A distributed fault detection method in swarm robotics systems based on machine learning

Abstract

The present invention relates to a distributed fault detection method in a swarm robotic system based on machine learning, belonging to the field of robotic control and fault detection technology. This method aims to address the single point of failure problem inherent in centralized fault detection by implementing fault detection and decision-making through a distributed architecture. Specifically, it includes steps such as information collection, feature extraction, machine learning classification, communication sharing, and voting. It utilizes a machine learning classification algorithm to improve fault detection accuracy, and uses a weighted majority voting mechanism to reduce misjudgments and enhance system robustness and fault tolerance. This method can be widely applied to fault detection and collaborative work in swarm robotic systems in complex environments.

Description

Technical Field

[0001] The present invention relates to a fault detection technology in a multi-agent system, and in particular to a distributed fault detection method in a swarm robotic system (SRS). Background Art

[0002] Swarm robotic systems consist of multiple intelligently interconnected nodes that collaborate and exhibit complex collective behaviors. While individual robots have limited capabilities for certain tasks, swarm robotic systems leverage their distributed nature to accomplish complex tasks without central control through communication and interaction between nodes. This distributed architecture is generally considered inherently robust. However, research has shown that even failures in just a few robots can significantly impact the performance and stability of the entire system. Existing fault detection methods are generally categorized into exogenous and endogenous methods. The former is used to detect faults in other robots, while the latter is used to detect faults within the system itself. These methods can be divided into qualitative and quantitative methods based on their classification algorithms. Qualitative methods are simple and resource-efficient, but their detection accuracy is limited. Quantitative methods rely on modeling robot behavior and can offer higher detection accuracy. Quantitative methods are further divided into model-based and data-driven algorithms. The former relies on domain knowledge, while the latter autonomously learns through data analysis to distinguish between normal and faulty behavior. With the continuous advancement of data storage technology and machine learning algorithms, data-driven methods are showing increasing potential for fault detection in swarm robotic systems. Summary of the Invention

[0003] The technical problem to be solved by the present invention is how to implement distributed fault detection in a swarm robot system, thereby improving the robustness and fault tolerance of the system and avoiding the single point failure problem of centralized fault detection.

[0004] In order to solve the above technical problems, the present invention adopts the following technical solution: The present invention designs a distributed fault detection method based on a swarm robot system, comprising the following steps:

[0005] In step 001, each robot uses its sensors and communication module to observe its neighboring robots and collect behavioral data. This data includes information such as distance, relative angle, speed, and acceleration to monitor and analyze the neighboring robots.

[0006] In step 002, the observed raw data is processed to generate features that represent the neighboring robot's behavior. These features can be binary (e.g., whether the robot is within a specified distance range) or numerical (e.g., speed, acceleration, etc.). These features provide the basis for subsequent classification, helping to identify whether the robot is exhibiting abnormal behavior.

[0007] In step 003, after feature extraction is complete, the robot conducts a preliminary assessment of its neighbors' behavior to identify possible anomalies. For example, if a robot is observed to remain stationary or move at an unreasonable speed, the system will mark it as "suspicious," providing a reference for the subsequent classification stage.

[0008] In step 004, the robot accumulates multiple observed feature data within a time window and passes these features as input to the machine learning classifier. By combining multiple observed features, the changing trends of the robot's behavior can be more comprehensively reflected.

[0009] In step 005, each robot uses a machine learning classifier (such as logistic regression or gradient boosting decision tree) to determine the fault status of its neighboring robots. The classifier classifies the robots based on the input feature data and outputs a fault determination. Logistic regression is suitable for real-time detection with low resource consumption, while gradient boosting decision trees can handle more complex feature relationships and provide higher classification accuracy.

[0010] In step 006, the classifier generates a fault label after analyzing the feature data. If the fault probability output by the classifier exceeds the set threshold, the robot is marked as a faulty robot; if it is below the threshold, it is marked as a normal working robot.

[0011] In step 007, after classification is complete, the robot transmits its classification results for its neighboring robots to other robots via the local communication network. This information sharing allows each robot to obtain the classification results of multiple robots for the same object, providing basic data for the final voting decision.

