A method and system for monitoring the status and fault warning of a sanitation robot cluster

Abstract

The present invention discloses a method and system for monitoring the status and fault warning of a sanitation robot cluster. By optimizing the execution order of maintenance tasks for each sanitation robot device, the devices with a high fault risk level and a high maintenance priority in each sanitation robot device are preferentially included in the maintenance tasks and executed in a timely manner; according to the execution effect of the device maintenance tasks, the relevant parameters of the fault prediction model are dynamically updated, the device fault prediction results are recalculated, and it is judged whether the operating status of the device has returned to normal; if it is recognized that the operating status of the device has returned to normal and the operating status data are within the normal operating status threshold range within the set time, the device maintenance task is ended, otherwise the maintenance plan needs to be adjusted and the maintenance measures are continuously optimized until the sanitation robot device completely returns to normal. The present invention realizes the high efficiency and accuracy of intelligent monitoring and fault warning of the status of the sanitation robot cluster, improves the device reliability and operating efficiency, and reduces the maintenance cost.

Description

Technical Field

[0001] The present invention relates to the technical field of robot operation control, and in particular discloses a method and system for monitoring the status and fault warning of a sanitation robot cluster. Background Art

[0002] During the long-term operation of the sanitation robot cluster, due to environmental factors and equipment aging, the status monitoring and fault warning methods have the following problems:

[0003] 1. Sensor accuracy degradation: During long-term operation, the accuracy and reliability of sensors decrease due to environmental influences or equipment aging, resulting in problems such as observation data drift and increased noise, which directly affect the accuracy of the fault prediction model.

[0004] 2. Difficulty in joint monitoring of key components: Batteries, fans, nozzles, sweeping discs, lidar and other key components have different working conditions, aging rates and failure modes, which makes it difficult to uniformly evaluate the status during joint monitoring. In addition, there are potential related failures between different components. How to identify the fault chain through multi-dimensional data fusion is a key technical difficulty.

[0005] 3. Calibration problems due to individual differences: The performance of sensors of different models and batches varies significantly, making it difficult to calibrate and optimize them uniformly. In addition, sensor performance changes dynamically with operating time, requiring targeted adjustment of fault warning parameters, but current methods are difficult to adapt to.

[0006] 4. Data consistency and continuity: As the cluster runs longer, sensor degradation may lead to incomplete and inconsistent observation data, data breakpoints or error accumulation. This problem easily forms a blind spot for fault prediction, seriously affecting the accuracy and real-time performance of monitoring.

[0007] 5. Optimization and computing efficiency bottleneck: Existing methods are difficult to simultaneously take into account the real-time data collection efficiency and fault risk priority processing requirements of large-scale equipment clusters. Especially when the number of equipment increases and the operating environment becomes more complex, the priority sorting and resource allocation of maintenance tasks become more difficult.

[0008] Therefore, the above-mentioned defects in the existing sanitation robot cluster operations, especially in the joint monitoring of multiple components, data consistency maintenance and dynamic optimization, are technical problems that need to be solved urgently. Summary of the invention

[0009] The present invention provides a method and system for monitoring the status and early warning of a sanitation robot cluster, aiming to solve at least one of the above-mentioned defects existing in the existing sanitation robot cluster operation.

[0010] One aspect of the present invention relates to a method for monitoring the status and fault warning of a sanitation robot cluster, including the following steps:

[0011] Obtain the operation status data of the sanitation robot cluster; the operation status data includes temperature, voltage and current data;

[0012] Compare the obtained operation status data with the preset normal operation status threshold. If the operation status data exceeds the normal operation status threshold, dynamically adjust the observation frequency of the operation status data and narrow the observation range;

[0013] Obtain the cumulative operation duration of the sanitation robot cluster. If the cumulative operation duration exceeds the preset safety duration threshold, adjust the duration weight parameter in the fault prediction model;

[0014] According to the adjusted duration weight parameter, recalculate the fault prediction result, comprehensively consider the cumulative operation duration and the operation status data, and judge whether there is a risk of fatigue loss for each sanitation robot device. If there is a risk of fatigue loss, generate a device fault warning signal;

[0015] Obtain the device fault warning signal, generate a targeted device maintenance suggestion plan in combination with the operation status data and the cumulative operation duration, and determine the maintenance priority of each sanitation robot device according to the fault risk level;

[0016] Optimize the execution order of the maintenance tasks of each sanitation robot device, and give priority to the devices with high fault risk level and high maintenance priority in each sanitation robot device to be included in the maintenance tasks and execute them in a timely manner;

[0017] According to the execution effect of the device maintenance tasks, dynamically update the relevant parameters of the fault prediction model, recalculate the device fault prediction result, and judge whether the device operation status has returned to normal;

[0018] If it is recognized that the device operation status has returned to normal and the operation status data is within the normal operation status threshold range within the set time, end the device maintenance task, otherwise, adjust the maintenance plan and continue to optimize the maintenance measures until the sanitation robot device completely returns to normal.

[0019] Further, the step of obtaining the operation status data of the sanitation robot cluster includes:

[0020] According to the operation status of the sanitation robot cluster, obtain the real-time data of the operation status data and construct a data acquisition model;

[0021] Optimize the data acquisition strategy by using the sparrow algorithm, and determine the optimal observation frequency and observation range of each parameter in the operation status data through iterative search;

[0022] During the optimization process, the data acquisition model is dynamically adjusted according to the changes in the observation frequency and observation range;

[0023] Preprocess the operation status data to remove outliers and noise data from the operation status data;

[0024] Use clustering algorithms to analyze the preprocessed operation status data and identify data characteristics and patterns under different operation states;

[0025] According to the clustering results, establish the mapping relationship between the operation state and data characteristics to achieve real-time monitoring and early warning of the operation status of the sanitation robot cluster;

[0026] Apply the optimized data acquisition strategy to the sanitation robot cluster, continuously monitor the data acquisition efficiency and quality, and dynamically adjust the optimization algorithm according to the feedback results to achieve the adaptive optimization of the data acquisition strategy.

[0027] Furthermore, compare the obtained operation status data with the preset normal operation status threshold. If the operation status data exceeds the normal operation status threshold, the steps of dynamically adjusting the observation frequency of the operation status data and narrowing the observation range include:

[0028] Transmit the operation status data to the data processing module;

[0029] The data processing module uses time series analysis algorithms to perform trend analysis on the historical data of each parameter in the operation status data to obtain the change trend curves of each parameter;

[0030] According to the preset normal threshold ranges of each parameter, judge whether the real-time data of each current parameter in the operation status data exceeds the normal threshold range. If the real-time data of each current parameter exceeds the normal threshold range, trigger an alarm;

[0031] If it is identified that the real-time data of a certain current parameter continuously exceeds the normal threshold range multiple times, dynamically adjust the observation frequency of the current parameter;

[0032] At the same time, according to the change trend curve, narrow the observation range of the current parameter;

[0033] Use the Kalman filter algorithm to filter the real-time data of the current parameter to remove the noise interference in the operation status data;

[0034] Transmit the adjusted observation frequency, observation range, and the filtered real-time data to the data visualization module, and the data visualization module displays the change situation of the current parameter in real time.

[0035] Further, the steps of obtaining the cumulative operation duration of the sanitation robot cluster and adjusting the duration weight parameter in the fault prediction model when the cumulative operation duration exceeds the preset safety duration threshold include:

[0036] Obtain the real-time operation data of the sanitation robot cluster and transmit the real-time operation data to the ground control center; the real-time operation data includes the operation status and operation duration information of each sanitation robot.

[0037] Analyze and calculate the cumulative operation duration of the sanitation robot cluster through the sparrow algorithm to obtain the total operation duration and average operation duration of the cluster.

[0038] Judge whether the total operation duration of the cluster exceeds the preset safety duration threshold. If the total operation duration of the cluster exceeds the safety duration threshold, trigger the adjustment mechanism of the fault prediction model.

[0039] Dynamically adjust the duration weight parameter in the fault prediction model according to the degree of exceeding the safety duration threshold, and assign a higher weight to the duration factor.

[0040] By increasing the duration weight, make the fault prediction model more biased towards the duration factor.

[0041] Based on the adjusted fault prediction model, re-evaluate the fault risk of the sanitation robot cluster and identify individual sanitation robots with high risk.

[0042] For high-risk sanitation robots, issue a return or landing command in a timely manner, and at the same time dispatch standby sanitation robots to take over the work tasks of high-risk sanitation robots.

[0043] Further, according to the adjusted duration weight parameter, recalculate the fault prediction result, comprehensively consider the cumulative operation duration and operation status data, and judge whether there is a risk of fatigue wear and tear for each sanitation robot device. If there is a risk of fatigue wear and tear, the steps of generating a device fault warning signal include:

[0044] Obtain the adjusted duration weight parameter and input the adjusted duration weight parameter into the fault prediction model.

[0045] According to the fault prediction model, calculate the fault prediction result of the sanitation robot device.

[0046] Obtain the operation duration data and status monitoring data of the sanitation robot device.

[0047] Input the operation duration data and status monitoring data of the sanitation robot device into the fatigue wear and tear risk assessment model.

[0048] The fatigue loss risk assessment model comprehensively analyzes the operation duration data and status monitoring data to determine whether there is a fatigue loss risk for the sanitation robot equipment;

[0049] If it is identified that there is a fatigue loss risk for the sanitation robot equipment, a device fault warning signal is generated;

[0050] The device fault warning signal is sent to the sanitation robot equipment management system.

[0051] Further, the steps of obtaining the device fault warning signal, combining the operation status data and the cumulative operation duration, generating a targeted device maintenance suggestion plan, and determining the maintenance priority of each sanitation robot equipment according to the fault risk level include:

[0052] Perform data cleaning and feature engineering on the operation status data and the cumulative operation duration, remove outliers and redundant data, and extract key features;

[0053] According to the device historical operation data, establish a fault warning model using the support vector machine algorithm, and train and optimize the fault warning model;

[0054] When receiving the device fault warning signal, combine the current status data and operation duration data of the device, and generate a targeted device maintenance suggestion plan through the decision tree algorithm;

[0055] Adopt the association rule mining algorithm to analyze the association relationship between the device fault type, fault cause and maintenance measures;

[0056] According to the severity of the device fault warning signal and the possibility of the fault occurrence, calculate the fault risk level of each sanitation robot equipment through the fuzzy comprehensive evaluation method;

[0057] Based on the fault risk level of each sanitation robot equipment and the limitation of maintenance resources, use the integer programming model to determine the maintenance priority information of each sanitation robot equipment;

[0058] Display the device maintenance suggestion plan and maintenance priority information through a visual dashboard.

[0059] Further, the steps of optimizing the execution order of the maintenance tasks of each sanitation robot equipment and giving priority to including the equipment with a high fault risk level and a high maintenance priority in the maintenance tasks and executing them in a timely manner include:

[0060] Obtain the historical fault data and maintenance records of multiple sanitation robot equipment, and calculate the fault risk level and maintenance priority value of each sanitation robot equipment through big data analysis;

[0061] According to the failure risk level and maintenance priority value, the sparrow algorithm is used to optimize and sort the execution order of the maintenance tasks of the sanitation robot equipment, and the optimal execution order sequence of the equipment maintenance tasks is obtained;

[0062] According to the optimized execution order of the equipment maintenance tasks, a equipment maintenance plan is generated, and the sanitation robot equipment with a high failure risk level and a high maintenance priority is preferentially included in the maintenance plan;

[0063] Obtain the real-time operation status data of the sanitation robot equipment. If it is identified that the key operation parameters of the sanitation robot equipment exceed the preset safety threshold range, an early warning maintenance task for the equipment is triggered, and the maintenance task priority of the current equipment is set to the highest;

[0064] According to the equipment maintenance plan, the maintenance personnel are scheduled in real time to execute the equipment maintenance tasks;

[0065] If it is identified that an early warning maintenance task appears, the personnel are preferentially scheduled to handle the early warning maintenance task;

[0066] The equipment operation data is collected and analyzed in real time through the Internet of Things platform, and the failure prediction model is continuously optimized by using machine learning algorithms, and the failure risk level of the equipment is dynamically updated;

[0067] According to the latest equipment failure risk level, the sparrow algorithm is periodically re-executed to optimize the execution order of the equipment maintenance tasks, generate a new equipment maintenance plan, and form a dynamic optimization closed loop for the equipment maintenance tasks.

