high affinity antibodies When the affinity function ƒt(Ab, Ag) is greater than the antibody recognition threshold, ƒt(Ab, Ag) = 1, otherwise ƒt(Ab, Ag) = 0, that is, ƒt(Ab, Ag) = 1 means that the antigen Ag is covered by the antibody cell Ab; the antibody cell (Ab, δ) that can cover the largest number of antigens is selected as the memory cell M through the affinity function, as shown in formula (6); (6) Randomly select one of the non-empty memory cell sets as the initial affinity-mature memory cell M. At this time, antigens that cannot be covered by memory cells will appear in TAgm, and these remaining antigens will form a new antigen set TAgmr; (7) ④. Generate memory cell sets The process of generating memory cell set TM is a cyclic process of repeated clonal mutation and proliferation of the remaining antigen set to generate affinity mature antibodies.

3. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is caused to perform the steps of the method according to claim 1 or 2.

4. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the computer program is executed by the processor, the processor is caused to perform the steps of the method according to claim 1 or 2.

Images