[0012] In step 008, after receiving the classification results, each robot makes a collective decision through a majority voting mechanism. If a majority of robots determine that a robot is faulty, it is collectively identified as faulty; if a majority of robots determine that it is functioning properly, it is considered to be functioning normally. This distributed voting mechanism effectively reduces the impact of a single robot's misjudgment on the overall system.

[0013] In step 009, after voting to determine whether a robot is faulty, the swarm robot system makes a decision based on the voting results. For faulty robots, the system can isolate or stop interacting with them, or reassign tasks to maintain the overall system's normal operation. Furthermore, the system can trigger self-recovery actions or send a fault alarm signal based on the fault condition.

[0014] As a preferred technical solution of the present invention: each robot collects behavioral data such as distance, angle, speed, etc. with its neighboring robots through high-precision sensors (such as ultrasonic, infrared sensors or lidar).

[0015] As a preferred technical solution of the present invention, in the machine learning classification phase of step 005, the system preferably uses a gradient boosted decision tree (GBDT) as the primary classification algorithm. GBDT can efficiently handle complex nonlinear features and achieve high-precision classification results by integrating multiple weak classifiers.

[0016] As a preferred technical solution of the present invention: in the voting mechanism of step 008, the weight of each robot's vote can be dynamically adjusted according to the accuracy of its sensor, data quality or the importance of its location.

[0017] As a preferred technical solution of the present invention: the system introduces a redundant communication mechanism, that is, each robot not only exchanges fault classification information with its direct neighbors, but also transmits the classification results to robots at a longer distance through multi-hop communication.

[0018] In order to solve the above technical problems, the present invention adopts the following technical solution: The present invention designs a distributed group robot fault detection system, which includes the following units:

[0019] Sensor unit: used to collect behavioral data such as distance, angle, speed, and acceleration of neighboring robots to provide basic information for fault detection;

[0020] Data processing and feature extraction unit: responsible for processing sensor data and extracting binary or numerical features that can characterize robot behavior;

[0021] Machine learning classification unit: uses the extracted feature data to determine whether the robot has a fault through a machine learning classifier;

[0022] Communication and sharing unit: responsible for the transmission and sharing of classification results between robots, ensuring that the information covers every robot in the entire system;

[0023] Voting decision unit: Through a weighted majority voting mechanism, collectively decide whether a robot has a fault;

[0024] Fault handling unit: After confirming a fault, it takes measures such as isolation, task reallocation or fault recovery to ensure the normal operation of the system.

[0025] The system of the present invention adopts a distributed fault detection and weighted voting mechanism, which has higher robustness, real-time performance and detection accuracy than the existing technology; by introducing a machine learning algorithm to improve the accuracy of fault detection, and flexibly handle the fault according to its severity, the stable operation of the system is guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of the distributed swarm robot system fault detection method designed by the present invention. DETAILED DESCRIPTION

[0027] like Figure 1 As shown, the present invention designs a distributed swarm robot system detection method to perform distributed fault detection and decision-making on the system. In actual application, the method specifically includes the following steps:

[0028] In step 001, each robot collects the behavior data of the neighboring robots through sensors and determines whether the data is successfully obtained: if yes, it proceeds to step 002; otherwise, it re-collects the data.

[0029] Step 002: The robot processes the raw data and generates a feature vector.

[0030] In step 003, the robot performs a preliminary evaluation of the extracted features to identify whether there are any anomalies and determine whether abnormal behavior is found: if so, it is marked as "suspicious state" and proceeds to step 004; otherwise, it directly proceeds to step 4.

[0031] Step 004, input the feature data within a time window into the machine learning classifier to determine whether sufficient feature data has been collected: if so, proceed to step 005, otherwise continue to accumulate data.

[0032] Step 005: Analyze the features through a machine learning classifier to determine whether the robot has a fault.

[0033] Step 006, generate a fault label based on the analysis results of the classifier, and determine whether the fault probability exceeds the set threshold: if yes, mark it as a "faulty robot", otherwise mark it as a "normal working robot" and go to step 007.

