A monitoring method, device, medium and product for intelligent electromechanical systems

Abstract

The application discloses a monitoring method and device for an intelligent electromechanical system, a medium and a product, relates to the field of intelligent electromechanical system analysis, and comprises the following steps: acquiring task execution data of the intelligent electromechanical system under environmental disturbance and hardware health state; performing failure analysis based on a task function link of the intelligent electromechanical system, and constructing a software and hardware failure factor set; determining a disturbance variable representing the environmental disturbance, and combining the task execution data to establish a software reliability degradation model; constructing a software and hardware comprehensive Bayesian network based on the software and hardware failure factor set and the software reliability degradation model; and performing reliability reasoning and failure diagnosis on the intelligent electromechanical system by using the software and hardware comprehensive Bayesian network. The application can realize the prediction, reasoning and diagnosis of the complete task reliability of the intelligent electromechanical system.

Description

Technical Field

[0001] This application relates to the field of intelligent electromechanical system analysis, and in particular to a monitoring method, equipment, medium and product for intelligent electromechanical systems. Background Technology

[0002] With the development of intelligent manufacturing, intelligent robots, and unmanned systems, intelligent electromechanical systems are widely used in industrial production, warehousing and logistics, service robots, and operations in complex environments. These systems typically integrate mechanical structures, sensors, embedded controllers, artificial intelligence algorithms, and actuators, exhibiting a deep integration of mechanics, electronics, software, and intelligence.

[0003] Compared to traditional electromechanical systems, the reliability issues of intelligent electromechanical systems are more complex. On the one hand, hardware components such as servos, motors, grippers, cameras, main control boards, and communication buses still suffer from traditional failures such as wear, aging, overheating, frame loss, communication anomalies, and loss of control commands. On the other hand, intelligent algorithms are highly sensitive to the environment and are prone to new failure modes such as target recognition failure, center point offset, depth measurement errors, coordinate transformation anomalies, and motion planning failures under conditions such as target occlusion, surface contamination, reflection, changes in lighting, background interference, and dynamic scene changes.

[0004] Existing reliability analysis methods are typically based on component failure rates, life testing, fault tree analysis, or traditional FMEA methods, and are mainly suitable for analyzing hardware physical degradation problems. However, these methods are difficult to describe the nonlinear degradation laws of intelligent algorithms under complex environmental disturbances, and also difficult to reflect the process of fault propagation along the "perception-decision-planning-execution" functional link and the eventual task failure.

[0005] Meanwhile, existing software testing methods mostly focus on program logic, code defects, or input / output consistency, lacking effective quantification methods for performance degradation in intelligent perception caused by environmental disturbances. For example, in robotic arm vision grasping tasks, surface contamination of the target may lead to incomplete color threshold segmentation results, offset contour center calculations, abnormal depth mapping, and incorrect grasping paths. Such failures do not necessarily stem from program errors, but rather from the insufficient adaptability of intelligent perception algorithms to complex physical environments.

[0006] Therefore, a comprehensive analysis method is needed that can combine actual task execution data, software-aware degradation patterns, hardware health status, and system functional links to achieve effective monitoring of intelligent electromechanical systems. Summary of the Invention

[0007] The purpose of this application is to provide a monitoring method, device, medium, and product for intelligent electromechanical systems, which can enable the prediction, reasoning, and diagnosis of the complete task reliability of intelligent electromechanical systems.

[0008] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a monitoring method for intelligent electromechanical systems, including: Acquire task execution data of intelligent electromechanical systems under environmental disturbances and hardware health conditions; and conduct failure analysis based on the task function links of intelligent electromechanical systems to construct a set of software and hardware failure factors. Determine the disturbance variables characterizing environmental disturbances, and establish a software reliability degradation model based on the task execution data; the software reliability degradation model is used to characterize the nonlinear mapping relationship between the disturbance variables and the task success rate, and transforms the discrete test results into a continuous probability function; Based on the set of software and hardware failure factors and the software reliability degradation model, a software and hardware integrated Bayesian network is constructed; the software and hardware integrated Bayesian network is used to determine the causal relationships and conditional probability tables between nodes. We utilize integrated hardware and software Bayesian networks to perform reliability inference and failure diagnosis for intelligent electromechanical systems.