[0068] Furthermore, according to the execution effect of the equipment maintenance tasks, the relevant parameters of the failure prediction model are dynamically updated, and the equipment failure prediction result is recalculated. The steps to judge whether the equipment operation status has returned to normal include:

[0069] Obtain the execution effect data of the equipment maintenance tasks; the execution effect data includes equipment operation parameters and maintenance operation records;

[0070] According to the obtained execution effect data, the relevant parameters of the failure prediction model are dynamically adjusted; the relevant parameters of the failure prediction model include model weights and thresholds;

[0071] Using the adjusted failure prediction model, recalculate the failure probability of the sanitation robot equipment to obtain the updated equipment failure prediction result;

[0072] Compare the updated equipment failure prediction result with the preset normal operation threshold to judge whether the equipment operation status has returned to normal;

[0073] If the equipment operation status has returned to normal, it is determined that the equipment maintenance task execution is effective, and the current maintenance strategy is maintained;

[0074] If the operating state of the device does not return to normal, it is determined that the execution of the device maintenance task is invalid, and the maintenance strategy needs to be further optimized;

[0075] According to the judgment result of whether the operating state of the device returns to normal, dynamically adjust the execution frequency and maintenance content of the device maintenance task to form a closed-loop optimization.

[0076] Further, if it is recognized that the operating state of the device returns to normal and all the operating state data are within the normal operating state threshold range within the set time, the device maintenance task is ended; otherwise, the maintenance plan needs to be adjusted, and the maintenance measures are continuously optimized until the sanitation robot device completely returns to normal. The steps include:

[0077] Obtain the real-time operating state data of the sanitation robot device, and judge whether the operating state of the device returns to normal according to the preset safety threshold range;

[0078] If it is recognized that the operating state of the device returns to normal, continuously monitor the device operating data, and judge whether there are any signs of failure in the sanitation robot device within the subsequent set time through big data analysis technology;

[0079] If it is recognized that there are no signs of failure in the sanitation robot device within the subsequent set time, it is determined that the sanitation robot device has completely returned to normal, generate an end instruction for the device maintenance task, and terminate the maintenance task of the device;

[0080] If it is recognized that the operating state of the device does not return to normal, or there are signs of failure within a subsequent period of time, trigger the adjustment mechanism of the maintenance plan, and automatically optimize the maintenance measures of the device through machine learning algorithms;

[0081] During the adjustment of the maintenance plan, adopt the decision tree algorithm, comprehensively consider the historical operating data, current operating state and maintenance cost factors of the device, and obtain the optimal combination of maintenance measures;

[0082] Send the optimized maintenance measures to the sanitation robot device, remotely control the device to perform corresponding maintenance operations through Internet of Things technology, and monitor the change of the operating state of the device in real time;

[0083] According to the change of the operating state of the device, loop through the above steps until the device completely returns to normal, meets the conditions for ending the maintenance task, and outputs a completion signal for the device maintenance task.

[0084] Another aspect of the present invention relates to a sanitation robot cluster state monitoring and fault warning system, which is applied to the above-mentioned sanitation robot cluster state monitoring and fault warning method. The sanitation robot cluster state monitoring and fault warning system includes:

[0085] An acquisition module for acquiring the operation status data of a sanitation robot cluster; the operation status data includes temperature, voltage, and current data;

[0086] A comparison module for comparing the acquired operation status data with a preset normal operation status threshold. If the operation status data exceeds the normal operation status threshold, the observation frequency of the operation status data is dynamically adjusted to narrow the observation range;

[0087] An adjustment module for acquiring the cumulative operation duration of the sanitation robot cluster. If the cumulative operation duration exceeds a preset safety duration threshold, the duration weight parameter in the fault prediction model is adjusted;

[0088] A first judgment module for recalculating the fault prediction result according to the adjusted duration weight parameter, comprehensively considering the cumulative operation duration and the operation status data, and judging whether there is a risk of fatigue wear for each sanitation robot device. If there is a risk of fatigue wear, a device fault warning signal is generated;

[0089] A determination module for acquiring the device fault warning signal, generating a targeted device maintenance suggestion plan in combination with the operation status data and the cumulative operation duration, and determining the maintenance priority of each sanitation robot device according to the fault risk level;

[0090] An optimization module for optimizing the execution order of the maintenance tasks of each sanitation robot device, and preferentially including the devices with a high fault risk level and a high maintenance priority in the maintenance tasks and executing them in a timely manner;

[0091] A second judgment module for dynamically updating the relevant parameters of the fault prediction model according to the execution effect of the device maintenance task, recalculating the device fault prediction result, and judging whether the device operation status has returned to normal;

[0092] A maintenance module for ending the device maintenance task if it is recognized that the device operation status has returned to normal and the operation status data is within the normal operation status threshold within a set time. Otherwise, the maintenance plan needs to be adjusted, and the maintenance measures are continuously optimized until the sanitation robot device is completely restored to normal.

[0093] The beneficial effects obtained by the present invention are:

[0094] The present invention discloses a method and system for monitoring the status and fault warning of a sanitation robot cluster, which includes obtaining the operation status data of the sanitation robot cluster; comparing the obtained operation status data with a preset normal operation status threshold. If the operation status data exceeds the normal operation status threshold, the observation frequency of the operation status data is dynamically adjusted to narrow the observation range; obtaining the cumulative operation duration of the sanitation robot cluster. If the cumulative operation duration exceeds a preset safety duration threshold, the duration weight parameter in the fault prediction model is adjusted; according to the adjusted duration weight parameter, the fault prediction result is recalculated. Considering the cumulative operation duration and operation status data comprehensively, it is judged whether there is a risk of fatigue loss for each sanitation robot device. If there is a risk of fatigue loss, a device fault warning signal is generated; obtaining the device fault warning signal, combining the operation status data and the cumulative operation duration, generating a targeted device maintenance suggestion plan, and determining the maintenance priority of each sanitation robot device according to the fault risk level; optimizing the execution order of the maintenance tasks of each sanitation robot device, and giving priority to the devices with high fault risk level and high maintenance priority in each sanitation robot device to be included in the maintenance tasks and executed in a timely manner; according to the execution effect of the device maintenance tasks, dynamically updating the relevant parameters of the fault prediction model, recalculating the device fault prediction result, and judging whether the device operation status has returned to normal; if it is recognized that the device operation status has returned to normal and the operation status data is within the normal operation status threshold range within the set time, the device maintenance task is ended, otherwise the maintenance plan needs to be adjusted and the maintenance measures are continuously optimized until the sanitation robot device completely returns to normal. The method and system for monitoring the status and fault warning of the sanitation robot cluster disclosed by the present invention realize the high efficiency and accuracy of intelligent monitoring and fault warning of the status of the sanitation robot cluster, improve the device reliability and operation efficiency, and reduce the maintenance cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] Figure 1 FIG. is a schematic flowchart of an embodiment of a method for monitoring the status and fault warning of a sanitation robot cluster according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0096] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings of the specification and specific embodiments.

[0097] As Figure 1 shown, the first embodiment of the present invention proposes a method for monitoring the status and fault warning of a sanitation robot cluster, including the following steps:

[0098] Step S100, obtaining the operation status data of the sanitation robot cluster; the operation status data includes temperature, voltage and current data.

[0099] Obtain the operation status data such as temperature, voltage, and current of the sanitation robot cluster. For different types of data parameters, use the sparrow algorithm to optimize the data acquisition strategy and determine the optimal observation frequency and observation range for each parameter.

[0100] Step S200: Compare the obtained operation status data with the preset normal operation status threshold. If the operation status data exceeds the normal operation status threshold, then dynamically adjust the observation frequency of the operation status data and narrow the observation range.

[0101] Based on the collected temperature, voltage, and current data, judge the change trend of each parameter. If the parameter change exceeds the preset normal threshold range, then dynamically adjust the observation frequency of the parameter, narrow the observation range, and improve the data acquisition accuracy and real-time performance of the parameter.

[0102] Step S300: Obtain the cumulative operation duration of the sanitation robot cluster. If the cumulative operation duration exceeds the preset safety duration threshold, then adjust the duration weight parameter in the fault prediction model.

[0103] Analyze the cumulative operation duration of the sanitation robot cluster through the sparrow algorithm. If the cluster operation duration exceeds the preset safety duration threshold, then adjust the duration weight parameter in the fault prediction model and adopt a more conservative fault prediction strategy.

[0104] Step S400: According to the adjusted duration weight parameter, recalculate the fault prediction result. Considering the cumulative operation duration and operation status data comprehensively, judge whether there is a risk of fatigue loss for each sanitation robot device. If there is a risk of fatigue loss, then generate a device fault warning signal.

[0105] According to the adjusted duration weight parameter, recalculate the fault prediction result. Considering the operation duration and status data comprehensively, judge whether there is a risk of fatigue loss for the sanitation robot device. If there is a risk, then generate a device fault warning signal.

[0106] Step S500: Obtain the device fault warning signal, combine the operation status data and cumulative operation duration, generate a targeted device maintenance suggestion plan, and determine the maintenance priority for each sanitation robot device according to the fault risk level.

[0107] Obtain the device fault warning signal, combine the status data such as temperature, voltage, and current and the operation duration data, generate a targeted device maintenance suggestion plan, and determine the maintenance priority for each device according to the fault risk level.

[0108] Step S600: Optimize the execution order of the maintenance tasks for each sanitation robot device, and give priority to including the devices with high fault risk level and high maintenance priority in the maintenance tasks and execute them in a timely manner.

[0109] The maintenance task execution order of multiple devices is optimized by using the sparrow algorithm, and the devices with high failure risk levels and high maintenance priorities are preferentially included in the maintenance tasks and executed in a timely manner.

[0110] Step S700: Dynamically update the relevant parameters of the fault prediction model according to the execution effect of the device maintenance task, recalculate the device fault prediction result, and determine whether the device operating state has returned to normal.

[0111] Dynamically update the relevant parameters of the fault prediction model according to the execution effect of the device maintenance task, recalculate the device fault prediction result, and determine whether the device operating state has returned to normal.

[0112] Step S800: If it is recognized that the device operating state has returned to normal and the operating state data is within the normal operating state threshold range within the set time, then end the device maintenance task; otherwise, it is necessary to adjust the maintenance plan and continue to optimize the maintenance measures until the sanitation robot device has fully returned to normal.

[0113] If the device operating state has returned to normal, key parameters such as temperature, voltage, and current are within the safety threshold range, and there are no repeated fault signs in the subsequent period of time, then the maintenance task of this device can be ended; otherwise, it is necessary to adjust the maintenance plan and continue to optimize the maintenance measures until the device has fully returned to normal.

[0114] Furthermore, for the sanitation robot cluster status monitoring and fault warning method provided in this embodiment, step S100 includes:

[0115] Step S110: According to the operating state of the sanitation robot cluster, obtain the real-time data of the operating state data and construct a data acquisition model.