[0034] Step 007: transmit the classification result to the neighboring robots through the local communication network to determine whether the result is shared successfully: if yes, proceed to step 008; otherwise, retry sharing.

[0035] Step 008, collectively decide whether a robot has a fault through a weighted majority voting mechanism, and determine whether the majority voting results confirm a fault: if yes, proceed to step 009, otherwise end the detection process of the robot.

[0036] Step 009, take corresponding measures based on the voting results to determine the severity of the fault: if yes, isolate the faulty robot or reassign tasks; otherwise, continue monitoring or end the detection.

[0037] In response to the above-mentioned distributed swarm robot fault detection method, the present invention further designs a system based on this distributed fault detection method. The system includes a sensor unit, a data processing and feature extraction unit, a machine learning classification unit, a communication and sharing unit, a voting and decision-making unit, and a fault handling unit, with specific designs for each of these units. The sensor unit is used to collect behavioral data between the system and neighboring robots. The sensor unit can monitor the surrounding environment in real time and generate raw data, providing basic information for fault detection; data processing and feature extraction unit: responsible for processing the raw data collected by the sensor unit, extracting features that can characterize the behavior of neighboring robots, and providing an analysis basis for subsequent classifiers; machine learning classification unit: responsible for inputting the extracted features into the machine learning classifier, using algorithms such as gradient boosting decision tree (GBDT) to classify neighboring robots into faulty and non-faulty states, and judging whether the robot is faulty based on a preset threshold; communication and sharing unit: used for sharing classification results between robots, using local communication or multi-hop communication to ensure that each robot can exchange information with neighboring robots and pass the classification results to more robots; voting decision unit: by collecting the classification results of other robots and using a weighted majority voting mechanism to determine whether a robot is faulty; fault handling unit: the fault handling unit takes corresponding measures according to the severity of the fault, such as isolating the faulty robot, reallocating tasks, or triggering fault recovery operations to ensure the normal operation of the overall function of the group system.

[0038] The method and system designed using the above technical solution have the following technical effects compared to the existing technology: The method and system described in the present invention can achieve efficient distributed fault detection in a swarm robot system. Each robot identifies faults through real-time observation, data processing, and machine learning classifiers, and a voting mechanism is used to ensure the accuracy of the classification results. Through multi-level decision-making steps, the system can promptly isolate faulty robots or reassign tasks, ensuring the continuity of swarm tasks and the robustness of the entire system. This distributed approach improves the system's ability to respond to environmental changes and robot failures, greatly enhancing the stability and reliability of swarm robots in complex environments.

[0039] The embodiments of the present invention are described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Various changes can be made within the scope of knowledge possessed by ordinary technicians in this field without departing from the spirit of the present invention.

Claims

1. A distributed fault detection method in a swarm robot system based on machine learning, comprising an information acquisition module, a data processing and feature extraction module, a fault classification module, a communication and sharing module, a voting decision module, and a fault handling module, characterized in that: The information acquisition module is used to collect distance, relative angle, speed and acceleration data of adjacent robots; The data processing and feature extraction module performs feature extraction on the collected data; The fault classification module uses a classifier to classify feature data and determine whether there is an abnormality in the neighboring node; The communication and sharing module transmits the classification results to the neighboring robots and transmits them through multi-hop mode; The voting decision module adopts a weighted majority voting mechanism based on the received classification results, dynamically adjusts the voting weights according to the sensor accuracy or data quality of different robots, and decides whether there is a faulty node; After determining that there is a faulty node, the fault processing module performs node isolation, task reallocation or self-recovery operations to ensure the stable operation of the swarm robot system.

2. The distributed fault detection method in a swarm robot system based on machine learning according to claim 1, characterized in that: The classifier is a gradient boosting decision tree classifier.

3. The distributed fault detection method in a swarm robot system based on machine learning according to claim 1, characterized in that: The information collection module collects behavior data of neighboring robots through multiple sensors.

4. The distributed fault detection method in a swarm robot system based on machine learning according to claim 1, characterized in that: The communication and sharing module transmits the classification results to multiple neighboring robots using a redundant communication method.

Images