[0009] Secondly, this application provides a monitoring device for intelligent electromechanical systems, comprising: The data acquisition and failure analysis module is used to acquire task execution data of the intelligent electromechanical system under environmental disturbances and hardware health conditions; and to perform failure analysis based on the task function link of the intelligent electromechanical system to construct a set of software and hardware failure factors. The software reliability degradation model building module is used to determine the disturbance variables characterizing environmental disturbances and to build a software reliability degradation model in combination with the task execution data; the software reliability degradation model is used to characterize the nonlinear mapping relationship between the disturbance variables and the task success rate, and to transform the discrete test results into a continuous probability function; The integrated hardware and software Bayesian network construction module is used to construct an integrated hardware and software Bayesian network based on the set of hardware and software failure factors and the software reliability degradation model; the integrated hardware and software Bayesian network is used to determine the causal relationships and conditional probability tables between nodes. The reliability reasoning and failure diagnosis module is used to perform reliability reasoning and failure diagnosis of intelligent electromechanical systems using a combination of hardware and software Bayesian networks.

[0010] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the monitoring method for intelligent electromechanical systems.

[0011] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned monitoring method for intelligent electromechanical systems.

[0012] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the aforementioned monitoring method for intelligent electromechanical systems.

[0013] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a monitoring method, device, medium, and product for intelligent electromechanical systems. It identifies disturbance variables characterizing environmental disturbances and establishes a software reliability degradation model based on task execution data. The relationship between the degree of environmental disturbance and task success rate is transformed into a computable reliability function, providing a quantitative basis for the reliability analysis of intelligent algorithms and enabling the quantification of the impact of environmental disturbances on the reliability of intelligent sensing algorithms. A comprehensive software-hardware Bayesian network is constructed based on the set of software and hardware failure factors and the software reliability degradation model. This comprehensive software-hardware Bayesian network unifies the expression of software degradation parameters, hardware health status, and task execution results, enabling… This application enables more accurate reasoning about system-level task reliability, achieving a comprehensive analysis of the software and hardware reliability of intelligent electromechanical systems. Utilizing a hardware-software integrated Bayesian network, it forward-reasons the probability of reliable complete task functionality based on the current environmental disturbance level and hardware health status, providing a basis for risk assessment, task condition judgment, and safety control before the intelligent electromechanical system's operation. When a task failure occurs, it can further infer the posterior probability of each failure cause, identifying the main factors leading to task failure, thereby supporting visual algorithm optimization, hardware maintenance, sensor calibration, and control strategy improvement. Ultimately, it enables dynamic prediction and diagnosis of the reliability of intelligent electromechanical systems. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a schematic flowchart of a monitoring method for intelligent electromechanical systems according to an embodiment of this application. Detailed Implementation

[0016] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0017] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] In one exemplary embodiment, such as Figure 1 As shown, a monitoring method for intelligent electromechanical systems is provided. This method is executed by a computer device and includes the following steps S101 to S104. Wherein: S101: Obtain task execution data of the intelligent electromechanical system under environmental disturbances and hardware health conditions; and perform failure analysis based on the task function link of the intelligent electromechanical system to construct a set of software and hardware failure factors. Among them, the intelligent electromechanical system has environmental perception, intelligent decision-making, motion planning and execution control functions; Regarding the acquisition of task execution data, this application compiles data based on the test records and system operation links of the ArmPi Ultra robotic arm physical platform. The task execution data includes: the white contamination area level of the target surface, the status of software functions (visual perception, depth measurement, coordinate calculation, motion planning), the health status of the servo motors, the hardware execution status (gripper status and grasping execution status), as well as the number of tests, the number of successful grasps, the number of failed grasps, the grasping success rate, and the main failure phenomena under different contamination levels.

[0019] Among them, the white pollution area level of the target surface is used to characterize environmental disturbance; the visual perception status mainly reflects whether the target can be effectively detected, whether the binary mask is complete, and whether there is a significant offset in the contour center; the depth measurement status mainly reflects whether the depth value is valid or whether there is an abnormal jump; the coordinate calculation status and motion planning status are used to characterize the software function status in the process from perception input to grasping action generation; the servo health status and gripper status are used to characterize the hardware execution chain status; and the task result data is obtained from the grasping test statistics.

[0020] S101 specifically includes: S11, Obtain the task function link of the intelligent electromechanical system; the task function link includes, in sequence, the sensing input, data processing, coordinate calculation, motion planning, execution control and task result output links; S12, acquire test data of the task function link under different environmental disturbances and hardware health states; the test data includes environmental disturbance level, software function state, hardware execution state and statistical results of task success or failure; S13, based on the task function link, perform software system failure mode and effect analysis, hardware system failure mode and effect analysis, and software and hardware coupling failure mode and effect analysis (FMEA) to identify key failure modes; the key failure modes include visual perception failure, depth measurement anomaly, coordinate calculation deviation, motion planning failure, insufficient servo output, gripper slippage, and communication synchronization anomaly. S14, the failed samples are classified into primary causes according to the functional link where the anomaly first occurs, forming the set of software and hardware failure factors.