[0116] According to the operating state of the sanitation robot cluster, obtain the real-time data of parameters such as temperature, voltage, and current, and construct a data acquisition model.

[0117] Step S120: Use the sparrow algorithm to optimize the data acquisition strategy, and determine the optimal observation frequency and observation range of each parameter in the operating state data through iterative search.

[0118] Use the sparrow algorithm to optimize the data acquisition strategy, and determine the optimal observation frequency and observation range of each parameter through iterative search.

[0119] Step S130: During the optimization process, dynamically adjust the data acquisition model according to the changes in the observation frequency and observation range.

[0120] During the optimization process, dynamically adjust the data acquisition model according to the changes in the observation frequency and observation range to improve the adaptability of data acquisition.

[0121] Step S140: Preprocess the operation status data to eliminate outliers and noise data in the operation status data.

[0122] Preprocess the parameter data such as the collected temperature, voltage, and current, eliminate outliers and noise data, and improve the data quality.

[0123] Step S150: Use a clustering algorithm to analyze the preprocessed operation status data and identify the data characteristics and patterns under different operation states.

[0124] Use a clustering algorithm to analyze the preprocessed data and identify the data characteristics and patterns under different operation states.

[0125] Step S160: According to the clustering results, establish a mapping relationship between the operation status and data characteristics to realize real-time monitoring and early warning of the operation status of the sanitation robot cluster.

[0126] According to the clustering results, establish a mapping relationship between the operation status and data characteristics to realize real-time monitoring and early warning of the operation status of the sanitation robot cluster.

[0127] Step S170: Apply the optimized data acquisition strategy to the sanitation robot cluster, continuously monitor the data acquisition efficiency and quality, and dynamically adjust the optimization algorithm according to the feedback results to realize the adaptive optimization of the data acquisition strategy.

[0128] Apply the optimized data acquisition strategy to the sanitation robot cluster, continuously monitor the data acquisition efficiency and quality, and dynamically adjust the optimization algorithm according to the feedback results to realize the adaptive optimization of the data acquisition strategy.

[0129] Furthermore, for the method for monitoring the status and early warning of faults of the sanitation robot cluster provided in this embodiment, step S200 includes:

[0130] Step S210: Transmit the operation status data to the data processing module.

[0131] Obtain the real-time acquisition data of multiple parameters such as temperature, voltage, and current, and transmit the data to the data processing module.

[0132] Step S220: The data processing module uses a time series analysis algorithm to perform trend analysis on the historical data of each parameter in the operation status data to obtain the change trend curve of each parameter.

[0133] In the data processing module, use a time series analysis algorithm to perform trend analysis on the historical data of each parameter to obtain the change trend curve of each parameter.

[0134] Step S230: According to the preset normal threshold ranges of each parameter, determine whether the real-time data of each current parameter in the operation status data exceeds the normal threshold range. If the real-time data of each current parameter exceeds the normal threshold range, an alarm is triggered.

[0135] According to the preset normal threshold ranges of each parameter, determine whether the real-time data of each current parameter exceeds the normal threshold range. If it exceeds, an alarm is triggered.

[0136] Step S240: If it is recognized that the real-time data of a certain current parameter exceeds the normal threshold range continuously for multiple times, dynamically adjust the observation frequency of the current parameter.

[0137] When the real-time data of a certain parameter exceeds the normal threshold range continuously for multiple times, dynamically adjust the observation frequency of this parameter. For example, adjust the data collection once per minute to once every 10 seconds.

[0138] Step S250: At the same time, according to the change trend curve, narrow the observation range of the current parameter.

[0139] At the same time, according to the change trend curve of this parameter, narrow the observation range of this parameter. For example, adjust the original observation range of 0 - 100 degrees to 80 - 100 degrees to improve the data collection accuracy of this parameter.

[0140] Step S260: Use the Kalman filter algorithm to filter the real-time data of the current parameter to remove the noise interference in the operation status data.

[0141] Use the Kalman filter algorithm to filter the real-time data of this parameter to remove the noise interference in the data and improve the real-time performance and reliability of the data.

[0142] Step S270: Transmit the adjusted observation frequency, observation range, and the filtered real-time data to the data visualization module, and the data visualization module displays the change situation of the current parameter in real time.

[0143] Transmit the adjusted observation frequency, observation range, and the filtered real-time data to the data visualization module, and display the change situation of this parameter in real time through forms such as curve graphs and bar graphs, providing an intuitive monitoring view for the operation and maintenance personnel.

[0144] Furthermore, for the method for monitoring the status and fault warning of the sanitation robot cluster provided in this embodiment, step S300 includes:

[0145] Step S310: Obtain the real-time operation data of the sanitation robot cluster, and transmit the real-time operation data to the ground control center; the real-time operation data includes the operation status and operation duration information of each sanitation robot.

[0146] Obtain the real-time operation data of the sanitation robot cluster, including the operation status and operation duration of each sanitation robot, and transmit the data to the ground control center.

[0147] Step S320: Analyze and calculate the cumulative operation duration of the sanitation robot cluster through the sparrow algorithm to obtain the total operation duration and average operation duration of the cluster.

[0148] Analyze and calculate the cumulative operation duration of the sanitation robot cluster through the sparrow algorithm to obtain the total operation duration and average operation duration of the cluster.

[0149] Step S330: Determine whether the total operation duration of the cluster exceeds the preset safety duration threshold. If the total operation duration of the cluster exceeds the safety duration threshold, trigger the adjustment mechanism of the fault prediction model.

[0150] Determine whether the total operation duration of the cluster exceeds the preset safety duration threshold. If it exceeds the threshold, trigger the adjustment mechanism of the fault prediction model.

[0151] Step S340: Dynamically adjust the duration weight parameter in the fault prediction model according to the degree of exceeding the safety duration threshold, and assign a higher weight to the duration factor.

[0152] Dynamically adjust the duration weight parameter in the fault prediction model according to the degree of exceeding the safety duration threshold, and assign a higher weight to the duration factor.

[0153] Step S350: Make the fault prediction model more biased towards the duration factor by increasing the duration weight.

[0154] Make the fault prediction model more biased towards the duration factor by increasing the duration weight, so as to adopt a more conservative fault prediction strategy.

[0155] Step S360: Based on the adjusted fault prediction model, re-evaluate the fault risk of the sanitation robot cluster and identify individual sanitation robots with high risk.

[0156] Based on the adjusted fault prediction model, re-evaluate the fault risk of the sanitation robot cluster and identify individual sanitation robots with high risk.

[0157] Step S370: For high-risk sanitation robots, issue a return or landing command in a timely manner, and at the same time dispatch standby sanitation robots to take over the work tasks of high-risk sanitation robots.

[0158] For high-risk sanitation robots, issue a return or landing command in a timely manner, and at the same time dispatch standby sanitation robots to take over their tasks to ensure the continuity and safety of the cluster tasks.

[0159] Furthermore, the method for monitoring the status and fault warning of the sanitation robot cluster provided in this embodiment, step S400 includes:

[0160] Step S410, obtain the adjusted duration weight parameter, and input the adjusted duration weight parameter into the fault prediction model.

[0161] Obtain the adjusted duration weight parameter and input the parameter into the fault prediction model.

[0162] Step S420, calculate the fault prediction result of the sanitation robot device according to the fault prediction model.

[0163] Calculate the fault prediction result of the sanitation robot device according to the fault prediction model.

[0164] Step S430, obtain the operation duration data and status monitoring data of the sanitation robot device.

[0165] Obtain the operation duration data and status monitoring data of the sanitation robot device.

[0166] Step S440, input the operation duration data and status monitoring data of the sanitation robot device into the fatigue loss risk assessment model.

[0167] Input the operation duration data and status monitoring data of the sanitation robot device into the fatigue loss risk assessment model.

[0168] Step S450, the fatigue loss risk assessment model comprehensively analyzes the operation duration data and status monitoring data to determine whether there is a fatigue loss risk for the sanitation robot device.

[0169] Through the fatigue loss risk assessment model, comprehensively analyze the operation duration and status data to determine whether there is a fatigue loss risk for the device.

[0170] Step S460, if it is recognized that the sanitation robot device has a fatigue loss risk, generate a device fault warning signal.

[0171] If the sanitation robot device has a fatigue loss risk, trigger the generation of a device fault warning signal.

[0172] Step S470, send the prepared fault warning signal to the sanitation robot device management system.

[0173] Send the device fault warning signal to the sanitation robot device management system to prompt the staff to repair and maintain the device in time.

[0174] Furthermore, the method for monitoring the status and fault warning of the sanitation robot cluster provided in this embodiment, step S500 includes:

[0175] Step S510: Clean the operation status data and cumulative operation duration, and perform feature engineering to remove outliers and redundant data, and extract key features.

[0176] Obtain multi-dimensional status data such as temperature, voltage, and current of the device and operation duration data, and through data cleaning and feature engineering, remove outliers and redundant data, and extract key features.

[0177] Step S520: Based on the historical operation data of the device, establish a fault warning model using the support vector machine algorithm, and train and optimize the fault warning model.

[0178] Based on the historical operation data of the device, establish a device fault warning model using the support vector machine algorithm, and through model training and optimization, improve the accuracy and recall rate of fault warning.

[0179] Step S530: When receiving a device fault warning signal, combine the current status data and operation duration data of the device, and generate a targeted device maintenance suggestion plan through the decision tree algorithm.

[0180] When receiving a device fault warning signal, combine the current status data and operation duration data of the device, and generate a targeted device maintenance suggestion plan through the decision tree algorithm.

[0181] Step S540: Use the association rule mining algorithm to analyze the association relationships among device fault types, fault causes, and maintenance measures.

[0182] Use the association rule mining algorithm to analyze the association relationships among device fault types, fault causes, and maintenance measures, and optimize the operability and effectiveness of the device maintenance suggestion plan.

[0183] Step S550: According to the severity of the device fault warning signal and the probability of the fault occurring, calculate the fault risk level of each sanitation robot device through the fuzzy comprehensive evaluation method.

[0184] According to the severity of the device fault warning signal and the probability of the fault occurring, calculate the fault risk level of each device through the fuzzy comprehensive evaluation method.

[0185] Step S560: Based on the fault risk levels of each sanitation robot device and the limitations of maintenance resources, use an integer programming model to determine the maintenance priority information of each sanitation robot device.

[0186] Based on the fault risk levels of the devices and the limitations of maintenance resources, use an integer programming model to determine the maintenance priorities of each device and formulate an optimal device maintenance plan.

[0187] Step S570: Display the equipment maintenance suggestion plan and maintenance priority information through a visual dashboard.

[0188] The personalized equipment maintenance suggestion plan and maintenance priority information are displayed to maintenance personnel through a visual dashboard to assist them in making equipment maintenance decisions and improving maintenance efficiency and accuracy.

[0189] Furthermore, for the sanitation robot cluster status monitoring and fault warning method provided in this embodiment, step S600 includes:

[0190] Step S610: Obtain the historical fault data and maintenance records of multiple sanitation robot devices, and calculate the fault risk level and maintenance priority value of each sanitation robot device through big data analysis.

[0191] Obtain the historical fault data and maintenance records of multiple devices, and calculate the fault risk level and maintenance priority value of each device through big data analysis.

[0192] Step S620: According to the fault risk level and maintenance priority value, use the sparrow algorithm to optimize and sort the execution order of the maintenance tasks of the sanitation robot devices, and obtain the optimal execution order sequence of the equipment maintenance tasks.

[0193] According to the fault risk level and maintenance priority value of the equipment, use the sparrow algorithm to optimize and sort the execution order of the maintenance tasks of the equipment, and obtain the optimal execution order sequence of the equipment maintenance tasks.

