Scaffold robot erection failure prediction system

Abstract

The application relates to the field of scaffold robots, in particular to a building fault prediction system applied to a scaffold robot, which comprises an element multi-source perception module, a time sequence management prediction module, a risk discrimination module, a risk active correction module and an output warning module.In the application, wind speed, vibration and joint load are synchronously collected and feature extraction is carried out, analysis is carried out through a wind speed-vibration combined discrimination matrix, comprehensive discrimination and weighted fusion of high-altitude falling correlation risks and structural instability risks are realized, the composite risks of gust impact and structural resonance are effectively identified, meanwhile, a single scaffold node can be taken as a benchmark, time consumption during building is recorded and a time sequence feature vector is extracted, an installation overtime risk is quantitatively evaluated, and the robot running speed is automatically adjusted, construction overtime faults are actively avoided within a limited operation window period, and the time controllability of a building task is improved.

Description

Technical Field

[0001] This invention relates to the field of scaffolding robots, specifically a fault prediction system for scaffolding robots. Background Technology

[0002] Scaffolding erection is a high-risk part of construction. Traditionally, it relies on manual operation, which has inherent drawbacks such as high labor intensity, high risk of falls, and low work efficiency. In recent years, with the integration of robotics technology and intelligent construction, specialized robots for high-altitude scaffolding erection have been gradually applied. These robots use multi-degree-of-freedom robotic arms to perform repetitive erection operations such as grasping, positioning, and connecting, thus replacing manual high-altitude work to a certain extent.

[0003] However, unlike ground industrial robots, scaffolding robots operate in a highly open environment. During operation, they are simultaneously subjected to the coupling effect of external environmental loads and internal mechanical states. Furthermore, the construction progress is constrained by weather windows, and the predetermined installation tasks must be completed within a limited time.

[0004] Currently, most safety monitoring solutions for scaffolding robots draw on the single-parameter alarm mode of traditional tower cranes or elevators. For example, they only alarm for exceeding the threshold for wind speed or only protect the current of the drive motor from overload. The monitoring of each physical quantity is independent of each other, and there is a lack of comprehensive judgment ability on the inherent relationship between wind speed, vibration response and joint load. Meanwhile, conventional systems also fail to conduct joint assessments and collaborative processing of external environmental risks and internal construction progress risks under a unified framework, and are unable to proactively adjust the robot's operating strategy based on progress deviations to avoid construction timeout risks, thus hindering further improvements in the safety and construction efficiency of scaffolding robots. Summary of the Invention

[0005] In this invention, wind speed, vibration, and joint load are simultaneously collected and feature extracted. Through analysis using a wind speed-vibration joint discrimination matrix, a comprehensive judgment and weighted fusion of the risks associated with high-altitude falls and structural instability is achieved. This effectively identifies the combined risks of gust impact and structural resonance. Simultaneously, using a single scaffold node as a benchmark, the system records the erection time and extracts temporal feature vectors to quantitatively assess the risk of installation timeouts and automatically adjusts the robot's operating speed. Within a limited working window, it proactively avoids construction timeout failures, improving the time controllability of the erection task. This addresses the problem that existing technologies rely on independent monitoring methods, which cannot comprehensively assess the safety and construction efficiency of scaffolding robots. Therefore, this invention proposes an erection failure prediction system for scaffolding robots.

[0006] The objective of this invention can be achieved through the following technical solution: a fault prediction system for scaffolding robots, comprising a multi-source element perception module, a time-series management prediction module, a risk discrimination module, a risk active correction module, and an output early warning module; The multi-source sensing module can classify the operation process of the scaffolding robot into elements and collect multi-source sensing data to obtain physical multi-element elements, including wind speed elements, vibration elements, and joint load elements. The timing management and prediction module can record the scaffolding robot's construction process and obtain timing elements, including construction location, construction time, and construction completion time. The risk discrimination module can perform physical risk analysis based on physical multi-source factors and time-series risk analysis based on time-series factors, and generate physical risk types, physical risk probabilities, and time-series risk types and time-series risk probabilities respectively. The risk active correction module actively analyzes the time-series risk type and probability, generates a correction signal, corrects the time-series risk type and probability, and sends the corrected time-series risk probability and type to the output early warning module. The output early warning module analyzes physical risk types and probabilities and generates physical risk warnings. Simultaneously, it generates risk warnings based on the time-series risk probabilities and types sent by the proactive risk correction module. In a preferred embodiment of the present invention, the multi-source sensing module collects the wind speed time-series signal of the work area in real time at a fixed sampling frequency, and performs sliding window filtering and outlier removal on the wind speed time-series signal to generate a standard wind speed feature vector. The multi-source sensing module collects the time-domain signal of the robot's three-dimensional vibration acceleration, extracts the frequency-domain amplitude features through fast Fourier transform, and concatenates the time-domain and frequency-domain features into a comprehensive vibration feature vector. The multi-source sensing module collects the output torque value of each joint in real time, and calculates it by combining the reduction ratio and transmission efficiency of each joint to obtain the actual load torque vector of each joint.