[0021] Taking the ArmPi Ultra robotic arm's 3D spatial grasping task as an example, the functional chain of the robotic arm's 3D spatial grasping task, namely "visual perception - depth measurement - coordinate calculation - motion planning - grasping execution", is decomposed. Secondly, software system FMEA, hardware system FMEA, and software-hardware coupling FMEA are carried out to identify key failure modes such as visual perception failure, depth measurement anomaly, coordinate calculation deviation, motion planning failure, insufficient servo output, gripper slippage, and communication synchronization anomaly. Finally, the failure samples are classified into primary causes according to the functional link where the anomaly first occurs, forming a set of failure factors that can be used for Bayesian network node selection and probability assignment.

[0022] Furthermore, the scattered software anomalies, hardware degradation, and task failures in intelligent electromechanical systems are uniformly organized into modelable reliability influencing factors.

[0023] S102, determine the disturbance variable characterizing the environmental disturbance, and establish a software reliability degradation model in combination with the task execution data; the software reliability degradation model is used to characterize the nonlinear mapping relationship between the disturbance variable and the task success rate, and transform the discrete test results into a continuous probability function; Specifically, the perturbation variables include, but are not limited to: the percentage of contaminated area on the target surface, the target occlusion ratio, the ambient light intensity, the background complexity, the reflectivity of the target, the color similarity of the target, or the target's movement speed.

[0024] Since the visual perception stage is located at the forefront of the robotic arm's grasping task functional chain, its output directly affects subsequent coordinate calculation, motion planning, and grasping execution. Simultaneously, white contamination on the target surface can disrupt the color features upon which LAB color space threshold segmentation relies, easily leading to incomplete binary masks, contour center shifts, and target detection failures. Therefore, the proportion of white contamination area on the target surface is selected as a typical environmental disturbance variable, and the grasping success rate is used as the task reliability output indicator to establish a nonlinear mapping relationship between environmental disturbance variables and task success rate. The specific data acquisition process is as follows: Under the same experimental environment, with identical robotic arm software structure, control logic, and hardware composition, only the proportion of white contamination area on the surface of the target color block was varied. The contamination area levels were set to seven levels: 10%, 20%, 35%, 50%, 65%, 80%, and 90%, with 10%–35% representing light contamination, 50%–65% moderate contamination, and 80%–90% heavy contamination. Twenty independent grasping tests were conducted at each contamination level, resulting in a total of 140 physical test samples. For each test, the contamination area level, number of tests, number of successful grasps, number of failed grasps, grasping success rate, and main failure phenomena were recorded.

[0025] According to the test results, as the proportion of white pollution area increases, the system's grasping success rate shows a significant non-linear downward trend. Under light pollution conditions, visual perception and coordinate calculation still have a certain degree of robustness, and failures are mostly manifested as execution-level problems such as gripping slippage; under moderate pollution conditions, incomplete binary masks and target center offset lead to distorted coordinate calculations, and the task success rate drops rapidly; under heavy pollution conditions, visual perception and depth mapping fail, and the system struggles to trigger effective grasping actions.

[0026] The software reliability degradation model is constructed using logistic function, exponential degradation model, Weibull distribution model, Probit regression model, piecewise linear model, Gaussian process regression model, support vector machine, random forest or neural network model.

[0027] The following explanation uses the logistic function to illustrate the software reliability degradation model. The logistic function is a non-linear probability model commonly used to describe binary classification probabilities, state transition probabilities, or S-shaped degradation trends. Using this function as a software reliability degradation model for the relationship between "white pollution area percentage - capture success rate," its form is: ; Where x represents the proportion of white pollution area, S(x) represents the expected capture success rate when the pollution area proportion is x, x0 is the inflection point of the curve, that is, the pollution area proportion when the capture success rate is about 50%, and k is a parameter to be estimated, used to reflect the rate of change of the success rate as the pollution area increases.

[0028] Based on the actual capture success rate data under the above seven pollution levels, the parameters were estimated by nonlinear least squares method, and the fitting results were k=9.1378 and x0=0.6255, indicating that the logistic empirical model has good consistency with the actual degradation process; when the white pollution area is close to about 62.55% of the total area of ​​the target color block, the system success rate enters the rapid decline range.