[0194] Step S630: Generate an equipment maintenance plan according to the optimized execution order of the equipment maintenance tasks, and give priority to including the sanitation robot devices with high fault risk level and high maintenance priority in the maintenance plan.

[0195] Generate an equipment maintenance plan according to the optimized execution order of the equipment maintenance tasks, and give priority to including the devices with high fault risk level and high maintenance priority in the maintenance plan.

[0196] Step S640: Obtain the real-time operation status data of the sanitation robot device. If it is recognized that the key operation parameters of the sanitation robot device exceed the preset safety threshold range, then trigger the warning maintenance task of the device, and set the maintenance task priority of the current device to the highest.

[0197] Obtain the real-time operation status data of the device. If the key operation parameters of the device exceed the preset safety threshold range, then trigger the warning maintenance task of the device, and set the maintenance task priority of the device to the highest.

[0198] Step S650: According to the equipment maintenance plan, schedule maintenance personnel to execute the equipment maintenance tasks in real time.

[0199] According to the equipment maintenance plan, schedule maintenance personnel in real time to perform equipment maintenance tasks.

[0200] Step S660: If a warning maintenance task is identified, prioritize scheduling personnel to handle the warning maintenance task.

[0201] If a warning maintenance task occurs, prioritize scheduling personnel to handle the warning maintenance task.

[0202] Step S670: Real-time collect and analyze equipment operation data through the Internet of Things platform, continuously optimize the fault prediction model using machine learning algorithms, and dynamically update the fault risk level of the equipment.

[0203] Real-time collect and analyze equipment operation data through the Internet of Things platform, continuously optimize the equipment fault prediction model using machine learning algorithms, and dynamically update the fault risk level of the equipment.

[0204] Step S680: According to the latest equipment fault risk level, periodically re-execute the sparrow algorithm to optimize the execution order of equipment maintenance tasks, generate a new equipment maintenance plan, and form a dynamic optimization closed-loop for equipment maintenance tasks.

[0205] According to the latest equipment fault risk level, periodically re-execute the sparrow algorithm to optimize the execution order of equipment maintenance tasks, generate a new equipment maintenance plan, and form a dynamic optimization closed-loop for equipment maintenance tasks.

[0206] Furthermore, for the method for monitoring the status and fault warning of a sanitation robot cluster provided in this embodiment, step S700 includes:

[0207] Step S710: Obtain the execution effect data of equipment maintenance tasks; the execution effect data includes equipment operation parameters and maintenance operation records.

[0208] Obtain the execution effect data of equipment maintenance tasks, including equipment operation parameters and maintenance operation records, etc.

[0209] Step S720: Dynamically adjust the relevant parameters of the fault prediction model according to the obtained execution effect data; the relevant parameters of the fault prediction model include model weights and thresholds.

[0210] Dynamically adjust the relevant parameters of the fault prediction model according to the obtained execution effect data, including model weights and thresholds, etc.

[0211] Step S730: Use the adjusted fault prediction model to recalculate the fault probability of the sanitation robot equipment to obtain an updated equipment fault prediction result.

[0212] Use the adjusted fault prediction model to recalculate the fault probability of the equipment to obtain an updated equipment fault prediction result.

[0213] Step S740: Compare the updated device fault prediction result with a preset normal operation threshold to determine whether the device operation status has returned to normal.

[0214] Compare the updated device fault prediction result with a preset normal operation threshold to determine whether the device operation status has returned to normal.

[0215] Step S750: If the device operation status has returned to normal, determine that the device maintenance task has been effectively executed and maintain the current maintenance strategy.

[0216] If the device operation status has returned to normal, determine that the device maintenance task has been effectively executed and maintain the current maintenance strategy.

[0217] Step S760: If the device operation status has not returned to normal, determine that the device maintenance task has not been effectively executed and further optimize the maintenance strategy.

[0218] If the device operation status has not returned to normal, determine that the device maintenance task has not been effectively executed and further optimize the maintenance strategy.

[0219] Step S770: Dynamically adjust the execution frequency and maintenance content of the device maintenance task according to the judgment result of whether the device operation status has returned to normal to form a closed-loop optimization.

[0220] Dynamically adjust the execution frequency and maintenance content of the device maintenance task according to the judgment result of whether the device operation status has returned to normal to form a closed-loop optimization.

[0221] Furthermore, for the sanitation robot cluster status monitoring and fault warning method provided in this embodiment, step S800 includes:

[0222] Step S810: Obtain the real-time operation status data of the sanitation robot device, and determine whether the device operation status has returned to normal according to a preset safety threshold range.

[0223] Obtain the real-time operation status data of the device, including key parameters such as temperature, voltage, and current, and determine whether the device operation status has returned to normal according to a preset safety threshold range.

[0224] Step S820: If it is recognized that the device operation status has returned to normal, continuously monitor the device operation data, and use big data analysis technology to determine whether there are any fault signs in the sanitation robot device within a subsequent set time.

[0225] If the device operation status has returned to normal, continuously monitor the device operation data, and use big data analysis technology to determine whether there are any fault signs in the device within a subsequent period of time.

[0226] Step S830: If it is recognized that there are no signs of failure in the sanitation robot device within the subsequent set time, it is determined that the sanitation robot device has fully returned to normal, an end instruction for the device maintenance task is generated, and the maintenance task of the device is terminated.

[0227] If there are no signs of failure in the device within a subsequent period of time, it can be determined that the device has fully returned to normal, an end instruction for the device maintenance task is generated, and the maintenance task of the device is terminated.

[0228] Step S840: If it is recognized that the operating state of the device has not returned to normal, or signs of failure appear within a subsequent period of time, a maintenance plan adjustment mechanism is triggered, and the maintenance measures of the device are automatically optimized through a machine learning algorithm.

[0229] If the operating state of the device has not returned to normal, or signs of failure appear within a subsequent period of time, a maintenance plan adjustment mechanism is triggered, and the maintenance measures of the device are automatically optimized through a machine learning algorithm.

[0230] Step S850: During the adjustment of the maintenance plan, a decision tree algorithm is adopted, and by comprehensively considering the historical operation data, current operating state, and maintenance cost factors of the device, an optimal combination of maintenance measures is obtained.

[0231] During the adjustment of the maintenance plan, a decision tree algorithm is adopted, and by comprehensively considering the historical operation data, current operating state, and maintenance cost and other factors of the device, an optimal combination of maintenance measures is obtained.

[0232] Step S860: The optimized maintenance measures are sent to the sanitation robot device, and through the Internet of Things technology, the device is remotely controlled to perform corresponding maintenance operations, and the change of the operating state of the device is monitored in real time.

[0233] The optimized maintenance measures are sent to the device, and through the Internet of Things technology, the device is remotely controlled to perform corresponding maintenance operations, and the change of the operating state of the device is monitored in real time.

[0234] Step S870: According to the change of the operating state of the device, the above steps are executed in a loop until the device fully returns to normal, the condition for ending the maintenance task is met, and a completion signal for the device maintenance task is output.

[0235] According to the change of the operating state of the device, the above steps are executed in a loop until the device fully returns to normal, the condition for ending the maintenance task is met, and a completion signal for the device maintenance task is output.

[0236] The present invention relates to a sanitation robot cluster status monitoring and fault warning system, which is applied to the above-mentioned sanitation robot cluster status monitoring and fault warning method. The sanitation robot cluster status monitoring and fault warning system includes an acquisition module, a comparison module, an adjustment module, a first judgment module, a determination module, an optimization module, a second judgment module, and a maintenance module. Among them, the acquisition module is used to acquire the operation status data of the sanitation robot cluster; the operation status data includes temperature, voltage, and current data; the comparison module is used to compare the acquired operation status data with the preset normal operation status threshold. If the operation status data exceeds the normal operation status threshold, the observation frequency of the operation status data is dynamically adjusted to narrow the observation range; the adjustment module is used to acquire the cumulative operation duration of the sanitation robot cluster. If the cumulative operation duration exceeds the preset safety duration threshold, the duration weight parameter in the fault prediction model is adjusted; the first judgment module is used to recalculate the fault prediction result according to the adjusted duration weight parameter, comprehensively consider the cumulative operation duration and the operation status data, and judge whether there is a risk of fatigue wear for each sanitation robot device. If there is a risk of fatigue wear, a device fault warning signal is generated; the determination module is used to acquire the device fault warning signal, generate a targeted device maintenance suggestion plan in combination with the operation status data and the cumulative operation duration, and determine the maintenance priority of each sanitation robot device according to the fault risk level; the optimization module is used to optimize the execution order of the maintenance tasks of each sanitation robot device, and give priority to the devices with high fault risk level and high maintenance priority in each sanitation robot device to be included in the maintenance tasks and executed in time; the second judgment module is used to dynamically update the relevant parameters of the fault prediction model according to the execution effect of the device maintenance task, recalculate the device fault prediction result, and judge whether the device operation status has returned to normal; the maintenance module is used to end the device maintenance task if it is recognized that the device operation status has returned to normal and the operation status data is within the normal operation status threshold range within the set time, otherwise, the maintenance plan needs to be adjusted, and the maintenance measures are continuously optimized until the sanitation robot device returns to normal completely.

[0237] The following uses specific embodiments to illustrate the sanitation robot cluster status monitoring and fault warning method provided by the present application. The sanitation robot cluster status monitoring and fault warning method includes the following steps:

[0238] Step S101: Acquire the operation status data such as temperature, voltage, and current of the sanitation robot cluster. For different types of data parameters, use the sparrow algorithm to optimize the data acquisition strategy, and determine the optimal observation frequency and observation range of each parameter.

[0239] According to the operating status of the sanitation robot cluster, real-time data of parameters such as temperature, voltage, and current are obtained to build a data acquisition model. The sparrow algorithm is used to optimize the data acquisition strategy, and the optimal observation frequency and observation range of each parameter are determined through iterative search. During the optimization process, the data acquisition model is dynamically adjusted according to the changes in the observation frequency and observation range to improve the adaptability of data acquisition. The collected parameter data such as temperature, voltage, and current are preprocessed to remove outliers and noise data to improve data quality. The preprocessed data is analyzed using a clustering algorithm to identify data features and patterns under different operating states. According to the clustering results, a mapping relationship between the operating status and data features is established to achieve real-time monitoring and early warning of the operating status of the sanitation robot cluster. The optimized data acquisition strategy is applied to the sanitation robot cluster, and the data acquisition efficiency and quality are continuously monitored. The optimization algorithm is dynamically adjusted according to the feedback results to achieve adaptive optimization of the data acquisition strategy.

[0240] Specifically, during the operation of the sanitation robot cluster, the sensor collects parameter data such as temperature, voltage, and current in real time, and samples every 1 second to build a data acquisition model. The sparrow algorithm is used to optimize the data acquisition strategy. After 50 iterative searches, the temperature observation frequency is determined to be 1Hz, the voltage observation frequency is 5Hz, and the current observation frequency is 10Hz. The observation ranges are -20℃~100℃, 0V~30V, and 0A~50A, respectively. During the optimization process, the data acquisition model is dynamically adjusted according to the changes in the observation frequency and observation range to improve the adaptability of data acquisition. The collected data is preprocessed, and the 3σ criterion is used to eliminate outliers. The Kalman filter algorithm is used to remove noise and improve data quality. The preprocessed data is analyzed using the K-means clustering algorithm. The optimal number of clusters is determined to be 3 by calculating the silhouette coefficient, and the data features and patterns under different operating states such as normal operation, abnormal operation, and failure are identified. According to the clustering results, the mapping relationship between the operating state and the data features is established to achieve real-time monitoring and early warning of the operating state of the sanitation robot cluster. The optimized data collection strategy is applied to the sanitation robot cluster, and the data collection efficiency and quality are continuously monitored. When the data collection efficiency is lower than 90% or the data quality is lower than 95%, the optimization algorithm is triggered to readjust the data collection strategy to achieve adaptive optimization.