[0007] In a preferred embodiment of the present invention, the risk discrimination module calculates the standard wind speed feature vector to obtain the wind speed risk judgment quantity, performs threshold judgment on the wind speed risk judgment quantity to determine whether there is a risk of gust impact, and generates a fall-related risk type and assigns a first physical risk probability. The vibration comprehensive feature vector is analyzed to obtain the vibration risk assessment quantity. The vibration risk assessment quantity is then analyzed to determine whether there is a type of structural instability risk and assign a second physical risk probability. The actual load torque vector is thresholded to determine whether there is a risk of joint overload and to assign a third physical risk probability.

[0008] In a preferred embodiment of the present invention, the risk discrimination module further performs composite risk identification by constructing a joint discrimination matrix of wind speed and vibration: Using the current wind speed risk assessment and vibration risk assessment as two-dimensional inputs, they are mapped to a preset wind speed and vibration joint discrimination matrix. The joint discrimination matrix is ​​divided into four quadrants: safe zone, attention zone, warning zone, and danger zone. When the mapping point falls into the warning zone or danger zone, the risk discrimination module will weight and fuse the first physical risk probability and the second physical risk probability to generate a comprehensive physical risk probability.

[0009] In a preferred embodiment of the present invention, the time-series management prediction module uses the scaffolding robot to complete the construction of a single scaffolding node as a positioning reference, and records in real time the number of nodes that have been completed, the number of nodes that have been built, the actual start time of the current node, and the time elapsed for the current node, and stores the above information in the construction log sequence according to the timestamp. The time series management prediction module extracts the historical construction time of each node based on the construction log sequence, calculates the historical average construction time and the historical standard deviation, and calculates the number of remaining nodes to be built based on the current construction node number and the total number of planned nodes. The historical average construction time, the historical standard deviation, the number of remaining nodes to be built, and the current time elapsed are combined into a time series feature vector.

[0010] In a preferred embodiment of the present invention, the time management prediction module uses the ratio of the planned total construction time to the historical average time as the benchmark progress rate, and combines the remaining number of nodes to be built and the current time already consumed to predict the actual remaining completion time; the difference between the actual remaining completion time and the planned remaining time is calculated to obtain the progress deviation. The timing management prediction module performs a threshold judgment on the progress deviation to determine whether there is an installation timeout risk, generates a timing risk type as progress lag risk, and calculates the timing risk probability by multiplying the ratio of the progress deviation to the positive deviation threshold by the basic timeout probability.

[0011] In a preferred embodiment of the present invention, after receiving the schedule lag risk and the corresponding timing risk probability, the risk active correction module calculates the required speed compensation coefficient based on the schedule deviation, and constrains the speed compensation coefficient between the preset maximum safe speed multiple and the minimum safe speed multiple, and generates a speed adjustment command to send to the controller of the scaffolding robot. The proactive risk correction module re-estimates the corrected actual remaining completion time and the corrected schedule deviation based on the update rate after speed compensation, and recalculates the corrected timing risk probability based on the corrected schedule deviation.

[0012] Compared with the prior art, the beneficial effects of the present invention are: In this invention, wind speed, vibration, and joint load are collected and features are extracted simultaneously. The wind speed-vibration joint discrimination matrix is ​​used for analysis to achieve a comprehensive judgment and weighted fusion of the risk associated with high-altitude falls and the risk of structural instability. This effectively identifies the combined risk of gust impact and structural resonance, and significantly improves the accuracy and reliability of physical fault prediction.