[0029] The discrete grasping test results are transformed into continuous probability functions, allowing the expected grasping success rate under any contaminated area percentage to be estimated by the software reliability degradation model. This model describes the system performance degradation trend under visual disturbance conditions and provides data for assigning conditional probabilities to nodes such as visual perception state, depth measurement state, coordinate calculation state, and grasping execution state in the subsequent integrated hardware and software Bayesian network.

[0030] S103, Based on the set of software and hardware failure factors and the software reliability degradation model, construct a software and hardware integrated Bayesian network; the software and hardware integrated Bayesian network is used to determine the causal relationship and conditional probability table between nodes; Based on a software and hardware integrated Bayesian network, which integrates software functional status, hardware health status and experimental data under different pollution scenarios, the system can achieve complete reliability prediction of the crawling function and diagnose the cause of failure after the task fails.

[0031] S103 specifically includes: S31, map each factor in the set of software and hardware failure factors to a binary state node; the binary state node includes environmental disturbance related nodes, software perception nodes, software decision planning nodes, hardware execution nodes, and complete task function nodes. S32, determine the directed causal edges between nodes according to the workflow of the task function link to form a failure propagation topology; S33, the probability function output by the software reliability degradation model is used as the prior probability or conditional probability, and the experimental data is combined to assign values ​​to each node to generate a conditional probability table.

[0032] Taking the ArmPi Ultra robotic arm's 3D spatial grasping task as an example, the level of white contamination area on the target surface is not treated as a network node, but rather as an external scenario condition, categorized into three levels: light contamination, moderate contamination, and heavy contamination. Node probabilities are set and inference analysis is performed under the same network structure. The integrated hardware and software Bayesian network consists of eight binary nodes: visual perception state, depth measurement state, coordinate calculation state, motion planning state, servo health state, gripper state, grasping execution state, and complete grasping function reliability. Each node has two states, representing the corresponding function or component in a binary state such as normal / abnormal, accurate / deviationate, or reliable / unreliable.

[0033] The causal relationships between nodes are determined based on the workflow of the robotic arm's 3D spatial grasping task: the visual perception state and depth measurement state jointly affect the coordinate calculation state; the coordinate calculation state further affects the motion planning state; the coordinate calculation state, motion planning state, servo health state, and gripper state jointly affect the grasping execution state; and the grasping execution state ultimately determines the reliability of the complete grasping function. This structure embodies the failure propagation relationship of "perceptual input—coordinate calculation—motion planning—execution state—task reliability".

[0034] S104 utilizes a hardware-software integrated Bayesian network to perform reliability reasoning and failure diagnosis for intelligent electromechanical systems.

[0035] S104 specifically includes: During the forward reasoning phase, the current environmental disturbance state and hardware health state are input as evidence into the Bayesian network to calculate the posterior probability of the complete task function being reliable. For example, after inputting the proportion of white pollution area on the target surface, the health state of the servo motor, the gripper state, and the communication state, the integrated hardware and software Bayesian network outputs the reliability probability of the robotic arm completing the complete grasping task.

[0036] In the reverse diagnostic phase, when a task failure occurs in the intelligent electromechanical system, the evidence of "unreliable complete task function" is input into a hardware-software integrated Bayesian network. This network then infers the posterior probability of each failure factor node and ranks the causes of failure according to their probability. For example, the intelligent electromechanical system can determine that task failure is more likely caused by visual perception failure, depth mapping anomalies, motion planning failure, servo degradation, insufficient gripper holding, or communication anomalies, thereby pinpointing the system's weak points.

[0037] Following S104 are: S41 collects new task execution data, hardware status monitoring data, and fault records during the continuous operation of the intelligent electromechanical system. S42, the integrated hardware and software Bayesian network is updated and corrected based on the new task execution data, hardware status monitoring data and fault records.

[0038] That is, the prior probability and conditional probability tables of the integrated software and hardware Bayesian network are updated based on new task execution data, hardware status monitoring data and fault records, so that the system reliability inference results are continuously corrected with the running data, thereby improving the applicability of the integrated software and hardware Bayesian network under different objectives, different environments and different hardware states.