[0241] Step S102: determine the change trend of each parameter based on the collected temperature, voltage and current data. If the parameter change exceeds the preset normal threshold range, dynamically adjust the observation frequency of the parameter, narrow the observation range, and improve the data collection accuracy and real-time performance of the parameter.

[0242] Obtain the real-time acquisition data of multiple parameters such as temperature, voltage, and current, and transmit the data to the data processing module. In the data processing module, a time series analysis algorithm is used to perform trend analysis on the historical data of each parameter to obtain the change trend curve of each parameter. According to the preset normal threshold range of each parameter, judge whether the real-time data of each parameter currently exceeds the normal threshold range. If it exceeds, trigger an alarm. When the real-time data of a certain parameter continuously exceeds the normal threshold range multiple times, dynamically adjust the observation frequency of this parameter. For example, adjust the data collection once per minute to once every 10 seconds. At the same time, according to the change trend curve of this parameter, narrow the observation range of this parameter. For example, adjust the original observation range of 0-100 degrees to 80-100 degrees to improve the data collection accuracy of this parameter. Use the Kalman filter algorithm to filter the real-time data of this parameter to remove the noise interference in the data and improve the timeliness and reliability of the data. Transmit the adjusted observation frequency, observation range, and filtered real-time data to the data visualization module, and display the change situation of this parameter in real time through forms such as curve charts and bar charts to provide an intuitive monitoring view for the operation and maintenance personnel.

[0243] Specifically, it is used to predict future parameter values:

[0244] (1)

[0245] In formula (1), is the step-ahead prediction value, is the time series mean, is the autoregressive coefficient, is the autoregressive order, represents the time series at a distance from the current prediction time point retroactively back to the past The actual value for each time step. This formula is used to predict future parameter values. The system collects data on multiple parameters such as temperature, voltage, and current in real time through sensors, once every minute, and transmits the data to the data processing module. The data processing module uses the ARIMA time series analysis algorithm to model and analyze the historical data of each parameter, obtaining the change trend curve of each parameter. For example, the normal threshold range for the temperature parameter is set at 0 - 100 degrees Celsius, and the real-time data of the current temperature parameter is 120 degrees Celsius, exceeding the normal threshold range. The system automatically triggers an alarm prompt. Since the temperature parameter has exceeded the normal threshold range three times consecutively, the system dynamically adjusts the observation frequency of the temperature parameter to collect data once every 10 seconds. At the same time, according to the change trend curve of the temperature parameter, the observation range is adjusted from 0 - 100 degrees Celsius to 80 - 120 degrees Celsius to improve the accuracy of data collection. The system uses the Kalman filter algorithm to filter the real-time data of the temperature parameter, effectively removing Gaussian white noise interference in the data and improving the signal-to-noise ratio of the data. The adjusted observation frequency, observation range, and filtered real-time temperature data are transmitted to the data visualization module, which displays the change trend of the temperature parameter in real time through a dynamic curve graph and statistically counts the number of times and duration of the temperature parameter exceeding the normal threshold range within each hour in the form of a bar graph, providing an intuitive monitoring view for the operation and maintenance personnel and facilitating equipment anomaly diagnosis and predictive maintenance.

[0246] (2)

[0247] In formula (2), is the alarm probability, is the alarm rate parameter, is the duration exceeding the threshold. This formula describes the probability model for alarm triggering.

[0248] (3)

[0249] In formula (3), is the alarm function, which triggers an alarm when the parameter exceeds the threshold range , . : represents the minimum threshold allowed for the value of the time series . : represents the maximum threshold allowed for the value of the time series . This formula is used to determine whether the parameter exceeds the normal range.

[0250] (4)

[0251] In formula (4), is the autoregressive polynomial, is the differencing operator, is a moving average polynomial, is the time series observation value, is a white noise sequence. is a parameter that controls the speed of trend change. is the time point in the time series. This formula describes the basic form of the ARIMA model.

[0252] Step S103: Analyze the cumulative running duration of the sanitation robot cluster through the sparrow algorithm. If the cluster running duration exceeds the preset safety duration threshold, adjust the duration weight parameter in the fault prediction model and adopt a more conservative fault prediction strategy.

[0253] The duration weight parameter is a control variable in the fault prediction model used to represent the influence degree of the cumulative running duration on fault prediction. It is a key parameter for measuring the influence degree of the cumulative running duration of sanitation robots on the fault prediction result. Dynamically adjusting this parameter can balance the prediction conservativeness and sensitivity of the model, so as to adapt to the risk characteristics of different operation stages. When the cumulative running duration of the cluster exceeds the safety threshold, it indicates that the overall system is in a higher potential risk state, and it is necessary to adjust the weight of the running duration factor. The adjusted weight will make the model pay more attention to the running duration and tend to predict possible faults earlier. The duration weight parameter affects the output result of the fault prediction model, making robots with higher running durations account for a higher proportion in the prediction. After the weight adjustment, the model adopts a more conservative strategy (A more conservative fault prediction strategy means adopting a more cautious approach in the fault prediction model, tending to early warn of potential fault risks to minimize the possibility of faults. The core of this strategy is to reduce the risk of missed reports, that is, it is better to report more "possible faults" rather than ignore any potential risks. For example, when the prediction model finds that certain conditions are close to the fault threshold, even if the standard is not fully reached, it will trigger a fault warning in advance. Or adjust the model parameters (such as the duration weight parameter) to make it more sensitive to the key factors affecting faults (such as cumulative running duration, temperature, voltage, etc.). A more conservative strategy is to minimize the impact of robot faults in key task scenarios, enhance the reliability and safety of the system, and improve the overall risk prevention and control ability in the key operation stage.). When the running duration is short, the weight may be low, and the model is less sensitive to the running duration; when the running duration is close to or exceeds the safety duration threshold, the weight parameter increases, and the model is more inclined to think that robots with long running durations are more likely to have faults.

[0254] Obtain the real-time operation data of the sanitation robot cluster, including the operation status, operation duration and other information of each sanitation robot, and transmit the data to the ground control center. Analyze and calculate the cumulative operation duration of the sanitation robot cluster through the sparrow algorithm to obtain the total operation duration and average operation duration of the cluster. Judge whether the total operation duration of the cluster exceeds the preset safety duration threshold. If it exceeds the threshold, trigger the adjustment mechanism of the fault prediction model. Dynamically adjust the duration weight parameter in the fault prediction model according to the degree of exceeding the safety duration threshold, and assign a higher weight to the duration factor. By increasing the duration weight, make the fault prediction model more biased towards the duration factor, so as to adopt a more conservative fault prediction strategy. Based on the adjusted fault prediction model, re-evaluate the fault risk of the sanitation robot cluster and identify individual sanitation robots with high risks. For high-risk sanitation robots, issue return or landing instructions in a timely manner, and at the same time dispatch standby sanitation robots to take over their tasks to ensure the continuity and safety of the cluster tasks.

[0255] Specifically, during the task execution process of the sanitation robot cluster, the operation status data of each sanitation robot, including flight speed, altitude, attitude, battery power, etc., and the cumulative operation duration are collected in real time through on-board sensors, and the data is transmitted to the ground control center through a wireless communication module. After receiving the data, the control center uses the sparrow algorithm to analyze and calculate the cumulative operation duration of the sanitation robot cluster. Assuming that there are 10 sanitation robots in the cluster, through the iterative optimization of the sparrow algorithm, the total operation duration of the cluster is obtained as 120 hours, and the average operation duration is 12 hours. The preset safety duration threshold of the system is 100 hours. By comparison, it is found that the total operation duration of the cluster has exceeded the threshold, triggering the adjustment mechanism of the fault prediction model. Dynamically adjust the duration weight parameter in the fault prediction model according to the degree of exceeding the safety duration threshold. The original duration weight is 3, and now according to the exceeding threshold ratio of 20%, the duration weight is increased to 5, assigning a higher weight to the duration factor. The adjusted fault prediction model comprehensively considers various status parameters of the sanitation robot, and through Bayesian network reasoning, conducts a fault risk assessment on each sanitation robot in the cluster, identifying 2 high-risk sanitation robots, and their fault probabilities are 8 and 75 respectively, exceeding the warning threshold of 7. For these 2 high-risk sanitation robots, the control center issues return instructions in a timely manner to let them return to the base safely, and at the same time dispatches 2 new sanitation robots from the standby sanitation robots to take over their tasks to ensure the uninterrupted execution of the cluster tasks. Through real-time monitoring, intelligent analysis and dynamic adjustment, the safety and reliability of the sanitation robot cluster are effectively improved, and the risk of task failure is reduced.

[0256] Step S104: Recalculate the fault prediction result according to the adjusted duration weight parameter, comprehensively consider the operation duration and status data, and determine whether there is a risk of fatigue loss for the sanitation robot equipment. If there is a risk, generate an equipment fault warning signal.

[0257] Obtain the adjusted duration weight parameter and input the parameter into the fault prediction model. According to the fault prediction model, calculate the fault prediction result of the sanitation robot equipment. Obtain the operation duration data and status monitoring data of the sanitation robot equipment. Input the operation duration data and status monitoring data of the sanitation robot equipment into the fatigue loss risk assessment model. Through the fatigue loss risk assessment model, comprehensively analyze the operation duration and status data to determine whether there is a risk of fatigue loss for the equipment. If there is a risk of fatigue loss for the sanitation robot equipment, trigger the generation of an equipment fault warning signal. Send the equipment fault warning signal to the sanitation robot equipment management system to prompt the staff to repair and maintain the equipment in a timely manner.

[0258] Specifically, first, adjust and optimize the parameters of the fault prediction model through algorithms such as the least squares method to obtain the optimal model parameters. For example, by collecting the historical fault data of 100 sanitation robot equipment and using the gradient descent algorithm, reduce the mean square error of the fault prediction model from 15 to 08, improving the prediction accuracy of the model. Then, input the optimized model parameters into the fault prediction model, and the model calculates the probability of the equipment failing in a future period based on the real-time status data of the equipment. At the same time, the system also obtains the operation duration data and status monitoring data of the sanitation robot equipment, such as the equipment has run for 1000 hours and the vibration data shows that the vibration amplitude of the equipment is relatively large. Input these data into the fatigue loss risk assessment model, which uses the fuzzy comprehensive evaluation method to comprehensively analyze the operation duration and status data of the equipment to determine whether there is a risk of fatigue loss for the equipment. When the fatigue loss index of the equipment exceeds 7, trigger the generation of an equipment fault warning signal. The warning signal will be automatically sent to the sanitation robot equipment management system, and the system will send a warning message to the relevant staff in a timely manner according to the warning level, such as when the probability of failure is greater than 80% and the fatigue loss index is greater than 8, to prompt them to repair and maintain the equipment to avoid sudden failures of the equipment and ensure the safe operation of the sanitation robot.

[0259] Step S105: Obtain the equipment fault warning signal, generate a targeted equipment maintenance suggestion plan in combination with status data such as temperature, voltage, and current and operation duration data, and determine the maintenance priority of each equipment according to the fault risk level.