[0013] In this invention, the erection time is recorded and the temporal feature vector is extracted based on a single scaffold node. The progress deviation is calculated based on the planned total construction time and the historical average time, and the installation timeout risk is quantitatively assessed. At the same time, a speed compensation coefficient is generated based on the progress deviation and the robot's running speed is automatically adjusted, forming a closed-loop temporal risk management mechanism from prediction to correction and reassessment. This mechanism proactively avoids construction timeout failures within a limited working window and improves the time controllability of the erection task. Attached Figure Description

[0014] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0015] Figure 1 This is a system block diagram of the present invention; Figure 2 This is a flowchart of the physical risk analysis for this invention. Detailed Implementation

[0016] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0017] Example 1: Please refer to Figures 1-2 As shown, the fault prediction system for scaffolding robots includes a multi-source element perception module, a time-series management prediction module, a risk identification module, a risk active correction module, and an output early warning module. The multi-source sensing module can classify elements and collect multi-source sensing data during the operation of the scaffolding robot to obtain physical multi-source elements, including wind speed elements, vibration elements, and joint load elements. The wind speed acquisition component of the multi-source sensing module is mounted on the top of the scaffolding robot. It collects the wind speed time-series signal of the work area in real time at a fixed sampling frequency, and performs sliding window filtering and outlier removal on the wind speed time-series signal to generate a standard wind speed feature vector. The vibration acquisition component of the multi-source sensing module includes multiple triaxial accelerometers installed on the robot base, the end of the robotic arm, and the scaffold connection node, which synchronously acquire the triaxial vibration acceleration time-domain signals of each measuring point, and extract the frequency domain amplitude features through fast Fourier transform, and splice the time domain and frequency domain features into a comprehensive vibration feature vector. The multi-source sensing module collects the output torque value of each joint in real time by using torque sensors installed on the output shaft of the drive motor of each joint. It then combines the reduction ratio and transmission efficiency of each joint to calculate the actual load torque vector of each joint.

[0018] The risk assessment module can perform physical risk analysis based on multiple physical factors, and generate physical risk types and probabilities using the following methods: The risk assessment module calculates the root mean square value and peak factor of the standard wind speed feature vector. When the root mean square value of the wind speed exceeds the first wind speed threshold and the peak factor is lower than the second wind speed threshold, it determines that there is a risk of gust impact, generates a fall-related risk type and assigns the first physical risk probability. The risk assessment module performs time-domain peak monitoring and frequency-domain dominant frequency offset analysis on the vibration comprehensive feature vector. When the vibration acceleration peak value at any measuring point exceeds the acceleration threshold or the dominant frequency offset exceeds the frequency offset threshold, it determines that there is a risk of structural resonance or loose connection, generates a type of structural instability risk and assigns a second physical risk probability. The risk assessment module compares the actual load torque vector with the rated load limit of each joint. When the duration of the actual load torque of any joint exceeding the rated load limit exceeds the duration tolerance threshold, it determines that there is a risk of joint overload, generates a type of joint overload risk, and assigns a third physical risk probability.

[0019] The risk assessment module also performs composite risk identification by constructing a wind speed-vibration joint assessment matrix. The root mean square value of the wind speed and the amplitude of the main vibration at the current moment are used as two-dimensional inputs and mapped to the preset wind speed-vibration joint assessment matrix. The joint assessment matrix is ​​divided into four quadrants: safe zone, attention zone, warning zone and danger zone. When the mapping point falls into the warning zone or danger zone, the risk judgment module will weight and fuse the first physical risk probability and the second physical risk probability to generate a comprehensive physical risk probability, and combine the fall-related risk and the structural instability risk into a comprehensive risk category for high-altitude falls.

[0020] Example 2: Please refer to Figures 1-2As shown, the timing management and prediction module can record the scaffolding robot's construction process and obtain timing elements, including construction location, construction time, and construction completion time. The timing management and prediction module uses the scaffolding robot to complete the construction of a single scaffolding node as a positioning benchmark, and records in real time the number of nodes that have been completed, the number of nodes that have been built, the actual start time of the current node, and the time elapsed for the current node. The module then stores the above information in the construction log sequence according to the timestamp. The time series management prediction module extracts the historical construction time of each node based on the construction log sequence, calculates the historical average construction time and the historical standard deviation, and calculates the number of remaining nodes to be built based on the current construction node number and the total number of planned nodes. The historical average construction time, the historical standard deviation, the number of remaining nodes to be built, and the current time elapsed are combined into a time series feature vector.