[0039] This application is applicable to industrial robotic arm assembly tasks, mobile robot handling tasks, unmanned vehicle visual recognition and path tracking tasks, UAV visual positioning tasks, intelligent inspection equipment defect identification tasks, and other intelligent electromechanical systems that rely on perception, decision-making, planning, and execution. It is used for reliability assessment, design improvement, and operation and maintenance of intelligent robotic arms, collaborative robots, and other intelligent electromechanical equipment. Taking the ArmPi Ultra robotic arm's three-dimensional spatial grasping task as an example, this application combines the robotic arm's grasping functional chain, white pollution disturbance test data, FMEA failure analysis results, and a Bayesian network inference model to infer the reliability probability of the system completing the full grasping task. It also identifies the key links most likely to cause failure when the task fails, demonstrating clear engineering application scenarios and physical significance. Specifically, it has the following effects: First, this application enables a comprehensive analysis of the software and hardware reliability of intelligent electromechanical systems. Traditional reliability analysis methods often model mechanical structures, electronic components, or software algorithms separately, making it difficult to describe the coupling relationships between environmental disturbances, software perception degradation, hardware state decline, and task failures in intelligent electromechanical systems. This application uses Bayesian networks to uniformly express software degradation parameters, hardware health status, and task execution results, enabling more accurate reasoning about system-level task reliability.

[0040] Second, this application quantifies the impact of environmental disturbances on the reliability of intelligent perception algorithms. For intelligent tasks such as visual grasping, algorithm failures are often not caused by traditional program errors, but by external factors such as target contamination, occlusion, reflection, and changes in lighting. This application transforms the relationship between the degree of environmental disturbance and task success rate into a computable reliability function through disturbance testing and logistic degradation modeling, providing a quantitative basis for the reliability analysis of intelligent algorithms.

[0041] Third, this application enables dynamic prediction, diagnosis, and updating of the reliability of intelligent electromechanical systems. By inputting the current environmental disturbance level and hardware health status, the Bayesian network can forward infer the probability of reliable complete task functionality, providing a basis for risk assessment, task condition judgment, and safety control before system operation. When a task failure occurs, it can further infer the posterior probability of each failure cause, identifying the main factors leading to task failure, thereby providing support for visual algorithm optimization, hardware maintenance, sensor calibration, and control strategy improvement. Simultaneously, as subsequent operational data and fault records accumulate, the prior probability and conditional probability tables in the Bayesian network can be dynamically corrected, enabling the reliability model to adapt to different application scenarios, different task objects, and different hardware states, improving the continuous applicability and engineering practicality of system-level reliability analysis.

[0042] Based on the same inventive concept, this application also provides a monitoring device for intelligent electromechanical systems for implementing the monitoring method for intelligent electromechanical systems described above. The solution provided by this device is similar to the implementation described in the above method. Therefore, the specific limitations in one or more embodiments of the monitoring device for intelligent electromechanical systems provided below can be found in the limitations of the monitoring method for intelligent electromechanical systems described above, and will not be repeated here.

[0043] In one exemplary embodiment, a monitoring device for intelligent electromechanical systems is provided, comprising: The data acquisition and failure analysis module is used to acquire task execution data of the intelligent electromechanical system under environmental disturbances and hardware health conditions; and to perform failure analysis based on the task function link of the intelligent electromechanical system to construct a set of software and hardware failure factors. The software reliability degradation model building module is used to determine the disturbance variables characterizing environmental disturbances and to build a software reliability degradation model in combination with the task execution data; the software reliability degradation model is used to characterize the nonlinear mapping relationship between the disturbance variables and the task success rate, and to transform the discrete test results into a continuous probability function; The integrated hardware and software Bayesian network construction module is used to construct an integrated hardware and software Bayesian network based on the set of hardware and software failure factors and the software reliability degradation model; the integrated hardware and software Bayesian network is used to determine the causal relationships and conditional probability tables between nodes. The reliability reasoning and failure diagnosis module is used to perform reliability reasoning and failure diagnosis of intelligent electromechanical systems using a combination of hardware and software Bayesian networks.

[0044] In an exemplary embodiment, a computer device is provided, which may be a server or a terminal. The computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is connected to the system bus via the I / O interfaces. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The I / O interfaces of the computer device are used for exchanging information between the processor and external devices. The communication interface of the computer device is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a monitoring method for intelligent electromechanical systems.

[0045] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0046] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0047] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0048] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0049] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0050] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0051] In this application, all actions to acquire signals, information, or data are carried out in compliance with the relevant data protection laws and policies of the country where the location is situated, and with the authorization granted by the owner of the relevant device.