[0260] Obtain multi-dimensional status data such as temperature, voltage, and current of the device, as well as operation duration data. Through data cleaning and feature engineering, remove outliers and redundant data, and extract key features. Based on the historical operation data of the device, establish a device fault warning model using the support vector machine algorithm. Through model training and optimization, improve the accuracy and recall rate of fault warning. When receiving a device fault warning signal, combine the current status data and operation duration data of the device, and generate a targeted device maintenance suggestion plan through the decision tree algorithm. Use the association rule mining algorithm to analyze the association relationships among device fault types, fault causes, and maintenance measures, and optimize the operability and effectiveness of the device maintenance suggestion plan. According to the severity of the device fault warning signal and the probability of fault occurrence, calculate the fault risk level of each device through the fuzzy comprehensive evaluation method. Based on the fault risk level of the device and the limitations of maintenance resources, use the integer programming model to determine the maintenance priority of each device and formulate the optimal device maintenance plan. Display the personalized device maintenance suggestion plan and maintenance priority information to the maintenance personnel through a visual dashboard to assist them in making device maintenance decisions and improve maintenance efficiency and accuracy.

[0261] Specifically, first, multi-dimensional status data of the device such as temperature, voltage, and current are collected by sensors. For example, the temperature data range is 0 - 100 °C, the voltage data range is 0 - 380 V, and the current data range is 0 - 100 A. Meanwhile, the operation duration data of the device is recorded. Then, the collected data is cleaned and subjected to feature engineering. Outliers are removed through the 3σ principle, and redundant data is removed using the principal component analysis (PCA) method. Key features of the data such as mean, variance, and kurtosis are extracted. Next, based on the historical operation data of the device, a fault warning model is established using the support vector machine (SVM) algorithm. The model parameters are optimized through grid search and cross-validation methods to make the accuracy and recall rate of fault warning reach over 95%. When a device fault warning signal is received, combined with the current status data and operation duration data of the device, a targeted device maintenance recommendation plan is generated through the C5 decision tree algorithm. The Apriori association rule mining algorithm is used to analyze the association relationships among device fault types, fault causes, and maintenance measures, and association rules with a confidence level greater than 80% are obtained for optimizing the device maintenance recommendation plan. According to the severity of the device fault warning signal and the likelihood of the fault occurrence, the fault risk level of each device is calculated through the fuzzy comprehensive evaluation method. The risk level is divided into three levels: high, medium, and low. Based on the fault risk level of the device and the limitation of maintenance resources, a 0-1 integer programming model is used to determine the maintenance priority of each device. The objective function is to minimize the device fault risk, and the constraint condition is the upper limit of maintenance resources. Finally, the personalized device maintenance recommendation plan and maintenance priority information are displayed to the maintenance personnel through a visual dashboard. A pie chart is used to display the fault type distribution, a bar chart is used to display the fault quantity trend, and a radar chart is used to display the device health index to assist the maintenance personnel in making device maintenance decisions.

[0262] Step S106: Optimize the execution order of maintenance tasks for multiple devices using the sparrow algorithm, and give priority to including devices with a high fault risk level and a high maintenance priority in the maintenance tasks and execute them in a timely manner.

[0263] Obtain the historical failure data and maintenance records of multiple devices, and calculate the failure risk level and maintenance priority value of each device through big data analysis. According to the failure risk level and maintenance priority value of the device, use the sparrow algorithm to optimize and sort the execution order of the device maintenance tasks, and obtain the optimal sequence of the device maintenance task execution order. According to the optimized execution order of the device maintenance tasks, generate a device maintenance plan, and give priority to including the devices with high failure risk level and high maintenance priority in the maintenance plan. Obtain the real-time operation status data of the device. If the key operation parameters of the device exceed the preset safety threshold range, trigger the early warning maintenance task of the device and set the maintenance task priority of the device to the highest. According to the device maintenance plan, schedule the maintenance personnel to execute the device maintenance tasks in real time. If there is an early warning maintenance task, give priority to scheduling personnel to handle the early warning maintenance task. Collect and analyze the device operation data in real time through the Internet of Things platform, and continuously optimize the device failure prediction model using machine learning algorithms, and dynamically update the failure risk level of the device. According to the latest device failure risk level, regularly re-execute the sparrow algorithm to optimize the execution order of the device maintenance tasks, generate a new device maintenance plan, and form a dynamic optimization closed loop for the device maintenance tasks.

[0264] Specifically, through big data analysis of the failure data and maintenance records of 1000 devices in the past 5 years, the failure risk level of each device is calculated using the association rule mining algorithm, with a value range of 1-5, and the higher the value, the higher the failure risk. At the same time, the analytic hierarchy process is used to comprehensively evaluate the maintenance priority of the devices from 5 dimensions such as device importance, failure frequency, and failure impact degree, and the maintenance priority value of each device is obtained, with a value range of 0-1. On this basis, the sparrow search algorithm is used to optimize the execution order of the device maintenance tasks. By setting the fitness function, with the maximum weighted sum of the device failure risk level and the maintenance priority value as the optimization goal, after 500 iterations of search, the optimal sequence of the device maintenance task execution order is obtained. According to the optimized order, a device maintenance plan is generated, and the devices with a failure risk level above 4 and a maintenance priority value greater than 0.8 are preferentially included in the plan. By installing sensors on the devices to collect key operation parameters such as temperature, vibration, and current of the devices in real time, when it is monitored that the temperature of a certain device exceeds the preset safety threshold of 80°C for 5 consecutive minutes, the system automatically issues a warning maintenance task and sets the maintenance task priority of the device to the highest. The system automatically schedules maintenance personnel to handle the maintenance tasks every 1 hour. When encountering a warning task, it immediately notifies the nearest maintenance personnel to go and handle it. By deploying the Internet of Things platform, the device operation data is collected every 30 seconds, and the long short-term memory neural network algorithm is used to train and optimize the device failure prediction model, and the device failure risk level is updated once a day. When the failure risk level of a certain device rises continuously for 3 days, the system automatically triggers the sparrow algorithm to re-optimize the execution order of the device maintenance tasks, generates the latest device maintenance plan, and forms a closed-loop management for the dynamic optimization of the device maintenance tasks.

[0265] Step S107: Dynamically update the relevant parameters of the failure prediction model according to the execution effect of the device maintenance task, recalculate the device failure prediction result, and judge whether the device operation state has returned to normal.

[0266] Obtain the execution effect data of the device maintenance task, including device operation parameters, maintenance operation records, etc.; dynamically adjust the relevant parameters of the failure prediction model according to the obtained execution effect data, including model weights, thresholds, etc.; use the adjusted failure prediction model to recalculate the failure probability of the device to obtain the updated device failure prediction result; compare the updated device failure prediction result with the preset normal operation threshold to judge whether the device operation state has returned to normal; if the device operation state has returned to normal, it is determined that the device maintenance task is executed effectively, and the current maintenance strategy is maintained; if the device operation state has not returned to normal, it is determined that the device maintenance task is executed ineffectively, and the maintenance strategy needs to be further optimized; dynamically adjust the execution frequency and maintenance content of the device maintenance task according to the judgment result of whether the device operation state has returned to normal to form a closed-loop optimization.

[0267] Specifically, when obtaining the data on the execution effect of the equipment maintenance task, the operating parameters of the equipment, such as temperature, pressure, vibration, etc., can be collected in real time through the sensors installed on the equipment. The data is collected every 5 minutes and uploaded to the database. At the same time, after the maintenance personnel complete the equipment maintenance operation, they need to enter information such as the content, time, and personnel of the maintenance operation into the system to form a complete maintenance operation record. According to the obtained equipment operating parameters and maintenance operation record data, machine learning algorithms such as support vector machine (SVM) are used to dynamically adjust the relevant parameters of the fault prediction model. Through methods such as grid search and cross-validation, the hyperparameters of the model, such as the penalty coefficient C and the type of kernel function, are optimized to enable the model to better fit the data and improve the accuracy of fault prediction. The adjusted fault prediction model can recalculate the fault probability of the equipment through the logistic regression algorithm to obtain the updated equipment fault prediction result. The updated equipment fault prediction result is compared with the preset normal operation threshold (such as the fault probability is less than 1). If the fault probability of the equipment is lower than the threshold, it is determined that the equipment operating state has returned to normal and the maintenance task execution is effective, and the current maintenance strategy can be maintained. If the fault probability of the equipment is higher than the threshold, it is determined that the equipment operating state has not returned to normal and the maintenance task execution is ineffective, and the maintenance strategy needs to be further optimized, such as shortening the maintenance cycle, increasing the maintenance content, etc. According to the judgment result of whether the equipment operating state has returned to normal, the execution frequency and maintenance content of the equipment maintenance task are dynamically adjusted. For example, the original monthly regular maintenance is changed to semi-monthly, and the original routine maintenance content is increased by replacing equipment parts. Through continuous feedback optimization, a closed-loop management of equipment maintenance is formed to continuously improve the reliability and stability of the equipment.

[0268] Step S108: If the equipment operating state has returned to normal, the key parameters such as temperature, voltage, and current are all within the safety threshold range, and there are no repeated fault signs in the subsequent period of time, the maintenance task of the equipment can be ended. Otherwise, the maintenance plan needs to be adjusted and the maintenance measures need to be continuously optimized until the equipment is completely restored to normal.

[0269] Obtain the real-time operating status data of the device, including key parameters such as temperature, voltage, and current. According to the preset safety threshold range, determine whether the device operating status has returned to normal. If the device operating status has returned to normal, continuously monitor the device operating data, and through big data analysis technology, determine whether there are any signs of failure in the device in the subsequent period of time. If there are no signs of failure in the device in the subsequent period of time, it can be determined that the device has fully recovered to normal, generate an end instruction for the device maintenance task, and terminate the device maintenance task. If the device operating status has not returned to normal, or there are signs of failure in the subsequent period of time, trigger the adjustment mechanism of the maintenance plan, and through machine learning algorithms, automatically optimize the device maintenance measures. During the adjustment of the maintenance plan, use decision tree algorithms to comprehensively consider factors such as the device's historical operating data, current operating status, and maintenance costs to obtain the optimal combination of maintenance measures. Send the optimized maintenance measures to the device, and through Internet of Things technology, remotely control the device to perform the corresponding maintenance operations and continuously monitor the changes in the device operating status. According to the changes in the device operating status, loop through the above steps until the device fully recovers to normal and meets the conditions for ending the maintenance task, and output the completion signal of the device maintenance task.

[0270] Specifically, by installing sensors on the device, key parameters such as the temperature, voltage, and current of the device are collected in real time. For example, the temperature sensor collects the device temperature as 65°C, the voltage sensor collects the operating voltage of the device as 220V, and the current sensor collects the operating current of the device as 2A. The collected data is compared with the preset safety threshold range. The temperature threshold is set to 0°C to 100°C, the voltage threshold is set to 200V to 240V, and the current threshold is set to 0A to 5A. By comparison, it can be judged whether the current device operation status is normal. If the device operation status returns to normal, the device operation data is continuously monitored. Through big data analysis technology, the historical operation data of the device is analyzed, and the time series analysis algorithm is used to predict the operation trend of the device in the next period of time to judge whether there may be fault signs. If the prediction result shows that there is no fault risk for the device within the next 1 month, it can be determined that the device has fully recovered to normal, and an end instruction for the device maintenance task is generated to terminate the device maintenance task. If the device operation status does not return to normal, or the prediction result shows that there is a high fault risk for the device within the next 1 month, a maintenance plan adjustment mechanism is triggered. Through the reinforcement learning technology in the machine learning algorithm, combined with the historical operation and maintenance data of the device, the device maintenance measures are automatically learned and optimized to obtain the best combination of maintenance measures. For example, through the analysis of historical data, it is found that when the device temperature exceeds 80°C, the device needs to be cooled down, and when the device voltage is lower than 210V, the device needs to be power-supplied and checked. Using the decision tree algorithm, considering factors such as the historical operation data, current operation status, and maintenance cost of the device, a decision tree for maintenance measures is generated. According to the branch conditions of the decision tree, the optimal combination of maintenance measures in different situations is obtained and sent to the device. Through the Internet of Things technology, the device is remotely controlled to perform corresponding maintenance operations, such as starting the device's heat dissipation system, adjusting the device's power supply voltage, etc., and the change of the device operation status is monitored in real time. According to the change of the device operation status, the above steps are cyclically executed to continuously optimize the maintenance plan until the key parameters such as the temperature, voltage, and current of the device all return to the normal range, meeting the conditions for ending the maintenance task, and outputting a completion signal for the device maintenance task.