[0021] The risk assessment module performs time-series risk analysis based on time-series elements, and generates time-series risk types and probabilities using the following method: The risk assessment module calculates the installation progress advance / lag index based on the time series feature vector as follows: The risk assessment module uses the ratio of the planned total construction time to the historical average time as the benchmark progress rate, and combines the remaining number of nodes to be built and the current time already consumed to predict the actual remaining completion time. The difference between the actual remaining completion time and the planned remaining time is calculated to obtain the schedule deviation. When the schedule deviation is greater than the positive deviation threshold, it is determined that there is a risk of installation timeout, and the time series risk type is generated as schedule lag risk. The probability of the time series risk is calculated by multiplying the ratio of the schedule deviation to the positive deviation threshold by the basic timeout probability. When the schedule deviation is less than the negative deviation threshold, it is determined that there is a risk of excessive schedule, and a time-series risk type is generated as excessive schedule risk, and a corresponding time-series risk probability is assigned.

[0022] After receiving the risk of schedule lag and the corresponding probability of timing risk, the risk proactive correction module calculates the required speed compensation coefficient based on the amount of schedule deviation. The speed compensation coefficient is equal to the ratio of the planned remaining time to the actual remaining completion time. The speed compensation coefficient is constrained between the preset maximum safe speed ratio and the minimum safe speed ratio, and a speed adjustment command is generated and sent to the motion controller of the scaffolding robot. The proactive risk correction module re-estimates the corrected actual remaining completion time and the corrected schedule deviation based on the update rate after speed compensation, and recalculates the corrected timing risk probability based on the corrected schedule deviation. The corrected timing risk probability and the corrected timing risk type are then sent to the output warning module.

[0023] The proactive risk correction module is also equipped with a speed smooth transition strategy; The speed smooth transition strategy decomposes the change in the speed compensation coefficient into multiple step increments. Only a single step increment is executed in each control cycle, so that the robot's movement speed gradually changes to the target compensation speed according to the preset acceleration slope. At the same time, during the speed adjustment process, the multi-source sensing module continuously monitors the actual load torque of each joint. If the actual load torque of any joint exceeds the preset safety margin ratio of the rated load limit, the current speed compensation coefficient is immediately locked to prevent further increase, and the speed lock status is sent to the output warning module as additional correction information. The output early warning module analyzes the physical risk type and probability and generates a physical risk early warning. At the same time, it provides risk early warning based on the time-series risk probability and type sent by the risk active correction module.

[0024] The output warning module outputs an audible and visual alarm signal with the corresponding color level based on the risk level range of the comprehensive physical risk probability or the third physical risk probability. The output early warning module generates progress deviation prompts based on the received corrected time-series risk probability and the difference between the current number of completed nodes and the planned number of nodes, and displays them visually in the form of a progress bar. The output warning module displays physical risk warning information and temporal risk warning information synchronously in the same display interface, and triggers the highest level comprehensive emergency alarm when both physical risk and temporal risk reach the warning level or above, while outputting a braking suggestion signal.

[0025] In summary, this invention simultaneously collects and extracts features from wind speed, vibration, and joint loads. Through analysis using a wind speed-vibration joint discrimination matrix, it achieves a comprehensive assessment and weighted fusion of risks associated with high-altitude falls and structural instability. This effectively identifies the combined risks of gust impacts and structural resonance. Furthermore, it can record the erection time and extract temporal feature vectors based on individual scaffold nodes, quantitatively assess the risk of installation timeouts, and automatically adjust the robot's operating speed. Within a limited working window, it proactively avoids construction timeout failures and improves the time controllability of the erection task.

[0026] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A fault prediction system for scaffolding robots, characterized in that, It includes a multi-source element perception module, a time-series management and prediction module, a risk identification module, a proactive risk correction module, and an output early warning module; The multi-source sensing module can classify the operation process of the scaffolding robot into elements and collect multi-source sensing data to obtain physical multi-element elements, including wind speed elements, vibration elements, and joint load elements. The timing management and prediction module can record the scaffolding robot's construction process and obtain timing elements, including construction location, construction time, and construction completion time. The risk discrimination module can perform physical risk analysis based on physical multi-source factors and time-series risk analysis based on time-series factors, and generate physical risk types, physical risk probabilities, and time-series risk types and time-series risk probabilities respectively. The risk active correction module actively analyzes the time-series risk type and probability, generates a correction signal, corrects the time-series risk type and probability, and sends the corrected time-series risk probability and type to the output early warning module. The output early warning module analyzes the physical risk type and probability and generates a physical risk early warning. At the same time, it provides a risk early warning based on the time-series risk probability and type sent by the risk active correction module.