[0052] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0053] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A monitoring method for intelligent electromechanical systems, characterized in that, include: Acquire task execution data of intelligent electromechanical systems under environmental disturbances and hardware health conditions; Failure analysis is performed based on the task function chain of the intelligent electromechanical system, and a set of software and hardware failure factors is constructed. Determine the disturbance variables that characterize environmental disturbances, and establish a software reliability degradation model based on the task execution data; the software reliability degradation model is used to characterize the nonlinear mapping relationship between the disturbance variables and the task success rate, and transforms the discrete test results into a continuous probability function; Based on the set of software and hardware failure factors and the software reliability degradation model, a software and hardware integrated Bayesian network is constructed; the software and hardware integrated Bayesian network is used to determine the causal relationships and conditional probability tables between nodes. We utilize integrated hardware and software Bayesian networks to perform reliability inference and failure diagnosis for intelligent electromechanical systems.

2. The monitoring method for intelligent electromechanical systems according to claim 1, characterized in that, The failure analysis is based on the task function chain of the intelligent electromechanical system, and a set of software and hardware failure factors is constructed, specifically including: The task function chain of the intelligent electromechanical system is obtained; the task function chain includes, in sequence, the sensing input, data processing, coordinate calculation, motion planning, execution control and task result output. Acquire test data of the task function link under different environmental disturbances and hardware health conditions; the test data includes environmental disturbance level, software function status, hardware execution status and statistical results of task success or failure; Based on the aforementioned task function links, software system failure mode and effect analysis, hardware system failure mode and effect analysis, and software-hardware coupling failure mode and effect analysis are performed to identify key failure modes. The key failure modes include visual perception failure, depth measurement anomaly, coordinate calculation deviation, motion planning failure, insufficient servo output, gripper slippage, and communication synchronization anomaly. The failed samples are categorized by the functional component where the anomaly first occurred, forming the set of software and hardware failure factors.

3. The monitoring method for intelligent electromechanical systems according to claim 1, characterized in that, The perturbation variables include the percentage of contaminated area on the target surface, the target occlusion ratio, the ambient light intensity, the background complexity, the target reflectivity, the target color similarity, or the target movement speed.

4. The monitoring method for intelligent electromechanical systems according to claim 1, characterized in that, The software reliability degradation model is constructed using logistic function, exponential degradation model, Weibull distribution model, Probit regression model, piecewise linear model, Gaussian process regression model, support vector machine, random forest or neural network model.

5. The monitoring method for intelligent electromechanical systems according to claim 1, characterized in that, The construction of a comprehensive software and hardware Bayesian network based on the set of software and hardware failure factors and the software reliability degradation model specifically includes: Each factor in the set of software and hardware failure factors is mapped to a binary state node; the binary state node includes environmental disturbance related nodes, software perception nodes, software decision planning nodes, hardware execution nodes, and complete task function nodes. Based on the workflow of the task function link, directed causal edges between nodes are determined to form a failure propagation topology. The probability function output by the software reliability degradation model is used as the prior probability or conditional probability, and combined with the experimental data, values ​​are assigned to each node to generate a conditional probability table.

6. The monitoring method for intelligent electromechanical systems according to claim 1, characterized in that, The method of using integrated hardware and software Bayesian networks for reliability reasoning and failure diagnosis of intelligent electromechanical systems also includes: During the continuous operation of the intelligent electromechanical system, new task execution data, hardware status monitoring data, and fault records are collected. The integrated hardware and software Bayesian network is updated and corrected based on new task execution data, hardware status monitoring data, and fault records.

7. A monitoring device for intelligent electromechanical systems, characterized in that, include: The data acquisition and failure analysis module is used to acquire task execution data of the intelligent electromechanical system under environmental disturbances and hardware health conditions. Failure analysis is performed based on the task function chain of the intelligent electromechanical system, and a set of software and hardware failure factors is constructed. The software reliability degradation model building module is used to determine the disturbance variables characterizing environmental disturbances and to build a software reliability degradation model in combination with the task execution data; the software reliability degradation model is used to characterize the nonlinear mapping relationship between the disturbance variables and the task success rate, and to transform the discrete test results into a continuous probability function; The integrated hardware and software Bayesian network construction module is used to construct an integrated hardware and software Bayesian network based on the set of hardware and software failure factors and the software reliability degradation model; the integrated hardware and software Bayesian network is used to determine the causal relationships and conditional probability tables between nodes. The reliability reasoning and failure diagnosis module is used to perform reliability reasoning and failure diagnosis of intelligent electromechanical systems using a combination of hardware and software Bayesian networks.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the monitoring method for intelligent electromechanical systems according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the monitoring method for intelligent electromechanical systems as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the monitoring method for intelligent electromechanical systems as described in any one of claims 1-6.

Images