[0271] Step S109: Increase the status monitoring and fault warning of key components of the sanitation robot.

[0272] To further improve the reliability of the sanitation robot cluster and the cluster operation efficiency, the present invention adds a status monitoring and fault warning mechanism for key components of the sanitation robot, specifically including:

[0273] Battery: By real-time monitoring parameters such as battery power and charge-discharge cycle times, an algorithm is used to evaluate the battery health status, and risks such as low battery power and excessive cycle times are timely warned, and the user is notified in advance to replace or charge.

[0274] Sweeping disk: Monitor indicators such as the rotation speed, wear degree, and cleaning efficiency of the sweeping disk. If the cleaning efficiency decreases or the wear is excessive, the system will issue a fault warning in advance and recommend that the user clean or replace it.

[0275] Fan: By monitoring parameters such as the fan speed, noise, and vibration, detect early signs of fan failure in a timely manner, such as abnormal speed, excessive vibration, etc., and perform maintenance or replacement in advance.

[0276] Suction nozzle: Monitor the suction force and blockage degree of the suction nozzle. If the suction force decreases or the blockage is severe, the system will give an alarm in advance and recommend that the user clean or replace it.

[0277] Filter: Monitor the status of the filter, including filtration efficiency, blockage degree, etc. If the filtration efficiency decreases or the blockage is severe, the system will issue a warning and recommend cleaning or replacement.

[0278] LiDAR: Real-time monitor the working status of the LiDAR, including laser emission power, received signal strength, etc. Once signal abnormalities or power drops are detected, the system will issue a warning.

[0279] Camera: Monitor the image quality, focal length, etc. of the camera. If the image quality deteriorates or the focal length is distorted, the system will inform the user in advance to clean or repair it.

[0280] Ultrasonic: Monitor the signal quality, response time, etc. of the ultrasonic sensor. If the signal quality decreases or the response time is abnormal, the system will issue a fault warning in advance and recommend checking or replacing the sensor.

[0281] By adding a status monitoring and warning mechanism for the above key components, the system can identify potential faults in advance, avoid the interruption of cluster tasks caused by component failures of the equipment, and thus improve the overall operation efficiency and reliability of the sanitation robot cluster.

[0282] To further improve the reliability of the sanitation robot cluster and the cluster operation efficiency, the present invention adds a status monitoring and fault warning mechanism for the key components of the sanitation robot, specifically including:

[0283] Battery: By real-time monitoring parameters such as battery power, charge and discharge cycle times, etc., comprehensively evaluate the battery health degree using the following formula:

[0284] (5)

[0285] In formula (5), is the battery health degree, is the number of battery charge and discharge cycles, is the depth of charge, is the maximum battery capacity. Issue a warning according to the battery health degree. When When it is lower than a certain threshold, the system will remind the user to replace or recharge.

[0286] Sweeping disc: By monitoring the change in the length of the sweeping bristles and the change in the ground contact pressure during sweeping, the following formula is used to calculate the wear degree of the sweeping disc:

[0287] (6)

[0288] In formula (6), is the wear degree of the sweeping disc, is the current value of the length of the sweeping bristles, is the initial length value of the sweeping bristles, is the current ground contact pressure value, is the maximum allowable ground contact pressure value. When reaches the preset threshold, the system will issue a fault warning to remind the user to clean or replace.

[0289] Fan: Monitor the flow rate and negative pressure of the fan, and calculate the fan performance through the following formula:

[0290] (7)

[0291] In formula (7), is the fan power, is the fan flow rate, is the negative pressure of the fan, is the rotational speed of the fan, is the fan efficiency. When exceeds the normal range, the system will give an alarm to remind the user to check and maintain the fan.

[0292] Suction nozzle: By monitoring the wear degree and suction force of the suction nozzle, judge the performance of the suction nozzle through the following formula:

[0293] (8)

[0294] In formula (8), is the suction force attenuation degree of the suction nozzle, is the initial suction force, is the current suction force. When reaches a certain value, the system will issue a warning that the suction nozzle needs maintenance.

[0295] Filter: By monitoring the usage time and water flow rate of the filter, evaluate the effectiveness of the filter using the following formula:

[0296] (9)

[0297] In formula (9), is the filter efficiency, is the usage time, is the maximum usage time, is the current water flow rate, is the maximum water flow rate. If is less than the preset threshold, the system will remind to clean or replace the filter.

[0298] LiDAR: Monitor the monitoring accuracy of the LiDAR. The following formula is used to evaluate the radar performance:

[0299] (10)

[0300] In formula (10), is the radar accuracy, is the maximum detection distance of the radar, is the minimum detection distance. When exceeds the allowable range, the system will trigger a fault warning.

[0301] Camera: Monitor the resolution of the camera. The following formula is used to calculate the camera image quality:

[0302] (11)

[0303] In formula (11), is the image quality of the camera, and are the actual resolution height and width of the camera respectively, and are the maximum resolution height and width respectively. When is lower than the threshold, the system will issue a camera fault warning.

[0304] Ultrasonic: Monitor the sensing accuracy of the ultrasonic sensor. The following formula is used to evaluate the sensing effect:

[0305] (12)

[0306] In formula (12), is the sensing accuracy of the ultrasonic, and are the maximum and minimum detection distances respectively. If is lower than a certain threshold, the system will issue a warning of reduced sensing accuracy.

[0307] By adding the status monitoring and warning mechanism for the above key components, the system can identify potential faults in advance, avoid the interruption of cluster tasks caused by component failures, and thus improve the overall operation efficiency and reliability of the sanitation robot cluster.

[0308] S110. Fault warning mechanism by comprehensively analyzing the state parameters of each key component of the sanitation robot and combining sensor data.

[0309] In order to more accurately evaluate the operating status of the sanitation robot cluster, the present invention proposes to comprehensively analyze the state parameters of each key component and sensor data to achieve more accurate fault warning. The specific steps are as follows:

[0310] 1. Comprehensive parameter analysis:

[0311] Integrate parameters such as battery health, sweeper wear degree, fan flow rate, suction nozzle wear, filter status, lidar monitoring accuracy, camera resolution, ultrasonic induction accuracy, etc. with data such as temperature, voltage, current, vibration, etc. collected by the sanitation robot sensors to establish a comprehensive evaluation model. Through the weighted average method, each parameter is comprehensively scored according to its weight affecting the equipment operation.

[0312] 2. Comprehensive evaluation model:

[0313] Assume that the influence of the state parameters of each component is: , , , , , , , , corresponding to the weight coefficients of the battery, fan, sweeper, suction nozzle, filter, lidar, camera, and ultrasonic respectively. The comprehensive health score can be expressed as:

[0314] (13)

[0315] In formula (13), is the battery health, is the sweeper wear degree, is the fan power, is the suction nozzle suction attenuation degree, is the filter efficiency, is the lidar accuracy, is the camera image quality, is the ultrasonic induction accuracy.

[0316] After combining the sensor data, this comprehensive score is judged through the fault warning model. If is lower than the preset safety threshold, the system will issue an alarm and recommend maintenance.

[0317] 3. Fault warning mechanism combined with sensor data:

[0318] During the data collection process, the sanitation robot monitors the device status in real time through sensors such as temperature, voltage, current, and vibration. The time series analysis model (such as ARIMA or LSTM) is used to model the historical data to predict the future health status of the device.

[0319] The sensor data (such as temperature, voltage, current, etc.) is combined with the above comprehensive score to form a composite scoring model. The model is continuously updated and optimized through machine learning methods (such as support vector machine SVM, decision tree, etc.) to make the fault warning more accurate.

[0320] When the composite scoring model evaluates that a certain sanitation robot has a high fault risk, the system will generate corresponding maintenance strategies according to the specific fault type and risk level, and give priority to arranging high-risk devices for inspection and maintenance.

[0321] 4. Comprehensive analysis and real-time monitoring:

[0322] During the operation of the sanitation robot cluster, the system calculates the comprehensive score of each device automatically by monitoring the status of each device in real time, including battery power, sweeper wear, fan flow rate, etc., and combining the sensor data. When the comprehensive score of the device is lower than the preset threshold for multiple consecutive times, the system will send a warning signal and optimize the maintenance plan for the device.

[0323] For example, when the flow rate and negative pressure parameters of the fan are abnormal, combined with the battery health and camera resolution data, the system can judge the overall operation of the device and take timely measures to avoid the failure of the overall task caused by the failure of a single component.

[0324] 5. Adjustment and optimization:

[0325] The system continuously collects and analyzes the real-time data of the device. Combining the above comprehensive score, the sparrow algorithm is used to optimize the execution order of the maintenance tasks of each device. For devices with lower scores and higher fault risks, priority is given to maintenance to reduce the risk of cluster operation interruption.

[0326] Through the comprehensive analysis of the state parameters of the above key components and sensor data, the present invention can more accurately evaluate the fault risk of the device, give early warnings and take effective maintenance measures, thereby improving the operation efficiency of the sanitation robot cluster and the long-term reliability of the device.

[0327] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made to these embodiments by those skilled in the art once they learn the basic inventive concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications that fall within the scope of the present invention. Obviously, those skilled in the art can make various changes and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A method for monitoring the status and early warning of sanitation robot clusters, characterized in that: The following steps are involved: Obtaining operating status data of the sanitation robot cluster; the operating status data includes temperature, voltage and current data; Compare the acquired operating status data with a preset normal operating status threshold, and if the operating status data exceeds the normal operating status threshold, dynamically adjust the observation frequency of the operating status data to narrow the observation range; Obtain the cumulative running time of the sanitation robot cluster, and if the cumulative running time exceeds a preset safety time threshold, adjust the time weight parameter in the fault prediction model; According to the adjusted duration weight parameter, the fault prediction result is recalculated, and the accumulated operation duration and operation status data are comprehensively considered to determine whether each sanitation robot equipment has a fatigue loss risk. If there is a fatigue loss risk, an equipment failure warning signal is generated; Obtain the equipment failure warning signal, combine the operation status data and the accumulated operation time, generate a targeted equipment maintenance suggestion plan, and determine the maintenance priority of each sanitation robot equipment according to the failure risk level; Optimize the execution order of maintenance tasks for each sanitation robot device, and give priority to the equipment with high failure risk level and high maintenance priority among the sanitation robot devices and execute them in time; According to the execution effect of the equipment maintenance task, the relevant parameters of the fault prediction model are dynamically updated, the equipment fault prediction results are recalculated, and it is determined whether the equipment operation status has returned to normal; If it is identified that the equipment operating status has returned to normal, and the operating status data are within the normal operating status threshold range within the set time, the equipment maintenance task is terminated, otherwise it is necessary to adjust the maintenance plan and continue to optimize the maintenance measures until the sanitation robot equipment is completely restored to normal; The step of obtaining the running status data of the sanitation robot cluster includes: According to the operating status of the sanitation robot cluster, obtain real-time data of the operating status data and build a data collection model; The sparrow algorithm is used to optimize the data collection strategy, and the optimal observation frequency and observation range of each parameter in the operating status data are determined through iterative search; During the optimization process, the data acquisition model is dynamically adjusted according to changes in the observation frequency and the observation range; Preprocessing the operating status data to remove abnormal values ​​and noise data in the operating status data; Use clustering algorithms to analyze the preprocessed operating status data and identify data features and patterns under different operating states; According to the clustering results, a mapping relationship between the operating status and data features is established to achieve real-time monitoring and early warning of the operating status of the sanitation robot cluster; The optimized data collection strategy is applied to the sanitation robot cluster, and the data collection efficiency and quality are continuously monitored. The optimization algorithm is dynamically adjusted according to the feedback results to achieve adaptive optimization of the data collection strategy.