2. The fault prediction system for scaffolding robots according to claim 1, characterized in that, The multi-source sensing module collects the wind speed time-series signal of the work area in real time at a fixed sampling frequency, and performs sliding window filtering and outlier removal on the wind speed time-series signal to generate a standard wind speed feature vector. The multi-source sensing module collects the time-domain signal of the robot's three-dimensional vibration acceleration, extracts the frequency-domain amplitude features through fast Fourier transform, and concatenates the time-domain and frequency-domain features into a comprehensive vibration feature vector. The multi-source sensing module collects the output torque value of each joint in real time, and calculates it by combining the reduction ratio and transmission efficiency of each joint to obtain the actual load torque vector of each joint.

3. The fault prediction system for scaffolding robots according to claim 1, characterized in that, The risk discrimination module calculates the standard wind speed feature vector to obtain the wind speed risk judgment quantity, performs threshold judgment on the wind speed risk judgment quantity to determine whether there is a risk of gust impact, and generates a fall-related risk type and assigns a first physical risk probability. The vibration comprehensive feature vector is analyzed to obtain the vibration risk assessment quantity. The vibration risk assessment quantity is then analyzed to determine whether there is a type of structural instability risk and assign a second physical risk probability. The actual load torque vector is thresholded to determine whether there is a risk of joint overload and to assign a third physical risk probability.

4. The fault prediction system for scaffolding robots according to claim 1, characterized in that, The risk assessment module also performs composite risk identification by constructing a joint assessment matrix of wind speed and vibration. Using the current wind speed risk assessment and vibration risk assessment as two-dimensional inputs, they are mapped to a preset wind speed and vibration joint discrimination matrix. The joint discrimination matrix is ​​divided into four quadrants: safe zone, attention zone, warning zone, and danger zone. When the mapping point falls into the warning zone or danger zone, the risk discrimination module will weight and fuse the first physical risk probability and the second physical risk probability to generate a comprehensive physical risk probability.

5. The scaffolding robot construction failure prediction system according to claim 1, characterized in that, The timing management and prediction module uses the scaffolding robot to complete the construction of a single scaffolding node as a positioning benchmark, and records in real time the number of nodes that have been completed, the number of nodes that have been built, the actual start time of the current node, and the time elapsed for the current node, and stores the above information in the construction log sequence according to the timestamp. The time series management prediction module extracts the historical construction time of each node based on the construction log sequence, calculates the historical average construction time and the historical standard deviation, and calculates the number of remaining nodes to be built based on the current construction node number and the total number of planned nodes. The historical average construction time, the historical standard deviation, the number of remaining nodes to be built, and the current time elapsed are combined into a time series feature vector.

6. The scaffolding robot construction failure prediction system according to claim 1, characterized in that, The time-series management prediction module uses the ratio of the planned total construction time to the historical average time as the benchmark progress rate, and combines the remaining number of nodes to be built and the current time already consumed to predict the actual remaining completion time; the difference between the actual remaining completion time and the planned remaining time is calculated to obtain the progress deviation. The timing management prediction module performs a threshold judgment on the progress deviation to determine whether there is an installation timeout risk, generates a timing risk type as progress lag risk, and calculates the timing risk probability by multiplying the ratio of the progress deviation to the positive deviation threshold by the basic timeout probability.

7. The fault prediction system for scaffolding robots according to claim 1, characterized in that, After receiving the schedule lag risk and the corresponding timing risk probability, the risk active correction module calculates the required speed compensation coefficient based on the schedule deviation, and constrains the speed compensation coefficient between the preset maximum safe speed ratio and the minimum safe speed ratio, and generates a speed adjustment command to send to the controller of the scaffolding robot. The proactive risk correction module re-estimates the corrected actual remaining completion time and the corrected schedule deviation based on the update rate after speed compensation, and recalculates the corrected timing risk probability based on the corrected schedule deviation.

Images