2. The sanitation robot cluster status monitoring and fault warning method according to claim 1, characterized in that: The step of comparing the acquired operating status data with a preset normal operating status threshold, and if the operating status data exceeds the normal operating status threshold, dynamically adjusting the observation frequency of the operating status data to narrow the observation range includes: Transmitting the operating status data to a data processing module; The data processing module uses a time series analysis algorithm to perform trend analysis on the historical data of each parameter in the operating status data to obtain a change trend curve of each parameter; According to the preset normal threshold range of each parameter, determine whether the real-time data of each parameter in the operating status data exceeds the normal threshold range, and if the real-time data of each parameter exceeds the normal threshold range, trigger an alarm; If it is identified that the real-time data of a certain parameter exceeds the normal threshold range for multiple times in a row, the observation frequency of the current parameter is dynamically adjusted; At the same time, according to the change trend curve, the observation range of the current parameter is narrowed; The Kalman filter algorithm is used to filter the real-time data of the current parameters to remove noise interference in the operating status data; The adjusted observation frequency, observation range, and real-time data after filtering are transmitted to the data visualization module, and the changes of the current parameters are displayed in real time via the data visualization module.

3. The sanitation robot cluster status monitoring and fault warning method according to claim 1, characterized in that: The step of obtaining the cumulative running time of the sanitation robot cluster and adjusting the time weight parameter in the fault prediction model if the cumulative running time exceeds a preset safety time threshold comprises: Acquire the real-time operation data of the sanitation robot cluster and transmit the real-time operation data to the control center; the real-time operation data includes the operation status and operation time information of each sanitation robot; The sparrow algorithm is used to analyze and calculate the cumulative running time of the sanitation robot cluster, and the total running time and average running time of the cluster are obtained; Determine whether the total operation time of the cluster exceeds the preset safety time threshold. If the total operation time of the cluster exceeds the safety time threshold, the adjustment mechanism of the fault prediction model is triggered; According to the degree of exceeding the safety time threshold, the time weight parameter in the fault prediction model is dynamically adjusted to give the time factor a higher weight; By increasing the duration weight, the fault prediction model is more biased towards the duration factor; Based on the adjusted fault prediction model, the failure risk of the sanitation robot cluster is re-evaluated, and high-risk sanitation robot individuals are identified; For high-risk sanitation robots, return instructions are issued in a timely manner, and spare sanitation robots are dispatched to take over the work tasks of the high-risk sanitation robots.

4. The sanitation robot cluster status monitoring and fault warning method according to claim 1, characterized in that: The step of recalculating the fault prediction result according to the adjusted duration weight parameter, comprehensively considering the accumulated operation duration and operation status data, and judging whether each sanitation robot equipment has a fatigue loss risk, and if there is a fatigue loss risk, generating an equipment fault warning signal includes: Obtaining an adjusted duration weight parameter, and inputting the adjusted duration weight parameter into a fault prediction model; According to the fault prediction model, a fault prediction result of the sanitation robot equipment is calculated; Obtain operating time data and status monitoring data of sanitation robot equipment; Input the operating time data and condition monitoring data of the sanitation robot equipment into the fatigue loss risk assessment model; The fatigue loss risk assessment model comprehensively analyzes the operating time data and the status monitoring data to determine whether the sanitation robot equipment has a fatigue loss risk; If it is identified that the sanitation robot equipment has a risk of fatigue wear, a warning signal for equipment failure is generated; The fault warning signal is sent to the sanitation robot equipment management system.

5. The sanitation robot cluster status monitoring and fault warning method according to claim 1, characterized in that: The steps of obtaining the equipment failure warning signal, combining the operation status data and the accumulated operation time, generating a targeted equipment maintenance suggestion plan, and determining the maintenance priority of each sanitation robot equipment according to the failure risk level include: Performing data cleaning and feature engineering on the running status data and the accumulated running time, removing outliers and redundant data, and extracting key features; Based on the historical operation data of the equipment, a fault warning model is established using a support vector machine algorithm, and the fault warning model is trained and optimized; When the equipment failure warning signal is received, a targeted equipment maintenance suggestion plan is generated through a decision tree algorithm in combination with the current status data and operation time data of the equipment; Use association rule mining algorithm to analyze the relationship between equipment failure type, failure cause and maintenance measures; According to the severity of the equipment fault warning signal and the possibility of the fault occurring, the fault risk level of each sanitation robot equipment is calculated by a fuzzy comprehensive evaluation method; Based on the failure risk level of each sanitation robot equipment and the limitation of maintenance resources, an integer programming model is used to determine the maintenance priority information of each sanitation robot equipment; The equipment maintenance recommendation plan and the maintenance priority information are displayed through a visual dashboard.

6. The sanitation robot cluster status monitoring and fault warning method according to claim 1, characterized in that: The step of optimizing the execution order of maintenance tasks of each sanitation robot device, and prioritizing the devices with high fault risk level and high maintenance priority among the sanitation robot devices to be included in the maintenance tasks and executed in a timely manner comprises: Obtain historical fault data and maintenance records of multiple sanitation robot devices, and calculate the fault risk level and maintenance priority value of each sanitation robot device through big data analysis; According to the fault risk level and maintenance priority value, the sparrow algorithm is used to optimize the execution order of the maintenance tasks of the sanitation robot equipment to obtain the optimal equipment maintenance task execution sequence; Generate an equipment maintenance plan based on the optimized equipment maintenance task execution sequence, and prioritize sanitation robot equipment with high failure risk level and high maintenance priority into the maintenance plan; Acquire the real-time operating status data of the sanitation robot equipment. If it is identified that the key operating parameters of the sanitation robot equipment exceed the preset safety threshold range, the early warning maintenance task of the equipment is triggered, and the maintenance task priority of the current equipment is set to the highest; According to the equipment maintenance plan, maintenance personnel are dispatched in real time to perform equipment maintenance tasks; If a warning maintenance task is identified, personnel will be dispatched to handle the warning maintenance task first; Collect and analyze equipment operation data in real time through the IoT platform, use machine learning algorithms to continuously optimize the fault prediction model, and dynamically update the equipment's fault risk level; According to the latest equipment failure risk level, the sparrow algorithm is re-executed regularly to optimize the execution sequence of equipment maintenance tasks, generate a new equipment maintenance plan, and form a dynamic optimization closed loop of equipment maintenance tasks.

7. The sanitation robot cluster status monitoring and fault warning method according to claim 1, characterized in that: If it is identified that the equipment operation status has returned to normal, and the operation status data are within the normal operation status threshold range within the set time, the equipment maintenance task is terminated, otherwise the maintenance plan needs to be adjusted and the maintenance measures continue to be optimized until the sanitation robot equipment is completely restored to normal. The steps include: Obtaining execution effect data of equipment maintenance tasks; the execution effect data includes equipment operating parameters and maintenance operation records; Dynamically adjusting relevant parameters of a fault prediction model according to the acquired execution effect data; the relevant parameters of the fault prediction model include model weights and thresholds; The adjusted fault prediction model is used to recalculate the failure probability of the sanitation robot equipment and obtain updated equipment failure prediction results; Compare the updated equipment fault prediction results with the preset normal operation threshold to determine whether the equipment operation status has returned to normal; If the equipment operation status returns to normal, it is determined that the equipment maintenance task is executed effectively and the current maintenance strategy is maintained; If the equipment operation status does not return to normal, it is determined that the equipment maintenance task is invalid and the maintenance strategy needs to be further optimized; Based on the judgment result of whether the equipment operating status has returned to normal, the execution frequency and maintenance content of the equipment maintenance tasks are dynamically adjusted to form a closed-loop optimization.

8. The sanitation robot cluster status monitoring and fault warning method according to claim 1, characterized in that: The steps of dynamically updating relevant parameters of the fault prediction model according to the execution effect of the equipment maintenance task, recalculating the equipment fault prediction result, and judging whether the equipment operation status has returned to normal include: Obtain real-time operating status data of sanitation robot equipment and determine whether the equipment operating status has returned to normal based on the preset safety threshold range; If it is identified that the equipment operation status has returned to normal, the equipment operation data will be continuously monitored, and through big data analysis technology, it will be determined whether the sanitation robot equipment has any signs of failure within the subsequent set time; If it is identified that the sanitation robot equipment has no signs of failure within the subsequent set time, it is determined that the sanitation robot equipment has completely recovered to normal, and an end instruction of the equipment maintenance task is generated to terminate the equipment maintenance task; If it is identified that the equipment's operating status has not returned to normal, or if there are signs of failure within a period of time, the maintenance plan adjustment mechanism will be triggered, and the equipment's maintenance measures will be automatically optimized through machine learning algorithms; In the process of adjusting the maintenance plan, the decision tree algorithm is used to comprehensively consider the historical operation data and current operation status of the equipment as well as the maintenance cost factors to obtain the optimal combination of maintenance measures; The optimized maintenance measures are sent to the sanitation robot equipment. Through the Internet of Things technology, the equipment is remotely controlled to perform corresponding maintenance operations and the operating status changes of the equipment are monitored in real time; According to the changes in the equipment operating status, the above steps are executed repeatedly until the equipment is completely restored to normal and the conditions for ending the maintenance task are met, and a completion signal of the equipment maintenance task is output.

9. A sanitation robot cluster state monitoring and fault warning system, applied to the sanitation robot cluster state monitoring and fault warning method according to any one of claims 1 to 8, characterized in that: The sanitation robot cluster status monitoring and fault warning system includes: An acquisition module is used to acquire the operating status data of the sanitation robot cluster; the operating status data includes temperature, voltage and current data; A comparison module, used to compare the acquired operation status data with a preset normal operation status threshold, and if the operation status data exceeds the normal operation status threshold, dynamically adjust the observation frequency of the operation status data to narrow the observation range; An adjustment module is used to obtain the cumulative running time of the sanitation robot cluster, and if the cumulative running time exceeds a preset safety time threshold, adjust the time weight parameter in the fault prediction model; The first judgment module is used to recalculate the fault prediction result according to the adjusted duration weight parameter, comprehensively consider the accumulated operation time and operation status data, and judge whether each sanitation robot equipment has a fatigue loss risk. If there is a fatigue loss risk, an equipment failure warning signal is generated; A determination module is used to obtain the equipment failure warning signal, generate a targeted equipment maintenance suggestion plan in combination with the operation status data and the accumulated operation time, and determine the maintenance priority of each sanitation robot equipment according to the failure risk level; An optimization module is used to optimize the execution order of maintenance tasks of each sanitation robot device, and give priority to the equipment with high fault risk level and high maintenance priority among the sanitation robot devices and execute them in time; The second judgment module is used to dynamically update the relevant parameters of the fault prediction model according to the execution effect of the equipment maintenance task, recalculate the equipment fault prediction result, and judge whether the equipment operation status has returned to normal; The maintenance module is used to end the equipment maintenance task if it is identified that the equipment operating status has returned to normal and the operating status data are within the normal operating status threshold within the set time. Otherwise, the maintenance plan needs to be adjusted and the maintenance measures continue to be optimized until the sanitation robot equipment is fully restored to normal.

Images